This content has been downloaded from IOPscience. Please scroll down to see the full text.
Download details:
IP Address: 140.113.38.11
This content was downloaded on 28/04/2014 at 05:09
Please note that terms and conditions apply.
Hot Filament for In Situ Catalyst Supply in the Chemical Vapor Deposition Growth of Carbon
Nanotubes
View the table of contents for this issue, or go to the journal homepage for more 2002 Jpn. J. Appl. Phys. 41 L67
(http://iopscience.iop.org/1347-4065/41/1A/L67)
Jpn. J. Appl. Phys. Vol. 41 (2002) pp.L67–L69 Part 2, No. 1A/B, 15 January 2002
c
2002 The Japan Society of Applied Physics
Hot Filament for In Situ Catalyst Supply in the Chemical Vapor Deposition Growth
of Carbon Nanotubes
Chia-Fu CHEN, Chien-Liang LINand Chi-Ming WANG
Department of Materials Science and Engineering, National Chiao Tung University, 1001 Ta Hsueh Road, Hsinchu 30049, Taiwan, Republic of China
(Received October 31, 2001; accepted for publication November 15, 2001)
A simplified chemical vapor deposition (CVD) method is described for in situ synthesis of multiwalled carbon nanotubes. The synthesis apparatus is similar to that used to deposit CVD diamond. However, an Fe–Cr wire is selected and coiled as the filament to grow nanotubes. The tubes grow because the filament acts as both a heat source for pyrolysis, and a source of metal for the catalyst. The evaporated metal atoms can be considered to catalyze the growth of carbon nanotubes. The system has the potential to inexpensively synthesize large amounts of nanotubes continuously by combining physical and chemical vapor deposition. [DOI: 10.1143/JJAP.41.L67]
KEYWORDS: carbon nanotubes, hot filament chemical vapor deposition, pyrolysis, catalyst
Since carbon nanotubes (CNTs) were first observed by Iijima1) in 1991, various methods have been developed for
their synthesis, including arc discharge,2) laser
vaporiza-tion,3)pyrolysis,4)and chemical vapor deposition (CVD).5–9)
Among these approaches, CVD methods hold some promise for scalability;8) however, they require carefully prepared
consumable substrate catalysts.6) Thus, a low-cost CVD
method which produces in situ catalysts for the growth of CNTs may be more valuable in practical applications. Several methods have been reported for growing CNTs in situ with-out pre-deposition of a catalyst layer. One such method pro-vides catalysts from the vapor source containing the metal ele-ment,10)while another uses the reactant gas which can act as a catalyst in certain substrates.11, 12)However, almost all meth-ods use chemical means to produce a catalyst in situ; few em-ploy physical vapor deposition (PVD) to produce such a cat-alyst. CNTs have been deposited as a by-product of diamond thick films in a hot filament CVD (HFCVD) system while copper was evaporated in situ from copper-covered parts near hot filaments to act as a catalyst during deposition.13)
How-ever, this method seems to be an impractical means of syn-thesizing CNTs. In this study, we report a simple, nontoxic, inexpensive in situ HFCVD method for preparing multiwalled nanotubes (MWNTs) on silicon. The method uses the fila-ment as the heat source for pyrolysis, and the catalytic evap-oration of the Fe–Cr filament.
A simplified CVD method for synthesizing nanotubes is described here. Figure 1 presents the schematic diagram of the direct HFCVD method. The proposed method eliminates nearly all of the complex and expensive machinery associ-ated with conventional nanotube growth. This system con-sists of an ac transformer, an Fe–Cr wire as the filament, a gas flow meter, a container of ethanol, a substrate holder, and a quartz tube used as a reaction vessel. The synthesis apparatus is similar to that used to deposit CVD diamond. However, tungsten is the most commonly used filament ma-terial in diamond deposition. An Fe–Cr wire is selected and coiled as the filament to grow nanotubes. The Fe–Cr wire used is commonly used in furnace heating. The wire is not high-grade and includes many impurities. Deposition is un-necessary in vacuo. Substrates were mirror-polished p-type, (100)-oriented wafers with a resistivity of 1–10·cm. Silicon wafers were first sonicated in acetone, and then washed using
Fe-Cr wire substrate substrate holder CO2 quartz tube ethanol ac transformer
Fig. 1. Schematic diagram of the direct HFCVD method.
DI water. The prototype system used a 1-mm-diameter Fe–Cr wire, and was heated by an ac current of 22 A at 40 V; the flow rate of the carrier gas CO2was 15 sccm, passing through
a container of ethanol. Growth was carried out for various du-rations, but it was found that growth for 15 min was sufficient to demonstrate the direct formation of CNTs on the surface of the silicon wafer.
A scanning electron microscope (SEM) (S-4000, Hitachi) was used to observe the original morphology and distribu-tion of the CNTs. A Renishaw micro-Raman spectroscope was used with an argon ion laser (514.5 nm line) to charac-terize the CNTs’ quality. Auger electron spectroscopy (AES) (Auger 670 PHI Xi, Physical Electronics) was used to identify the surface compositions of the CNTs. Finally, the nanotubes were scratched off and sonicated in acetone for 10 min before being dropped onto a carbon microgrid, and imaged with a transmission electron microscope (TEM) (tecnai 20, Philips) to further characterize them.
