SWNTs have a number of important properties, including variable bandgap, high current-carrying capacity, hydrogen absorption, and field emission. In most materials, the bandgap is an inherent property; however, a study by Wilder et al. (1998) showed that a nanotube’s bandgap is inversely proportional to its diameter, which opens up the possibility of choosing a nanotube’s bandgap. Stable current densities of carbon nanotubes are as high as 109 A/cm2 (Collins et al. 2000). In contrast, copper wires burn out at about 106 A/cm2. SWNTs have been observed to absorb up to 8% hydrogen by weight (Dillon et al. 1997), making them a feasible form of hydrogen storage for fuel cells. Nanotubes have been found to be excellent field emitters, making them suitable for flat panel display applications (Deheer et al. 1995).

However, the physical properties of nanotubes are still being discovered and disputed. The difficulty is that nanotubes have a very broad range of electronic, thermal, and structural properties that change with diameter and chirality, and diameter and chirality cannot be adequately controlled during synthesis. Yet, because of the promise that nanotubes hold, it is essential to overcome these obstacles and attempt to gain an accurate understanding of these properties.

In this experiment, four SWNT samples were produced using various pulsewidths. The effects of peak pulse power on SWNT type are analyzed and discussed. UV-Vis-NIR spectra were taken for the samples, and their absorbance properties compared. Pump wavelengths were measured, and their effect on decay times determined.

SWNTs were synthesized with a 755 nm alexandrite laser using the laser pulse synthesis technique described by Dillon et al. (1999). The pulsewidth was varied with runs at 2.0 ms, 1.0 ms, 500 ns, and 200 ns. In shortening the pulsewidth, a higher peak power is achieved. It will be interesting to see if this change in the synthesis process has any effect on the yield or optical properties of the SWNTs. The carbon targets contained 2.78% m.w. cobalt and 2.77% m.w. nickel, which are both metal catalysts in SWNT production. Argon was passed through the 1200° C chamber at a rate of 100 sccm, and pressure was held at 620 Torr. The argon provided an inert atmosphere necessary for nanotube growth. The rastor rate was 0.10300 Hz x 0.0400 Hz (approximately a 50% overlap). The power density was, on average, 27 W/cm².

The samples were then refluxed with 100 mL of deionized water and 24 mL of 3M HNO₃ at 130° C for 16.5 hours. During this time, the sample dispersed and metals became ionized. The metal ions were filtered out using 0.2 mm alumina filter and deionized water. The sample and filter were dried in a 50° C oven for 30 minutes, making it possible to separate them from each other. The sample was then baked in a 550° C furnace for 30 minutes to remove the carbon matrix via oxidation. For purposes of optical characterization, the SWNTs were cast into 0.1% m.w. films in nafion perfluorinated ion-exchange resin purchased from Aldrich.

                    m_final = x*(M_NiO) + x*(M_CoO)                                                  (1)

   m_final = x*(M_NiO) + x*(M_CoO1.33)                                             (2)

From x, the mass of nickel and cobalt in the original sample can be calculated and the metal content (wt/wt %) can be found using (3):

                   Metal Content = (x*M_Ni + x*M_Co)*100/ m_initial                          (3)

Linear absorption measurements were carried out using a Cary 500 double beam spectrometer at a spectral resolution of 1 nm.

Transient absorption measurements were taken with a Clark-MXR CPA-2001 regeneratively-amplified Ti:sapphire laser operating at 989 Hz. The 775 nm output pulses pump an optical parametric amplifier. b-BaBO4 (BBO) was used to double the frequency of a fraction of the 775 nm beam. By focusing a small portion of the Ti:sapphire output on a 2 mm sapphire window, white light probe pulses ranging from 440-950 nm were generated. Probe pulses can be delayed up to 1300 ps relative to the pump. The pump and probe pulsewidths are 200 and 125 fs, respectively.

Table 1. UV-Vis NIR spectra of SWNTs of varied pulsewidth in nafion vs. air.

