Disscusion and Conclusions
Sputtering Experiment
Rapid Thermal Processor Experiment
Chemical Vapour Deposition
Sputtering Experiment
The results of this experiment show that the height of the carbon nanotubes reach a maximum point between 3 and 4 nm of iron catalyst thickness (expected), which is good evidence that supports the reasoning of the Dalhousie research group. CNT’s were grown on silicon wafers with an iron catalyst thickness of 0 – 5 nm for most of their experiments, as they are looking for long nanotubes. The longer the nanotubes are, the higher the surface area per capita, and more Nafion catalyst can be supported by the nanotubes. However, these nanotubes are also very closely packed together, which means that the Nafion doesn’t necessarily cover much of the nanotube, only the top area. At five to six nanometers of catalyst thickness, the nanotubes of the three samples used in this project remain relatively long compare to their diameter, but they are also more separated than those grown on iron thickness of 0 to 5 nm. Separation between nanotubes also increased as the catalyst thickness grew, making short, spaced nanotubes with large diameters. In sample one, the nanotubes are spaced 30 – 80 nm apart at 5 nm of iron catalyst, with a length of 2.4 µm. The spacing in sample two at 4.93 nm (3.2 µm in height) and sample three at 5.11 nm (3.85 µm in height) of catalyst is 50 – 100 nm. In a sample grown prior to this experiment (titled “Gary’s sample”), there is a spacing of 20 – 60 nm between nanotube tops at 4.5 nm of catalyst thickness. These measurements suggest that optimal spacing and length occurs when the thickness of the iron catalyst is around 5.0 nm, and that the conditions of sample three (15 W, RTP 550°C, 5°C/s) may be best to obtain the best results in Nafion deposition, as the nanotubes were longer and spaced further apart than the other samples at this catalyst thickness.
One problem that was discovered with the use of an iron catalyst thickness of 0 to 10 nm is that when the samples were put through the intense heat and light of the RTP process, large patches of the iron oxide peeled off, making areas on the sample where no carbon nanotubes could be grown. This happened for all of the samples put through RTP for my experiment, producing only one useable sample for each run (the samples that had the least amount of iron peel-off, as too much peel-off would affect how the carbon source would have been deposited). A possible cause for this anomaly is that when the iron catalyst is deposited in the sputtering chamber, ‘piles’ of iron atoms gather on the silicon strips instead of the iron falling in the linear way it should due to the amount being sputtered or the power that it is sputtered at. When the iron is heated in the RTP, these clumps expand in size, more so than the silicon oxide that they are attached to. This is because the coefficient of thermal expansion is larger for the iron oxide (12 x 10-6 K-1) than that of glass (3 x 10-6 K-1), meaning that the material expands more in length and width when heat is applied to it. When the samples are rapidly cooled, the iron oxide cracks and chips off of the silicon substrate, leaving patches where no carbon nanotubes can be grown. Without the iron oxide ‘islands’, the carbon atoms have nothing to land on and attach to, and carbon nanotubes cannot be grown. We attempted to correct the peel-off problem by lowering the heating rate from 30°C/s to 5°C/s, but the cooling rate of the RTP run was not changed, and remained constant at 20°C/s for each sample run.
RTP peel off from sample 1: the striped iron oxide can be seen at the "high end" of the silicon wafer, or where the iron catalyst was applied the thickest.
As seen in Fig.1 and Fig.2, the carbon nanotubes grown on an iron catalyst substrate of increasing linear thickness from 0 to 10 nm have a peak height at around 3.5 to 5 nm of catalyst thickness and increase linearly in diameter.This shows that the longest nanotubes can be grown on substrates of thickness 0 to 5 nm and that thicker nanotubes result from thicker Fe catalyst applications. The heights of the nanotubes peaked at an average iron catalyst thickness of 3.2 nm, and the diameter of the nanotubes increased to 70 – 90 nm at 10 nm of catalyst thickness for all of the samples. It was also discovered that the spacing between the nanotubes grown on the iron catalyst that increased from 0 to 10 nm was larger than the spacing between the nanotubes on samples grown on iron catalyst 0 to 5 nm thick. This research supplies reason for only growing the nanotubes on iron catalyst 0 to 5 nm thick when looking for a maximum yield of long nanotubes because the heights of the tubes plateau between 3 and 4 nm of catalyst thickness. However, it also shows that thicker nanotubes with larger spacing between them can be grown when using a Fe-film thickness of 0 to 10 instead. Further experiments can be derived from the results of this one, including making substrates of iron catalyst thickness that keep an average constant thickness of about 4 nm (average thickness at which height of nanotubes peaks) or of 9.5 nm (average thickness at which the maximum diameter and spacing between nanotubes is obtained) to discover whether or not the growth of desired nanotubes can be controlled.
