NANO LETTERS
CVD Synthesis of Single Wall Carbon Nanotubes under “Soft” Conditions
2002 Vol. 2, No. 5 525-530
Avetik R. Harutyunyan, Bhabendra K. Pradhan, U. J. Kim, Gugang Chen, and P. C. Eklund* Department of Physics, 104 DaVey Laboratory, The PennsylVania State UniVersity, UniVersity Park, PennsylVania 16802 Received January 22, 2002; Revised Manuscript Received February 25, 2002
ABSTRACT A series of Fe and Fe/Mo catalysts, supported on alumina, were examined for Single-Walled Nanotube (SWNT) growth by Chemical Vapor Deposition (CVD) in methane. Molybdenum (20 wt %) was found to provide a clear synergistic benefit, lowering the growth temperature and eliminating the need for catalyst activation in hydrogen. The dependence of the carbon mass yield and sample quality on the CVD growth conditions is discussed. Catalyst activity in both the oxide and reduced (metallic) form were investigated under low methane flow (40 cm3/min) and in the temperature range 600 < T < 900 °C. We found that the Fe/Mo oxide catalyst was active at temperatures as low as 680 °C. Under these “soft” conditions (680 °C, 40 cm3/min methane), the Fe-oxide catalyst (without Mo) was not active for tube growth, but it could be activated by an in situ reduction to metal in flowing He/(10% H2). Tube diameters in the range 0.7 to 1.7 nm were produced under most of the growth conditions studied, as determined by Raman scattering. In most cases, the SWNTs were produced in bundles ∼10 nm in diameter, as observed by transmission electron microscopy. No evidence for coproduction of multiwalled tubes was found.
Introduction. Single wall carbon nanotubes (SWNTs) were first discovered by scientists at NEC and IBM in 1993.1,2 Today SWNTs are synthesized by three main methods: arc discharge,1-3 pulsed laser vaporization,4-6 and chemical vapor deposition.7-10 The growing interest of SWNTs for applications11 and fundamental science11-17 demands new approaches and flexibility for the synthesis. Many researchers consider chemical vapor deposition (CVD) as the only viable approach to large-scale production. As a result, research is underway to optimize the CVD growth, i.e., to investigate the effect of catalyst composition, variation of supporting/ substrate materials, synthesis temperature and hydrocarbon gases. Recently, Co-Mo metal catalysts have been found to selectively produce SWNTs at 700 °C using carbon monoxide as the carbon source.18 Fe/Mo bimetallic catalysts have also been evaluated for SWNT production from methane at high temperatures 900 °C.19 Synthesis of SWNTs at temperatures between 700 and 850 °C by catalytic decomposition of carbon monoxide and ethylene on alumina supported Fe/Mo catalysts has also been reported.9 Discovering CVD growth conditions that favor lower temperature and lower hydrocarbon flow is clearly important for the economical scale-up production. In this work, we have investigated the use of Fe and Fe/ Mo metallic and oxide catalysts for the low-temperature CVD production of SWNTs from methane under low flow (40 cm3/ min). We find that Mo exhibits a significant synergistic behavior in this regime and that Fe/Mo functions well at low temperature (680 °C), eVen as an oxide catalyst. Fe10.1021/nl0255101 CCC: $22.00 Published on Web 03/27/2002
© 2002 American Chemical Society
oxide (without Mo) is not active at this temperature, but becomes active when reduced to a metallic catalyst. The effect of the wt % catalyst loading onto the alumina support on the SWNT yield was also investigated and is reported here. Experimental Section. (Fe) or (Fe/Mo) catalysts supported on ∼2 µm diameter Alumina (Al2O3) particles were prepared following reference.19 Metal salts (99.999%, Alpha AESAR), i.e., Fe(NO3)3‚9Η2Ο, Fe2(SO4)3‚5Η2Ο, and (NH4)6Mo7O24‚ 4Η2Ο, were dissolved in methanol, and mixed thoroughly (1 h) with methanol suspensions of alumina (99.9%, Alpha AESAR). The solvent was then evaporated and the resultant cake heated to 90-100 °C for 3 h, removed from the furnace and ground in an agate mortar. The fine powders were then calcined for 1 h at 400-500 °C and then re-ground before loading into the CVD