UV-Vis NIR spectra were taken of each of the samples and a nafion blank in reference to air (Figure 1). Due to technical difficulties, the Cary 500 spectrophotometer did not record absorbance in the 587-647 nm range for all samples.

Figure 1. UV-Vis NIR spectra of SWNTs of varied pulsewidth in nafion vs. air.

Dynamics spectra were taken for the sample synthesized with a 2.0 s pulsewidth using transient absorption spectroscopy. One long-range dynamical measurement was taken where the delay was varied from 0.5 to 1300 ps, and the sample was pumped at 450 nm and probed at 650 nm. The sample experienced fast decay followed by a slower one. The faster decay was fitted to a double exponential function, indicating two separate but simultaneous decay processes. The function was fit from 0.116 ps to 10.2 ps, and gave time constants 140 fs and 960 fs occurring at a ratio of 23:1. The slower decay was fit to a single exponential function from 10.2 ps to 1300 ps, and had a decay time of 120 ps (Figure 2).

Two further dynamics spectra were taken from –0.5 to 10 ps delay. One of these was also pumped at 450 nm, but probed at 530 nm (Figure 3). A double exponential curve was fitted from 0.140 ps to 9.7 ps and the time constants were calculated to be 130 fs and 940 fs in a 24:1 ratio. Figure 4 shows the spectrum of the sample being pumped at 550 nm and probed ant 640 nm. The time constants were calculated to be 150 fs and 990 fs in a 12:1 ratio.

Figure 3 . Transient absorption dynamics spectra of sample LP020717R (generated with a 2.0 s pulsewidth) pumped at 450 nm and probed at 530 nm from –0.5 to 10 ps delay.

Figure 4 . Transient absorption dynamics spectra of sample LP020717R (generated with a 2.0 s pulsewidth) pumped at 550 nm and probed at 640 nm from –0.5 to 10 ps delay.

The two samples with the shorter pulsewidths, 200 ns and 500 ns, show distinctly less absorbance than the 1.0 and 2.0 s pulsewidths. It is possible that the high-energy intensity attributed to such short pulsewidths could produce less absorbent nanotubes.

However, due to the similar absorbance spectra for all the samples, no structural differences can be derived from the UV-Vis NIR spectra. This is an important result because it suggests that the peak pulse power in laser vaporization synthesis does not affect the types of SWNTs produced.

In all three dynamics spectra from transient absorption spectroscopy measurements, two very quick decays are observed on the order of 100 and 900 fs. In each case, these decays are in all probability due to the same events. The attribution of these time constants to a physical process is challenging. However, based on the dynamics observed in other nanomaterial systems, the fastest decay may correspond to charge carrier cooling, and the 900 fs decay may represent the delocalization of charge along the nanotube. The slower decay of 120 ps, seen in Figure 2a, can be attributed to electron-hole recombination.

In comparing figures 2 and 3, we see that for the same pump wavelength of 450 nm, the higher energy probe wavelength of 530 nm resulted in quicker decay times and a slightly increased influence of the faster decay. More transient absorption measurements can be taken with varying probe energies in order to derive a more specific relationship. In Figure 4, we see that for a less energetic pump pulse, both decays become longer and that the influence of the 990 fs decay increases dramatically. This could be the result of different tube types being excited by different photon energies of the pump pulse. However, it is not possible to draw definitive conclusions here as the absorption data in Figure 1 is inconclusive in the pumping region.

The main conclusions from this stage of the project are that the UV-Vis-NIR spectra show no significant structural differences, although nanotubes synthesized with a shorter pulsewidth tend to show less absorption, potentially because of nanotube degradation due to high-energy intensity. Transient absorption data indicate that the two fast decays, potentially due to charge carrier cooling and charge delocalization along the nanotube, become slower with a less energetic pump pulse. Also, the decay attributed to charge carrier cooling becomes slightly more influential at a higher energy probe pulse.

This work represents progress in the measurement of time-resolved relaxation following photoexcitation. This technique will continue to be applied to reveal further details on excited state lifetime, on electronic and vibrational relaxation, on the interactions between nanotubes and other nano or bulk materials, and eventually on specific tube types.