Introduction

During the synthesis of carbon nanotubes, small changes in the process conditions can have large effects on the morphology of the carbon nanostructures. Carbon can form nanotubes or nanofibres. Nanotubes are vertically aligned cylindrical sheets of graphite, while nanofibres and bent and tangled cylindrical graphite structures (Xiong 2006). In 2006, Hart and Slocum wrote a paper discussing the fact that nanotube morphology varies at different growth locations; they called these morphology variations ‘loading effects.’ Pictures of their nanotubes show that because flow rates of argon over the sample differed depending on the location, the nanotubes grew differently. The flow rate for argon is just one of many process conditions that affect the structure of nanotubes. Understanding how the changing of different process conditions affects the growth and structure of nanotubes is necessary for the production of nanotubes that are optimal for certain conditions (Hofmann et al, 2005).


There are three separate processed necessary in nanotube synthesis and each process can be done using many different techniques. A necessary component of nanotube growth is the presence of a first-row transition metal to catalyze the formation of tubular graphite sheets, which are the nanotubes. The metal can be deposited onto a substrate through sputtering or electron beam deposition (Grobert, 2007).


The metal catalyst must then be heated so that it anneals. This process transforms the fairly flat layer of metal to ‘nanoparticles.’ These are small patches of metal that catalyze nanotube growth. Annealing can be done with heat, called thermal annealing, or by exposing the metal to a plasma environment, caused plasma treatment (Grobert, 2007).

Top

After the annealing the actual growth must take place. This can happen in many different growth environments including; chemical vapour deposition, laser vaporisation, and carbon arc discharge. (Grobert, 2007) During our experiments into the morphology of nanotubes, iron catalyst was sputtered onto a substrate (a silicon wafer treated with 10 um of silicon dioxide); thermally annealed, and carbon was deposited using chemical vapour deposition (CVD).


Nanotubes grow on nanoparticles of iron, and these small iron particles can only be formed with small amounts of iron. Sputtering allows for the accurate deposition of small amounts of iron. The iron source in the sputtering chamber acts as a resistor in an electric circuit and because of this, it heats up to the point that it sublimes and ionizes into plasma in the sputtering chamber. The iron plasma travels through the grounding shield, a metal guard for the metal source, and onto the silicon wafers. The sputtering occurs at a low vacuum about 10-3 torr but in a high partial pressure of argon rather than contaminant gasses like oxygen (Sanderson, 2007).


The next necessary process in the production of nanotubes is annealing the iron catalyst. Annealing takes the relatively flat layer of iron and forms the nanoparticles that catalyze the nanotube growth. The nanoparticles can be formed through the application of heat to the iron, or exposing the iron to plasma. The heat or plasma treatment provides the iron atoms with enough energy to break the metallic bonds that hold them in place. The iron atoms then move around and eventually recombine into little iron droplets in a process that is similar to the formation of water droplets on a rain jacket or a car hood.


Thermal annealing is a much more cost effective than plasma annealing, which makes it more useful in the mass production of carbon nanotubes. Common practice has the thermal annealing taking place in a Rapid Thermal Processor (RTP). The RTP is used because it increases the temperature of the wafers and the iron by about 30ºC per second (George, 2007). Plasma annealing must take place in a low vacuum because ionizing all the gas particles in the chamber at atmospheric pressure would require excessive amounts of energy (Hart, 2006). The machinery required to create a vacuum and the plasma is much more expensive than the RTP, making the thermal annealing for cost effective.

Top

The rapid increase of temperature is important because the annealing would still occur at temperature below the set temperature, and because the substrate and iron would spend a decent amount of time in these temperatures, this would affect the recordings for the amount of time spent in the annealing process. The annealing rate would vary with the temperature and with a constantly but slowly changing temperature the annealing measurements would be almost impossible to duplicate these experiments.


The nanotubes can grow though carbon arc discharge, laser vaporization, or chemical vapour deposition (CVD) (Grobert, 2007). However the most effective both in terms of time and cost is the chemical vapour deposition (Hofmann et al., 2005) (Bronikowski, 2006), but even this can be done using different techniques. These include, basic chemical vapour deposition (CVD) and plasma-enhanced chemical vapour deposition (PECVD). In Hofmann study in 2005 their PECVD occurred at pressure of less that 10-6 mbar, which requires costly equipment to create such a vacuum and time to prepare the vacuum. In 2006 Karwa et al. performed experiments where they grew nanotubes inside stainless steel tubes using regular chemical vapour deposition. The carbon source, ethylene, was pumped through the heated stainless steel tubes and deposited on the annealed nanoparticles on the inside of the tube, creating nanotubes. This is a much simpler than the PECVD, and this is why we opted for CVD instead of PECVD.


A necessary component of any CVD is the etching of the metal catalyst before and during synthesis (Musso, 2005). Most metals react very well with oxygen, but these reactions clog the reaction sites that the carbon species uses to form the graphite sheets of nanotubes. Etching agents eat away at the metal-oxygen species allowing for more productive nanotube growth (Juang, 2004). Etching agents include H+ ions, H2O which becomes H+, ammonia (NH3), and in PECVD, plasma (Grobert, 2007) (Hart, 2006) (Hofmann, 2005). Liu J’s study in 2005, showed that changing the flow rate or amount of NH3 present during the CVD can affect the temperatures at which vertically aligned nanotubes forests grow.


The basic procedure we followed used a silicon wafer treated with 10 micometres of silicon dioxide as the substrate. We sputtered a linearly increasing wedge of iron that went from 0 to 5 nanometres thick. We then thermally annealed the iron and the substrate at 550ºC for 10 minutes. After this, we grew the nanotubes in regular chemical vapour deposition, using ethylene as the carbon source and ammonia as the etching agent. The gas ratio was 3:1 ethylene to ammonia for a total of 25 sccm for 20 minutes. Following the growth we made measurements of the nanotube morphology from pictures taken with a scanning electron microscope. The sputtering experiment manipulated the amount of iron deposited and the rate by which it was deposited on the silicon substrate. The experiment into the annealing conditions looked at changing the length of time and the temperature at which the iron was annealed at. The chemical vapour deposition experiment, varied the temperature and the time of the vapour deposition. Apart from the experimental conditions the samples were synthesized in the basic parameters described above.

Top