apparatus. The catalyst composition was confirmed using Energy Dispersive X-ray (EDX) analysis in a scanning electron microscope. The CVD growth of SWNT used in this work was carried out in a quartz tube flow reactor (38 mm i. d., and 90 cm long) centered in a three-zone horizontal tube furnace. Carefully weighed catalyst samples (30-80 mg) were placed in a quartz boat at the center of the reactor tube in the furnace. In some cases, a catalyst reduction step (or “activation”) was performed in situ in the CVD reactor by first passing 100 cm3/min flow of 10% H2/90% He (99.9%) at 500 °C for 10-20 h. The reducing atmosphere was then replaced by argon (99.99%) and the temperature was raised at ∼10 °C/min to the desired
growth temperature. SWNTs were then grown by passing a mixture of methane (40 cm3/min) diluted in argon (350 cm3/ min) over the catalyst at a temperature in the range 680900 °C for ∼1 h. The reactor was then allowed to cool to room temperature with argon gas flowing. The carbon product on the alumina support was then weighed to determine the carbon yield of the CVD process. We define carbon yield here as the fractional mass increase (mf - mo)/ mo, where mf and mo are, respectively, the final mass of the catalyst with carbon deposit and the initial mass of the catalyst. Of course, not all the carbon mass is in the form of SWNTs. Nevertheless, the amount of amorphous carbon detected in electron microscope images was small, and our practical definition of the relative yield is believed to provide a reasonable assessment of SWNT production in these experiments. The structure of the SWNTs was studied with transmission electron microscopy (JEOL JEM l200EX) at 120 KV. Raman Spectra were collected using a Bomem DA3 + FT Raman Spectrometer using Nd:YAG laser excitation (λ ) 1064.5 nm) at 0.4 mW power. A JY-ISA HR460 single grating spectrometer with CCD detector with a “supernotch” filter (Kaiser Optical) was used to collect Raman spectra with 488 nm excitation from an argon ion laser. All spectra were collected in air in the backscattering geometry at room temperature on the alumina-supported material taken from a cool CVD reactor. Results and Discussion. In Figure 1 we display TEM images of SWNTs grown at 900 °C in CH4 for 90 min using Feoxide (Figure 1a) and Fe/Mo-oxide (Figure 1b) catalysts. The compositions were: Fe:Al2O3 ) 1:16 and Fe:Mo:Al2O3 ) 1:0.2:16, where the Al2O3 refers to the alumina support, and the ratio a:b:c refers to the wt % loading (e.g., Fe:Mo:Al2O3 ) 1:0.2:16 indicates the wt. ratios Fe/Mo ) 5, Al2O3/Fe ) 16, etc.). Our preliminary CVD studies on the Fe/Mo system indicated that 20 wt % Mo (relative to Fe) seemed to produce the maximum synergistic benefit. All the data on Fe/Mo catalysts presented here are for the “optimized” (Fe/Mo:5: 1) weight ratio. The images in Figure 1 are included to show the typical SWNT bundle structure observed. The dark, particle-like structure near the top of Figure 1a comes from the alumina support. For both oxide catalysts at 900 °C, we observed an abundance of SWNTs. Analysis of many TEM images of SWNTs bundles produced with Fe-oxide exhibit an average bundle diameter of ∼10 nm. At high resolution, no amorphous carbon could be detected on the bundle exterior. It is interesting to note that TEM images taken on SWNTs produced at 900 °C with the Fe/Mo oxide catalyst (Figure 1b) also showed many indiVidual SWNT (see inset to Figure 1b). They exhibited an average diameter of ∼1.5 nm although large diameter individual tubes (∼3 nm) were also observed (inset Figure 1b). We did not observe the coproduction of multiwalled carbon nanotubes in any of the samples at this temperature, or at lower temperatures. In general, at the same CVD operating temperature, the bi-metallic Fe/Mo catalyst (oxide or reduced) was found to produce higher SWNT mass yields than the Fe catalyst. The synergism between Mo and Fe was very evident. At 800 °C 526
Figure 1. TEM images of bundles of SWNTs synthesized at 900 °C (40 cm3/min methane) using supported Fe (a) and Fe/Mo (b) catalysts in oxide form: (Fe:Al2O3 ) 1:16) and (Fe:Mo:Al2O3 ) 1: 0.2:16).
