Fullerenes and Carbon Nanotubes Formed in an Electric Arc at and ...

5 downloads 0 Views 437KB Size Report
[60]Fullerene and carbon nanotubes have been obtained with an electric arc discharge in helium or argon at the pressure of 0.1–3.0 bar (i.e. from. J0.9 up to ...
FULLERENES, NANOTUBES, AND CARBON NANOSTRUCTURES Vol. 12, No. 3, pp. 593–602, 2004

Fullerenes and Carbon Nanotubes Formed in an Electric Arc at and Above Atmospheric Pressure J. J. Langer,* S. Golczak, S. Z˙abin´ski, and T. Gibin´ski Laboratory for Materials Physicochemistry and Nanotechnology, Faculty of Chemistry, A. Mickiewicz University at Poznan´, S´rem, Poland

ABSTRACT [60]Fullerene and carbon nanotubes have been obtained with an electric arc discharge in helium or argon at the pressure of 0.1 –3.0 bar (i.e. from 20.9 up to 2.0 bar above the atmospheric pressure, respectively). The yield of the process is about 10 times higher in helium with a maximum at the pressure of 0.2 bar, while in argon the maximum is observed at 0.4 bar. It is not very important, what kind of the inert gas (He or Ar) is used in a process carried out at the atmospheric pressure and higher. In this case, the yield is usually better or comparable with other alternative methods, e.g. the solar furnace synthesis. We found that the oxygen plays a crucial role reducing the yield of fullerenes and carbon nanotubes. Even traces of O2 must be removed from the reactor chamber, otherwise yellow

*Correspondence: J. J. Langer, Laboratory for Materials Physicochemistry and Nanotechnology, Faculty of Chemistry, A. Mickiewicz University at Poznan´, Grunwaldzka 10, PL-63100, S´rem, Poland; E-mail: [email protected]. 593 DOI: 10.1081/FST-200026944 Copyright # 2004 by Marcel Dekker, Inc.

1536-383X (Print); 1536-4046 (Online) www.dekker.com

594

Langer et al. coloured fullerene derivatives (oxidation products) are formed instead of C60. Key Words:

Fullerenes; Carbon nanotubes; Electric arc; Synthesis.

INTRODUCTION Fullerenes and carbon nanotubes are still of growing interest due to their potential applications.[1] Search for new synthetic and purification techniques are among the most important goals for many laboratories. An improved method of preparation of pure double-walled carbon nanotubes, based on the floating chemical vapour deposition (CVD) process, has been reported recently.[2] It is generally accepted that the fullerenes and carbon nanotubes are formed in special conditions: at a very high temperature (a laser beam or an electric arc discharge evaporation of graphite) and in an inert gas atmosphere at a reduced pressure (e.g. He, about 0.3 bar).[3 – 6] On the other hand, fullerenes, fullerene derivatives, and fullerene-like structures have been detected in the soot produced for technical applications and also in the diesel motor soot.[7] The electric arc discharge is still one of the best methods to produce fullerenes, which is why we decided to check how it works in strongly modified conditions: near and above atmospheric pressure, and also what is the role of an inert gas in such a case. This work presents the results of our experiments.

EXPERIMENTAL The fullerenes and carbon nanotubes were prepared by evaporation of carbon electrodes in an electric arc discharge process. We have used graphite electrodes of a diameter of 5 mm and the length of 150 mm or anthracite electrodes of a diameter of 6 mm and the length of 150 mm. The electrodes were horizontally mounted near the bottom of a vertically oriented reaction chamber. DC and AC power supply units were used and operated at the voltage of 30– 50 V and the current of 50 – 200 A. The chamber was evacuated of the air and filled in with a noble gas (argon or helium) at the pressure from 0.1 to 3.0 bar. To modify and optimise the process we have changed the inert gas (Ar to He), the pressure (from 0.1 to 3.0 bar), and the current (50 –200 A, most frequently 90 A) using both DC and AC power supply units operated at the voltage of 30– 50 V. To protect the product against traces of oxygen, a copper grid was placed near (below and above) the carbon plasma flame. The soot condensed in a