Top
Rapid Thermal Processor Experiment
When comparing the differences in height for the samples that were annealed at 550ºC, it is observed that the height of Fe catalyst thickness from where height measurements can taken varies between the samples. On the sample annealed for 10 minutes, vertically aligned nanotubes only start appearing after when the iron catalyst reaches an initial height of 1.5 nanometres. Below this level the carbon formed a tangled mass of nanofibres that are quite useless. For the sample that was annealed for 20 minutes there are vertically aligned nanotubes at just over 1 nanometre of iron catalyst thickness. The sample annealed at 550ºC for 30 minutes had vertically aligned nanotubes with only .5 nanometres of iron catalyst. One explanation for this result could be that 550ºC does not give the iron atoms that much energy so when they are exposed to that temperature for only 10 minutes when there is not as much iron present the atoms cannot move around enough to form suitable nanoparticles. Exposing the silicon wafer to the same temperature but for a longer period of time, this allows the iron to move time to move around and form the nanoparticles essential for nanotube growth. This phenomenon can also explain why the height of the nanotubes from the sample that was annealed at 550ºC for 30 minutes plateaus at a fairly different location from the other two samples annealed at the same temperature. The iron atoms may have moved from areas with initially more iron to places with smaller initial amounts of iron. Because of the movement of iron, the plateau placement will be different from its placement in the other samples corresponding to the same trend in height but phase shifted slightly.
By comparing the heights of nanotubes annealed at different temperatures for the same amount of time, it can be seen that the main factor regarding the pattern for nanotube height is the length of time that the iron is annealed for. This is different from the apparent relationship between annealing temperature, time exposed, and the diameter of the nanotubes. The nanotube diameter gets more consistent the longer the sample is annealed for. The 10 minute sample has huge variation in the diameter while nanotubes from the 30 minute sample have much more consistent diameters all along the silicon strip. This trend suggests that the longer the iron is annealed for the more consistent the diameter of the nanotubes will be. In the sample that was annealed at 720ºC, the diameter for the nanotubes is uniform except at one point. This point is one that does not fit the curve for the height data, so it may be a situational anomaly. The fact that the initial height of the iron catalyst has, apparently, no effect on the diameter of the nanotubes from the sample annealed at 720ºC suggests that increasing the temperature of the annealing increases the diameter consistency of nanotubes. Since both the time exposed to the annealing temperature and the temperature of the annealing contribute energy to the iron atoms; the results of this experiment suggest that nanotubes grown on iron that has received more energy during the annealing process have more uniform diameters.
More trials have to be done with annealing at 720ºC, preferably trials that anneal the iron at this temperature for 15, 20, and 30 minutes, in order to explore the effect that RTP annealing has on the growth of nanotubes. However, during these trials pictures should be taken of two sites that have the same initial catalyst thickness to eliminate that chance of situational errors affecting the data. With more experiments into the effect that different growth conditions have on nanotube morphology. This knowledge is essential to produce nanotubes that are ideal for certain functions.
Top
Chemical Vapour Deposition Experiment
Results collected indicate that temperatures affect the general morphology of carbon nanotubes. For the sample at the lowest temperature, no carbon nanotubes were grown, suggesting that there was not enough thermal energy in the system to favour graphitization. The sample grown at the middle temperature produces carbon nanotubes very similar to the carbon nanotubes grown previously at a temperature a little bit lower. The nanotubes on the middle temperature are a little bit larger than those grown on the one with a lower temperature, suggesting that the size of the carbon nanotubes grown is effected by slight variations in temperature. This theory is also supported by the data from the sample at the highest temperature, where the carbon nanotubes were up to five time bigger than the samples at the lower temperatures. High temperatures also produce a mix of carbonization and graphitisation leading to carbon species, such as the toilet paper roll growth and a thick crap layer.
The higher the temperature of the furnace, the more thermal energy is delivered to making radicals. The more carbon radicals are present, the longer the carbon nanotubes will grow. However, at high temperatures, there wasn’t enough ammonia to make the active sites for the excess radicalized carbon radicals, so the radicals formed in carbonization in non-nanotube carbon species.
Top