and 680 °C, Mo was required to activate the oxide catalyst for SWNT growth (Fe-oxide alone could not produce tubes). Only at our highest temperature (900 °C) could Fe-oxide (without Mo) produce significant quantities of SWNTs. Furthermore, for the reduced Fe and Fe/Mo catalysts, the mass yields were always higher when Mo was present in the catalyst particle. A summary of carbon mass yields (wt % carbon relative to the catalyst/alumina support) for the various CVD experiments can be found in Table 1. A second series of experiments were conducted using the same Fe and Fe/Mo-oxide catalysts. This time, they were first reduced under flowing 10% H2/90% He gas at 500 °C, reduction was found to activate the Fe catalyst at CVD growth temperatures as low as 680 °C. In Figure 2, we show TEM images of SWNTs grown at 680 °C with the reduced Fe (Figure 2a) and reduced Fe/Mo (Figure 2b) catalysts. As Nano Lett., Vol. 2, No. 5, 2002
Table 1: Synthesis Conditions, Yield, Raman Radial Breathing Modes and Corresponding Tube Diameters for the Samples Synthesized Using Fe and Fe/Mo Catalysts raman radial breathing mode peaks (cm-1) catalyst Fe:Mo:Al2O3 1y:0.2:16 metal oxides 1:0:16 metal oxides 1:0.2:16 metal oxides 1:0:16 metal oxides 1:0.2:16 metal oxides 1:0:16 metal oxides 1:0:16 reduced metala 1:0.2:25 reduced metala 1:0.2:16 reduced metala 1:0.2:11 reduced metala 1:0.2:9 reduced metala 1:0.2:6.5 reduced metala a
temp yield (°C) (wt %)
λ ) 1064 nm excitation
tube diameter (nm)b
λ ) 488 nm excitation
λ ) 1064 nm excitation
λ ) 488 nm excitation
900
21
165; 273;
163; 178; 181; 203; 306
1.46; 0.86
1.48; 1.34; 1.31; 1.17; 0.76
900
16
165; 273
N/Ac
1.46; 0.86
N/Ac
800
20
163; 178; 181; 203; 262
1.46; 1.03; 0.9; 0.86; 0.82
1.48; 1.34; 1.31; 1.17
800
12
165; 230; 260; 273; 283 no peaks
680
10
1.46; 1.03; 0.9; 0.86; 0.82
1.48; 1.31; 1.17
680
6
680
10
680
9
680
no peaks
165; 230; 260; 273; 283 no peaks
163; 181; 203
159; 262; 274; 285
N/Ac
1.52; 0.9; 0.85; 0.82
N/Ac
262; 274; 285
156; 172; 180; 201
0.9; 0.85; 0.82
1.55; 1.4; 1.33; 1.18
18
154; 262; 274; 285
162; 177; 180; 203; 258
1.58; 0.9; 0.85; 0.82
1.49; 1.36; 1.33; 1.06; 0.91
680
16
147; 160; 262
165; 177; 180; 203; 258
1.66; 1.51; 0.9
1.46; 1.36; 1.33; 1.06; 0.91
680
12
N/Ac
177; 203; 258
N/Ac
1.38; 1.17; 0.91
680
8
∼270
172; 200
no peaks
∼0.85
Reduced in flowing 10% H2/90% He at 500 °C for 20 h. b d ) 224 cm-1‚nm/(ω (cm-1) - 12 (cm-1)).21
in Figure 1, we can only observe the “free” end of either an individual tube or a bundle of tubes. The other end is buried in the support and presumably is terminated on a catalyst particle. Importantly, we did not observe any metal particles at the “free” end of the tubes, and when individual tube ends were found, they appeared to be closed with a hemispherical fullerenic cap. The addition of Mo in the reduced catalyst form seemed to produce better quality tubes, many in long straight bundles, as shown in Figure 2b. In Figure 3 we display Raman spectra (1064 nm excitation) for the reaction products produced at various CVD growth temperatures with Fe-oxide and an optimized Fe/Mo-oxide catalyst (1:0.2:16). The region between 350 cm-1 and 1300 cm-1 is not shown, as no structure was evident in this region. The strongest peak of the high-frequency tangential C-atom displacement band (or T-band) of SWNTs appears at ∼1591 cm-1.20 It should be noted that changing the excitation frequency changes the sub-population of nanotubes observed in the spectrum. This follows from the observation that the Raman scattering from SWNTs is a resonant scattering process.20-22 For a particular excitation frequency, only a small subset of tubes in the sample whose diameter (or symmetry (n,m)20) provide for strong optical absorption at this excitation frequency can resonantly scatter.22 Therefore, the changes in the Raman spectrum with excitation frequency, that are most easily detected, are observed in the low frequency region where the frequency of the radial mode bands are strongly related to tube diameter (ωr ≈ 1/d). From Nano Lett., Vol. 2, No. 5, 2002
1.4; 1.19 c
N/A, data not collected.