Formation of Fullerenes and Carbon Nanotubes

595

chamber was additionally shielded to reduce degradation caused by intense UV irradiation. Fullerene C60 and C70 were extracted from the soot with toluene, separated and purified on a chromatography column (SiO2 – Norit A 1 : 1.5, benzene or benzene –hexane 9 : 1). The products, dried in an argon atmosphere at the temperature of 1008C and the reduced pressure of 0.01 Pa, were characterised with the use of TLC (SiO2 0.25 mm Merck, benzene) and spectrometry methods. C60 was crystallised from toluene (Fig. 1). Carbon nanotubes were prepared at the same conditions as fullerenes and isolated mostly from the deposit, but they have also been found in the soot condensed in a cooler, upper section of the reaction chamber. The yield, evaluated semi-quantitatively as a surface density of nanotubes observed in a deposit (SEM/TEM micrographs), was generally higher at a lower temperature of the electric arc (a lower current, e.g. 75 –110 A) and a higher pressure (e.g. 2 bar) applied. This observation well correlates with the amount of nanotubes obtained. Carbon nanotubes were characterised with EPR (Radiopan, operated at X band) and the electron microscopy: SEM (Philips SEM 515) and TEM (Jeol JEM 1200 II EL).

RESULTS AND DISCUSSION We used the electric arc discharge in noble gases (Ar or He) to evaporate carbon (graphite or anthracite) electrodes at the atmospheric pressure and even

Figure 1. C60 crystal (0.5 mm in size) formed from toluene solution.

596

Langer et al.

higher (up to 3.0 bar). We found the resulting soot contained not only fullerenes and fullerene derivatives (oxidation products), but also—to some degree—carbon nanotubes that normally appear mostly in a deposit (this is the result of intensive convection in a dense atmosphere). The yield of the process is about 10 times higher in helium with a maximum at the pressure of 0.2 bar, while in argon the maximum is lower and appeared at 0.4 bar (Fig. 2). Surprisingly, we have observed fullerenes, fullerene derivatives, and carbon nanotubes in almost all experiments performed at an elevated pressure (1 bar and above), with the yield of pure [60]fullerene about 0.5% (in the soot) which is not very high but still comparable with other alternative methods, e.g., the benzene–oxygen flame[8] or solar furnace[9,10] synthesis. In this case, the results are almost independent on what kind of the inert gas is used (He or Ar). Fullerene C60 was purified and characterised with NMR, MS, UV-VIS, and FTIR spectra. We found for C60: 13C-NMR: 143.4 ppm (benzene-d6); MS (LSIMS, NBA): 721.361 m/z [M þ 1]; UV-VIS (hexane): 404, 328, 268, 260, 257, 220 nm; FTIR [KBr, wave number (cm21)/relative intensity]: 1428/0.21, 1181/0.15, 575/0.23, 526/0.40; XRD (powder, 2Q): 5.4 (111), 8.8 (220), 10.3 (311); and TLC Rf ¼ 0.9. The amount of C70 obtained was not sufficient for full spectral characterisation. The yield of pure [60]fullerene is limited mainly by the reaction with oxygen present in the reaction vessel, and not only an elevated pressure, as reported previously for the low-pressure process.[3 – 6] Even traces of oxygen must be removed from the reactor chamber (about 0.1 ml of O2 is enough to oxidise all fullerene produced in our experiment), otherwise yellow coloured fullerene derivatives (oxidation products, TLC Rf ¼ 0.3, FTIR: strong absorption at 1094 and 1020 cm21) are formed instead of fullerenes. Changing the inert gas to Ar results in a decrease in the yield of fullerenes and a lowering of the difference between the yield found for graphite and anthracite (Fig. 2). At a low pressure of He, graphite electrodes have appeared to be much better than anthracite for the preparation of fullerenes, but not carbon nanotubes (in this case anthracite gave us better results, taking into account the concentration and the morphology of nanotubes formed). This is caused by contamination of anthracite leading to formation of highly polar fullerene derivatives (soluble in acetonitrile) that have not yet been fully identified (these compounds are not observed by TLC in extracts prepared from clean anthracite electrodes). We found that generally the processes leading to fullerenes and carbon nanotubes are competitive: the first one is more efficient at the higher current (higher temperature) and the lower pressure, while the second (formation of carbon nanotubes) needs a lower current (lower temperature) and a higher pressure. To simplify the system examined, despite interesting results with anthracite, further experiments on formation of carbon nanotubes at an elevated pressure have been performed with pure graphite

Formation of Fullerenes and Carbon Nanotubes

597

Figure 2. The yield of pure C60 formed with graphite and anthracite electrodes (alternating current of 130 A) as a function of the pressure of an inert gas (He and Ar).

electrodes. In argon, the best results we found at the pressure of 2 bar and the DC current of 75 –110 A (Fig. 3), while no nanotubes have been observed at 55 A which is the lowest current applied (Fig. 4). Carbon nanotubes (single- and multi-wall structures) have been found not only as the electrode deposit, but also in the soot formed owing to aggregation of carbon particles existing in the plasma.