SEM images in Fig. 2 show the surface morphology of the CNTs. A random tube network including curved tubes was formed. Tubes were about 60–80 nm in diameter and a few tens of microns long. The location of CNTs on the sample was investigated using the SEM images. The black images reveal that most of the CNTs were under the lowest position of the coiled filament. TEM analysis of the CNTs was per-formed to confirm that these structures are truly CNTs, and
L68 Jpn. J. Appl. Phys. Vol. 41 (2002) Pt. 2, No. 1A/B C.-F. CHENet al.
Fig. 2. SEM images of CNTs on silicon, obtained with 15 sccm CO2 car-rier gas. Upper image: scale bar 9µm. Lower image: scale bar 600 nm.
not carbon fibers. Figure 3 shows a representative TEM mi-crograph of the nanotubes. A comparison of our result with images presented elsewhere11, 13) shows that these tubes are multiwalled CNTs. The darker nanotube walls indicate that the nanotubes are multiwalled and hollow rather than solid fibers. The bending and twisting defects of the CNTs shown in Fig. 3 are consistent with the SEM images in Fig. 2. Figure 3 also shows that the CNTs have inner diameters of around 20 nm.
Figure 4 displays the Raman spectrum of CNTs grown by direct HFCVD. Below 3500 cm−1, the spectrum shows four strong peaks at 1350 (D line), 1582 (G line), 2691, and 2940 cm−1, a weak peak at 3230 cm−1, and a weak peak at
∼ 1615 cm−1(Dline) resulting from the line at 3230 cm−1.14)
While the peak position and shape depended slightly on the location measured, Fig. 4 presents the main features of the spectra, which agree with those observed by other groups.9, 14) A comparable intensity of the G and D lines implies a high density of structural defects15)in the curved graphene sheets. The SEM and TEM images of the CNTs in this study reveal that the CNTs contain carbon nanoparticles. The appearance of the strong D line can be interpreted as representing a large number of crystalline domains on the nanometer scale, en-hanced by the surfaces of the tubes.14)
Fig. 3. Transmission electron microscope image of the multiwalled CNTs, produced by direct HFCVD. 1000 1500 2000 2500 3000 3500 200 400 600 800 1000 1350 1582 1615 2691 2940 3230
Intensity (arb. units)
Raman Shift (cm-1)
Fig. 4. Micro-Raman spectrum of the CNTs grown on the silicon wafer.
The reason for the growth of the CNTs is discussed here. Almost every previous report has indicated that the growth of the CNTs must be attributable to the catalytic effect, and the use of hydrocarbon as the carbon source. In this synthesis, ethanol was chosen as the carbon source for growing CNTs. In addition to CO2, Ar gas was also used as a carrier gas in
growing CNTs. However, Ar yields fewer CNTs than CO2
does. Therefore, CO2 may be a key component in enhancing
the growth of CNTs, but the reason is not yet known. The fact that Ar can also be used suggests that no catalyst is present in the gas or vapor environment employed in this study;
there-Jpn. J. Appl. Phys. Vol. 41 (2002) Pt. 2, No. 1A/B C.-F. CHENet al. L69
fore, the filament is the only possible catalyst source. The fil-ament consists of Fe (72.43%), Cr (23.42%), Mn (3.45%) and Ni (0.69%), which was characterized using energy dispersive spectrometry (EDS) attached to the SEM. In the present ex-periment, the temperature of the filament is about 1200◦C, which is lower than its melting point by 100–200◦C. Under such conditions, the situation is similar to that of conven-tional vacuum evaporation coating, but with fewer evaporated atoms due to the lower temperature and absence of vacuum conditions. As shown in Fig. 1, the mixture of carrier gas and ethanol flows toward the filament. In addition to this flow di-rection, other flow directions were also tested, but few and scattered CNTs formed on the substrate. Therefore, we con-sider that the evaporated metal atoms following the flow di-rection toward the silicon substrate will provide the catalyst for further growth of the CNTs.
CNTs have been reported13)to be deposited as a by-product
of diamond thick films in an HFCVD system. That study suggested that the copper-covered parts near the hot fila-ments were evaporated in situ during deposition, mainly in the form of atomic clusters that might have enhanced catal-ysis due to the size effect. The surface stoichiometry of our nanotubes, characterized using AES, shows small amounts of Si (3.75%), Al (4.89%), O (3.33%), Ni (4.99%), Fe (6.37%) and Cr (4.78%) in the carbon (71.89%). Si may arise due to the substrate effect, and Al is the impurity associated with deposition at high temperature. Ni, Fe and Cr have been con-sidered to act as catalysts in the synthesis of CNTs.12, 16–18)Fe
is the major element, at 6.37 at%, suggesting it plays a lead-ing role in catalyzlead-ing the formation of these CNTs. In partic-ular, Fe is considered the best catalyst for thermal CVD.19)
EDS attached to the TEM also shows that Fe atoms appear; however, we did not find Fe particles so we have no direct proof to explain the mechanism of the CNT growth. On the other hand, we did not have sufficient experience using W wire as the filament for growing CNTs. The major reason is that W wire cannot stand a long time without fracture un-der the condition of using a mixture of CO2 and ethanol.