the Raman spectra displayed in Figure 3, we can conclude that the Fe-oxide (without Mo) produces tubes only at 900 °C. Furthermore, with Mo added to the Fe-oxide catalysts, the spectra show clearly that SWNTs could be grown at temperatures as low as 680 °C, a dramatic improvement in the catalyst performance with added Mo. Finally, it is worth mentioning that the Raman T-band spectrum shows that the optimized Fe/Mo-oxide catalyst is not active at 600 °C. The SWNT radial breathing mode frequency of an isolated SWNT has been shown theoretically to exhibit an inverse dependence on tube diameter d, i.e., ω r ≈ 1/d (isolated tube). Atomic displacements in this mode are all equal and radial (it is a true breathing mode for the tube). Typically, these radial Raman bands appear in the low-frequency range (∼100-300 cm-1). For a SWNT bundle, it has been shown that an additional constant term (12 cm-1) is needed to include the effect of tube-tube interactions within a bundle. In this case, ω r ) 12 + 224(cm-1 ‚ nm)/d.21,22 In Table 1, we display the SWNT diameters calculated from the expression above and based on the observed radial mode frequencies. Some small differences in the SWNT radial breathing mode region in Figure 3 (Fe-oxide, Fe/Mo-oxide; 1064 nm excitation; variable growth temperature) can be observed. However, no real trends in these radial mode bands can be detected for the oxide catalysts as a function of CVD reaction temperature. In Figure 4, we display only the Raman spectra for SWNTs grown with the optimized Fe/Mo-oxide catalyst (1:0.2:16) 527
Figure 3. Raman spectra for carbon materials grown at various CVD temperatures with Fe and Fe/Mo catalysts in oxide form. The CVD growth temperatures and catalyst composition are indicated above each spectrum. The T ≈ 300 K spectra were collected using 1064 nm excitation with carbon still on the alumina support.
Figure 2. TEM images of bundles of SWNTs synthesized at 680 °C (40 cm3/min methane) using supported Fe (a) and Fe/Mo (b) catalysts in reduced (metallic) form: (Fe:Al2O3 ) 1:16) (Fe:Mo: Al2O3 ) 1: 0.2:16). The catalysts were reduced in situ under flowing H2 at 500 °C.
as a function of CVD growth temperature; this time the spectra were collected using 488 nm radiation. The region between 350 and 1200 cm-1 is also not shown, as no peaks appeared in this region. Using the SWNT T-band intensity and band shape as our caliper, and consistent with the Raman spectra collected with 1064 nm excitation (Figure 3), we find that the tubes participating in the resonant scattering with 488 nm radiation confirm that the Fe/Mo-oxide catalyst is active down to 680 °C. In both Figures 3 and 4, the notch in the T-band at ∼1575 cm-1 is seem to be deepest at the CVD growth temperature of 800 °C. In some sense, the depth of this notch is a measure of sample quality, as it must deepen when the Raman line widths decreases (tube wall disorder 528
would increase these Raman line widths; amorphous carbon coating on the tube walls might also serve as a line broadening mechanism). Using this qualitative “T-band notch” criterion, the 800 °C CVD reaction appears to have produced the “best” tubes. The broad band at ∼1280 cm-1 (1064 excitation) and at ∼1350 cm-1 (488 nm excitation) seen in the spectra displayed in Figures 3 and 4 are well-known to stem from in various forms of disordered sp2 carbon.22-26 The band is “dispersive”, i.e., it’s position (cm-1) depends on excitation frequency. In fact the band shift by ∼70 cm-1 as the excitation is changed from 1064 to 488 nm. The cross section for scattering at this frequency requires disorder. It is usually referred to as the “disorder”-band or “D-band”. Here, the band can be associated with scattering from all disordered sp2 carbon in the sample (including the SWNTs). Consistent with our discussion above regarding disorder and the depth of the T-band “notch”, when the T-band notch is deep, we can also observe that the D-band is weak, and vice versa. It should be pointed out that the intensity of the carbon D-band is normally suppressed using 1064 nm excitation; the spectra collected at 488 nm are, as a consequence, more sensitive to this D-band contribution. We now briefly consider the radial mode scattering in Figure 4 (Fe/Mo-oxide catalyst/488 nm excitation/variable Nano Lett., Vol. 2, No. 5, 2002
Figure 4. Raman spectra for carbon materials produced at the CVD temperatures of 680, 800, and 900 °C for the optimized Fe/Mo catalyst (1:0.2:16) in reduced form. The data were collected using 488 nm excitation with the carbon still on the alumina support. Left panel: radial breathing mode region. The spectrum on the bottom is for the alumina support that exhibits two peaks that are marked with an asterisk (/). Right panel: SWNT Tangential (T) mode region. The broad band at 1350 cm-1 is the “D-Band” (see text).