598

Langer et al.

Figure 3. Carbon nanotubes produced at and above atmospheric pressure (TEM micrographs, a scale bar corresponds to 200 nm): (a) isolated nanotubes, (b) a mixture.

The most striking are carbon microrods [Fig. 5(a)] randomly oriented and totally composed of very pure nanotubes [Fig. 5(b)] that have been formed at 2 bar of Ar and the current of 75 A. Such microstructures, formed at 2 bar of Ar and 110 A, are organised in layers [Fig. 6(a)] and they consist of almost pure carbon nanotubes (a maximum yield), also randomly oriented inside microrod

Figure 4. Carbon nanostrucutres formed at 2 bar of Ar and 55 A, with no nanotubes (SEM).

Formation of Fullerenes and Carbon Nanotubes

599

Figure 5. (a) Unique carbon microrods (SEM micrograph) composed of (b) carbon nanotubes (TEM micrograph). A scale bar corresponds to 100 nm.

600

Langer et al.

Figure 6. (a) Microlayers (SEM micrograph) composed of (b) carbon nanotubes (SEM micrograph), and (c) their EPR spectrum.

Formation of Fullerenes and Carbon Nanotubes

601

structures [Fig. 6(b)]. In this case, an anisotropy of the EPR signal has appeared [Fig. 6(c)], which is typical for solids (powders) with paramagnetic centres (unpaired electrons) of weak spin exchange interactions and an anisotropic hyperfine coupling and g factor. This can be assigned to semiconducting carbon nanotubes. Similar in the shape, but without any nanotubes, carbon “microtrees”, have been observed in a deposit obtained as the result of a “flash” chemical vapour deposition (CVD) process in a methane– helium atmosphere at a reduced pressure.[11] There is probably a general mechanism leading to the formation of carbon microstructures: microrods and “microtrees” (in both cases, no catalyst was used), but it is still not clear.

CONCLUSIONS We found, that fullerenes and carbon nanotubes are formed at normal and above atmospheric pressure, with the yield comparable with other alternative methods. This can help to simplify the synthetic procedure making it even possible to work in a continuous regime (at the ambient pressure). We have found a competitive nature of reactions leading to fullerenes and carbon nanotubes in the electric arc discharge process. Formation of carbon microrods of a unique complex structure (composed of nanotubes) has been observed.

REFERENCES 1. Baughman, R.H.; Zakhidov, A.A.; de Heer, W.A. Carbon nanotubes—the route toward applications. Science 2002, 297, 787 – 792. 2. Wei, J.; Ci, L.; Jiang, B.; Li, Y.; Zhang, X.; Zhu, H.; Xu, C.; Wu, D. Preparation of highly pure double-walled carbon nanotubes. J. Mater. Chem. 2003, 13 (6), 1340– 1344. 3. Khemani, K.C.; Prato, M.; Wudl, E. A simple soxhlet chromatographic method for the isolation of pure C6O þ C70. J. Org. Chem. 1992, 57, 3254. 4. Koch, A.; Khemani, A.C.; Wudl, F. Preparation of fullerenes with a simple benchtop reactor. J. Org. Chem. 1991, 56, 4543. 5. Lamb, L.D.; Huffman, R.D. Fullerene production. J. Phys. Chem. Solid 1993, 54, 1635. 6. Aije, H.; Alvarez, M.M.; Anz, J.S.; Beck, R.D.; Diederich, F.; Fostinopulos, K.; Whetten, R.L. Characterization of the soluble allcarbon molecules C60 and C70. J. Phys. Chem. 1990, 94, 8630.

602

Langer et al.

7. Langer, J.J.; Gibin´ski, T.; Piatek, M. Fullerenes in the human environment, in preparation. 8. Pope, C.J.; Harr, J.A.; Howard, J.B. Chemistry of fullerenes C60 and C70 formation in flames. J. Phys. Chem. 1993, 97, 11,001. 9. Chibante, L.P.F.; Thess, A.; Alford, J.M.; Diener, M.D.; Smalley, R.E. Solar generation of the fullerenes. J. Phys. Chem. 1993, 97, 8696. 10. Pitts, R.R.; Hale, J.M.; King, D.E.; Fields, C.E. Formation of fullerenes in highly concentrated solar flux. J. Phys. Chem. 1993, 97, 8701. 11. Ajayan, P.M.; Nugent, J.M.; Siegel, R.W.; Wei, B.; Kohler-Redlich, Ph. Growth of carbon microtrees. Nature 2000, 404, 243. Received August 4, 2003 Accepted November 10, 2003