The long nanotube strands created by the direct synthesis technique is a good alternative to the fibers and filaments spun from nanotube slurries. The mechanical and electrical properties of these strands are indeed superior to the latter fibers. The strands can be produced in high yield and continuously. The thickness of the strands and their length may be further optimized, by tuning the processing conditions, to produce practically useful nanotube based devices, such as strong, highly conducting micro-cables or mechanically robust electrochemical micro-actuators.
These very long nanotube strands are handled and manipulated easily (see inset of FIG. 4 above, where the strand is tied into a knot) suggesting that these are not as brittle as nanotube aggregates prepared by other techniques. The Young's modulus estimates for these structures from the direct tensile tests are not as high as values expected for individual nanotubes or small nanotube bundles. But the numbers obtained for the modulus are lower bound estimates due to the uncertainty in knowing the exact cross-sectional area of the strands, supporting load. The nanotube strands are not monolithic structures and consist of parallel nanotube ropes separated by interstitial space.
The present inventors noted that without hydrogen flow, the yield of nanotube growth rapidly decreases and no long strands were produced. Replacing n-hexane with other hydrocarbons results in lower yields of single-walled nanotubes and once again no macroscopic self-assembled strands were created.
From the SEM measurements, the present inventors estimate that the approximate volume fraction of nanotubes in the strands is less than 48 percent by analyzing the spacing between the nanotube ropes in the strands. If one considers only this cross-sectional area supporting the load during the tensile test, the modulus values for the strands will jump to about 100 to about 150 GPa, consistent with published literature on the modulus of large SWNT bundles. Although individual SWNT has elastic modulus of about 1 TPa, the value can fail to about 100 GPa for nanotube bundles, due to the inter-nanotube defects (for example, imperfect lattice of nanotube bundles due to different nanotube diameters) present along the bundles.
FIG. 1 illustrates a CVD apparatus used for the floating catalyst method according to a preferred embodiment. The apparatus contains a reaction chamber 1. Preferably, the reaction chamber 1 is a ceramic tube (outer diameter: 68 mm, inner diameter: 58 mm, and length 1600 mm), which is mounted vertically inside the electronic furnace 2, having a rated temperature of 1200.degree. C. and a rated power 6 kW, for example. Other temperature and power ratings may also be used.
An evaporator 3 is located in an upper part of the reaction chamber 1. The evaporator contains a first inlet 4 for gases, such as the carrier gas, and a second inlet 5 for the nanotube source solution, such as a mixture solution of n-hexane, ferrocene and thiophene. The mixture solution is pumped from a storage vessel 6 via a liquid microflow rate pump 7 into the evaporator 3. The evaporator is adapted to vaporize the solution into the carrier gas at a temperature of about 150 to about 200.degree. C. The carrier gas carries the evaporated solution into the reaction chamber 1 in the form of a gas phase. The CVD apparatus also contains a collector 8 at the bottom of the reaction chamber where the carbon nanotubes are collected, a filter 9, and a gas outlet 10 connected to the bottom portion of the reaction chamber 1. However, the present invention should not be considered limited to the apparatus shown in FIG. 1 and other suitable CVD apparatus may be used instead.
A floating catalyst method is now be described. Thiophene and ferrocene are dissolved in liquid n-hexane in the storage vessel 6 and the mixture solution of liquid n-hexane containing the thiophene and ferrocene is then provided (i.e., sprayed) into the hydrogen carrier gas stream in the evaporator 3. The liquid n-hexane containing the thiophene and ferrocene is evaporated (i.e., vaporized) and fed from the top into the heated vertical reaction chamber 1. The hydrogen flow rate is adjusted to provide the optimum conditions for nanotube strand formation. The n-hexane is catalytically pyrolized in the reaction chamber 1 to form the carbon nanotube strands, and the gas flow carries the strands downstream through the reaction chamber. The nanotube strands are collected in the collector 8 at the bottom of the reaction chamber 1. The nanotube strands are then removed from the collector. Preferably, ferrocene is directly dissolved in the n-hexane and flowed into the evaporator along with the carbon source solution, without the pre-reduction under hydrogen atmosphere so as to simplify processing. However, if desired, ferrocene may be pre-reduced under a hydrogen atmosphere.
The following process ranges and parameters are provided for illustration of the preferred embodiments: First, a pre-processing gas, such as argon gas, is passed through the reaction chamber 1, at the rate of 100 ml/min, while the furnace 2 temperature increases to the reaction temperature. The flowing argon gas is switched to flowing hydrogen gas when the temperature reaches about 1000.degree. C.
After reaching the preset reaction temperature of about 1100.degree. C. to about 1200.degree. C., the n-hexane, ferrocene and thiophene mixture solution was introduced into the reaction chamber 1 together with the hydrogen gas through evaporator 3 so as to begin the reaction. Preferred ferrocene range in the mixture solution is about 0.01 to about 0.02 g/ml. Preferred thiophene ranges in the mixture solution is about 0.2 about 0.6 weight percent. The term "about" allows a deviation of about 5%, preferably about 10% from the stated value. The preferred flow rate of the mixture solution is about 0.2 to about 0.8 ml/min, while the preferred flow rate of the hydrogen gas is about 150 to about 300 ml/min. The reaction is conducted for a preset period of time, such as 30 to 90 minutes, preferably about 60 minutes, to generate the nanotube strands and then terminated by flowing argon gas at about 100 ml/min instead of flowing hydrogen. The nanotube strands are then collected after cooling the reaction chamber and furnace to room temperature. Other suitable temperature, concentration and flow ranges may also be used.
The morphology and microstructures of the products were examined using SEM (shown in FIG. 2) and TEM (shown in FIG. 3). The diameter distribution and the degree of crystallization of produced nanotubes were examined using micro Raman. The results indicate that the floating catalyst method in a vertical furnace can be used to realize mass production of the macroscopic or super long continuous single walled carbon nanotubes. The products consist of a large amount of densely packed aligned single walled carbon nanotube strands, in which the single walled nanotubes can be up to 20 to 30 cm in length. The diameter of nanotubes is in the range of 1-2 nm and the highest purity of single walled carbon nanotubes can be up to 85 to 95%.
Figure 2
The macroscopic strands produced using n-hexane and hydrogen carrier gas contain about 5 wt % impurities, which can be analyzed by thermo-gravimetric analysis (TGA) to be catalyst (Fe) particles and amorphous carbon. Majority of the catalyst particles can be removed by high temperature vacuum annealing or refluxing the strands in nitric acid for several minutes.