Figure 5. Raman spectra for carbon materials produced at the CVD temperatures of 680 °C for the optimized Fe/Mo catalyst (1:0.2:x) in reduced form at various loadings (x). The data were collected using 488 nm excitation with the carbon still on the alumina support. Left panel: radial breathing mode region. Peaks identified with the alumina support are marked with an asterisk (*). Right panel: SWNT Tangential (T) mode region. The braod band at ∼1350 cm-1 is the “D-Band” (see text).
growth temperature; left-hand panel). It should be noted that two spectral features present in the radial mode region are associated with the catalysts/alumina (these features are indicated with an asterisk; the Raman spectrum for the unreacted catalysts/alumina also appears at the bottom of the left-hand panel of the figure). These alumina bands have significantly reduced intensity with 1064 nm excitation and are not seen in Figure 3. The radial breathing mode data (frequencies and tube diameters) are also collected in Table 1. With 488 nm excitation, there is some evidence that higher CVD growth temperatures produce smaller diameter tubes (new higher frequency radial mode bands), i.e., the appearance of peaks at ∼265 cm-1 (CVD T ) 800 °C), and at ∼310 cm-1 (CVD T ) 900 °C). However, a more detailed study would be necessary to strengthen the tentative conclusion that higher temperatures produce smaller diameter tubes. Nano Lett., Vol. 2, No. 5, 2002
Finally, we have studied the effect of the catalyst loading on the SWNT produced at the lowest temperature (680 °C) using reduced Fe/Mo catalysts (carbon mass yields appear in Table 1). From the mass yield data, we cannot conclude there is an optimal loading. Raman scattering was also used as a qualitative probe of sample quality. In Figure 5, we show Raman spectra collected with 488 nm excitation for five samples synthesized at the lowest CVD growth temperature on reduced Fe/Mo (1:0.2) catalyst; the catalyst loading onto the alumina support is shown in parentheses above each spectrum. The catalyst reduction was made in situ prior to the CVD growth experiment. Referring to the right-hand panel (T-bands), all five spectra exhibit strong T-bands with varying depth observed for the “T-band notch”. Again, using the T-band notch depth as a qualitative caliper of SWNT product quality, we find that the optimal loading for the 529
reduced catalyst is (1:0.2:11); this assessment is consistent with the observation that the weakest D-band scattering is also observed at this same loading. The radial breathing mode spectra for the samples produced at variable loading of reduced Fe/Mo catalyst are displayed in the left-hand panel of Figure 5. The radial breathing mode frequencies and the respective calculated tube diameters for these five catalyst loadings studied are also collected in Table 1. It is difficult to make any strong conclusions about the relationship between tube diameter and catalyst loading from these data. However, it is significant that the tube diameters observed via Raman scattering are relatively small, i.e., in the vicinity of 0.7 to 1.7 nm tubes. Conclusion. SWNTs were synthesized by CVD in methane at very “soft” conditions, i.e., low temperature (680 °C) and low methane flow rate (40 cm3/min). Under these soft conditions, even the reduced Fe catalyst was active. Also under these conditions (T ) 680 °C, 40 cm3/min) the addition of 20 wt % Mo (i.e., Fe:Mo ) 5:1) was sufficient to make the catalyst active in oxide form, i.e., no activation in Η2 was found necessary, eVen at 680 °C. As the reduction (or activation) of the growth catalyst in H2 represents an expensive, additional step in the CVD process, this is an important result from the present study. In general, we find a synergistic benefit from the addition of 20 wt % Mo to the Fe-catalyst. No evidence for the coproduction of multiwalled tubes was found. Acknowledgment. This work was supported, in part, by Honda Motors Fundamental Research Laboratories (A.R.H. and B.K.P) and by the NSF MRSEC (DMR-0080019) Program at PennState (G.C.). References (1) Iijima, S.; Ichihashi, T. Nature 1993, 363, 603. (2) Bethune, D. S.; Kiang, C. H.; de Vries, M. S.; Gorman, G.; Savoy, R.; Vasquez, J.; Beyers, R. Nature 1993, 363, 605. (3) Journet, C.; Maser, W. K.; Bernier, P.; Loiseau, A.; Delachapelle, M. L.; Lefrant, S.; Deniard, P.; Lee, R.; Fischer, J. E. Nature 1997, 388, 756. (4) Guo, T.; Nikolaev, P.; Thess, A.; Colbert, D. T.; Smalley, R. E. Chem. Phys. Lett. 1995, 243, 49.