In-deed, we have confirmed that when using a mixture of H2
and ethanol to deposit carbon, no CNTs were formed. This may be attributed to tungsten having a much higher melting point (2974◦C). Therefore, we believe that in situ evaporated catalysts from the filament should be used to grow CNTs. Ac-cordingly, this method combines PVD and CVD to synthesize CNTs. In addition, since in situ evaporated catalysts from the filament were used to grow the CNTs, an alternating filament grid arrangement may support the scaling of the method, and continuous production may be attainable by in situ growth of CNTs which can be peeled off during the process.
A simplified CVD method is reported for in situ synthesis
of multiwalled carbon nanotubes. The most important feature of this direct HFCVD synthesis method is that no obvious cost or technological obstacles arise in scaling up the method for continuous high-quantity production. This simplified design possesses many advantages. The carrier gas is nonflammable, ethanol is nontoxic, and all complex vacuum seals are elimi-nated. CO2or Ar gas can be used as the carrier gas in growing
the CNTs, and CO2yields more CNTs than Ar does. The
di-rection of the gas mixture flow, namely, vertically toward the substrate, will benefit the CNT growth because the evaporated metal atoms can more easily contact the substrate surface. The direct HFCVD system has the potential to inexpensively syn-thesize large quantities of CNTs continuously.
The authors would like to thank the National Science Council of the Republic of China, Taiwan for financially supporting this work under Contract No. NSC 89-2213-E-009-228. We are grateful to Mr. Der-Fu Chang for carrying out the TEM measurements.
1) S. Iijima: Nature 354 (1991) 56.
2) C. Journet, W. K. Maser, P. Bernier, A. Loiseau, M. Lamy de la Chapelle, S. Lefrant, P. Deniard, R. Lee and J. E. Fischer: Nature
388 (1997) 756.
3) T. Guo, P. Nikolaev, A. Thess, D. T. Colbert and R. E. Smalley: Chem. Phys. Lett. 243 (1995) 49.
4) M. Terrones, N. Grobert, J. Olivares, J. P. Zhang, H. Terrones, K. Kordatos, W. K. Hsu, J. P. Hare, P. D. Townsend, K. Prassides, A. K. Cheetham, H. W. Kroto and D. R. M. Walton: Nature 388 (1997) 52. 5) M. Jose-Yacaman, M. Miki-Yoshida, L. Rendon and J. G. Santiesteban:
Appl. Phys. Lett. 62 (1993) 202.
6) S. Fan, M. G. Chapline, N. R. Franklin, T. W. Tombler, A. M. Cassell and H. Dai: Science 283 (1999) 512.
7) C. J. Lee, J. Park, J. M. Kim, Y. Huh, J. Y. Lee and K. S. No: Chem. Phys. Lett. 327 (2000) 277.
8) M. Su, B. Zheng and J. Liu: Chem. Phys. Lett. 322 (2000) 321. 9) Q. Zhang, S. F. Yoon, J. Ahn, B. Gan, Rusli and M. B. Yu: J. Phys.
Chem. Solids 61 (2000) 1179.
10) A. M. Rao, D. Jacques, R. C. Haddon, W. Zhu, C. Bower and S. Jin: Appl. Phys. Lett. 76 (2000) 3813.
11) Z. P. Huang, J. W. Xu, Z. F. Ren, J. H. Wang, M. P. Siegal and P. N. Provencio: Appl. Phys. Lett. 73 (1998) 3845.
12) M. Okai, T. Muneyyochi, T. Yanguchi and S. Sasaki: Appl. Phys. Lett.
77 (2000) 3468.
13) B. Gan, J. Ahn, Q. Zhang, S. F. Yoon, Rusli, Q. F. Huang, H. Yang, M. B. Yu and W. Z. Li: Diamond & Relat. Mater. 9 (2000) 897. 14) W. Li, H. Zhang, C. Wang, Y. Zhang, L. Xu, K. Zhu and S. Xie: Appl.
Phys. Lett. 70 (1997) 2684.
15) W. S. Bacsa, D. Ugarte, A. Chˆatelain and W. A. de Heer: Phys. Rev. B
50 (1994) 15 473.
16) Z. W. Pan, S. S. Xie, B. H. Chang, L. F. Sun, W. Y. Zhou and G. Wang: Chem. Phys. Lett. 299 (1999) 97.
17) L. C. Qin, D. Zhou, A. R. Krauss and D. M. Gruen: Appl. Phys. Lett.
72 (1998) 3437.
18) J. I. Sohn, C. J. Choi, S. Lee and T. Y. Seong: Appl. Phys. Lett. 78 (2001) 3130.