530
(5) Maser, W. K.; Muoz, E.; Benito, A. M.; Martnez, M. T.; de la Fuente, G. F.; Maniette, Y.; Anglaret, E.; Sauvajol, J. L. Chem. Phys. Lett. 1998, 292, 587. (6) Rao, A. M.; Bandow, S.; Richter, E.; Eklund, P. C. Thin Solid Films 1998, 331, 141. (7) Dai, H.; Rinzler, A. G.; Nikolaev, P.; Thess, A.; Colbert, D. T.; Smalley, R. E. Chem. Phys. Lett. 1996, 260, 471. (8) Peigney, A.; Laurent, C.; Dobigeon, F.; Rousset, A. J. Mater. Res. 1997, 12, 613, (9) Hafner, J. F.; Bronikowski, M. J.; Azamiam, B. R.; Nikolaev, P.; Rinzler, A. G.; Colbert, D. T.; Smalley, R. E. Chem. Phys. Lett. 1998, 296, 195, (10) Su, M.; Zheng, B.; Liu, J. Chem. Phys. Lett. 2000, 322, 321. (11) Tanaka, K.; Yamabe, T.; Fukui, K. The Science and Technology of Carbon Nanotubes; Elsevier: Oxford, 1999. (12) Yakobson, B. I.; Smalley, R. E. Am. Sci. 1997, 85, 324. (13) Hafner, J. H.; Cheung, C.; Oosterkamp, T. H.; Lieber, C. M. J. Phys. Chem. B 2001 105, 743. (14) Wong, S.; Joselevich, E.; Woolley, A.; Cheung, C.; Lieber, C. Nature 1998, 394, 52. (15) Saito, Y.; Hamaguchi, K.; Hata, K.; Tohji, K. Ultramicroscopy 1998, 73, 1. (16) Fan, S.; Chapline, M.; Franklin, N.; Tombler, T.; Cassell, A.; Dai, H. Science 1999 283, 512. (17) Martel, R.; Schmidt, T.; Shea, H. R.; Hertel, T.; Avouris, P. Appl. Phys. Lett. 1998, 73, 2447. (18) Kitayanan, B.; Alvarez, W. E.; Harwell, J. H.; Resasco, D. E. Chem. Phys. Lett. 2000, 317, 497. (19) Cassell, A. M.; Raymakers, J. A.; Kong, J.; Dai, H. J. Phys. Chem. B 1999, 103, 6484. (20) Dresselhaus, M. S.; Dresselhaus, G.; Eklund, P. C. Science of Fullerenes and Carbon Nanotubes; Academic Press: New York, 1996. (21) Dresselhaus, M. S.; Eklund, P. C. AdVances in Physics 2000, 49, 705. (22) Rao, A. M.; Richter, E.; Bandow, S.; Chase, B.; Eklund, P. C.; Williams, K. A.; Fang, S.; Subbaswamy, K. R.; Menon, M.; Thess, A.; Smalley, R. E.; Dresselhaus, G.; Dresselhaus, M. S. Science 1997, 275, 187. (23) Dresselhaus, M. S.; Dresselhaus, G.; Pimenta, M.; Eklund, P. C. Analytical Applications of Raman Spectroscopy; Pelletier, M., Ed.; 1999, 367. (24) Matthews, M. J.; Pimenta, M. A.; Dresselhaus, G.; Dresselhaus, M. S.; Endo, M. Phys. ReV. B 1999, 59, R6585. (25) Pimenta, M. A.; Jorio, A.; Brown, S. D. M.; Filho, A. G. S.; Dresselhaus, G.; Hafner, J. H.; Lieber, C. M.; Saito, R.; Dresselhaus, M. S. Phys. ReV. B 2001, 64, R041401. (26) Dresselhaus, M. S.; Dresselhaus, G.; Jorio, A.; Filho, A. G. S.; Saito, R. Carbon 2002, submitted.
NL0255101
Nano Lett., Vol. 2, No. 5, 2002
