Behavior of C60 under hydrothermal conditions: Transformation to amorphous carbon and formation of carbon nanotubes

Article abstract of DOI:10.1006/jssc.2001.9220

The behavior of fullerenes C₆₀ under hydrothermal conditions between 200 and 800°C, and under 100 MPa pressure, in the absence and in the presence of nickel is reported. The highest temperature of fullerene stability in water was 400°C after a 48-h-long treatment. Increasing the temperature and/or time of the hydrothermal treatment resulted in transformation of fullerene to amorphous carbon. At 700°C, high-quality open-ended multiwalled carbon nanotubes were formed in the vicinity of nickel particles. These nanotubes typically had an outer diameter of 30-40 nm and a wall thickness of 5 nm, with a graphitization level similar to that of carbon nanotubes prepared by chemical vapor deposition. The potential for large-scale synthesis of the carbon nanotubes by the hydrothermal technique is discussed. Since the present study was conducted under conditions which are common in the geological environment, our results imply that carbon nanotubes may form in natural hydrothermal systems.

Full text of DOI:10.1006/jssc.2001.9220

Journal of Solid State Chemistry 160, 184}188 (2001)
doi:10.1006/jssc.2001.9220, available online at http://www.idealibrary.com on
Behavior of C60 under Hydrothermal Conditions: Transformation
to Amorphous Carbon and Formation of Carbon Nanotubes
Wojciech L. Suchanek,*ꢃꢀ Joseph A. Libera,-ꢃꢁ Yury Gogotsi,-ꢃꢂ and Masahiro Yoshimura*
*Materials and Structures Laboratory, Center for Materials Design, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8503 Japan; and
-Department of Mechanical Engineering, University of Illinois at Chicago, M/C 251, Chicago, Illinois 60607-7022
Received December 21, 2000; in revised form April 11, 2001; accepted April 20, 2001; published online June 11, 2001
focused sunlight, direct inductive heating of carbon, or soot-
The behavior of fullerenes C60 under hydrothermal conditions
between 200 and 8003C, and under 100 MPa pressure, in the
absence and in the presence of nickel is reported. The highest
temperature of fullerene stability in water was 4003C after
a 48-h-long treatment. Increasing the temperature and/or time of
the hydrothermal treatment resulted in transformation of
fullerene to amorphous carbon. At 7003C, high-quality open-
ended multiwalled carbon nanotubes were formed in the vicinity
of nickel particles. These nanotubes typically had an outer dia-
meter of 30+40 nm and a wall thickness of 5 nm, with a graph-
ing hydrocarbon #ames (4). The major synthesis routes for
carbon nanotubes are the carbon arc method (5), laser
ablation (6), and chemical vapor deposition (CVD) (7}14).
All these techniques require high temperatures (usually at
least 700}8003C for the nanotubes and 10003C for the
fullerenes), vacuum systems, gas #ow, and complicated
equipment. Yields are relatively low and the products often
require puri"cation.
With the current synthesis of fullerenes and carbon
itization level similar to that of carbon nanotubes prepared by nanotubes being still very expensive and in the absence of
chemical vapor deposition. The potential for large-scale synthesis
of the carbon nanotubes by the hydrothermal technique is dis-
cussed. Since the present study was conducted under conditions
which are common in the geological environment, our results
imply that carbon nanotubes may form in natural hydrothermal
systems. ( 2001 Academic Press
Key Words: hydrothermal synthesis; carbon nanotubes;
fullerene; C60; amorphous carbon; CVD; hydrothermal systems;
geothermometer; hydrothermal mineral deposits.
known large-scale natural deposits of these carbon allot-
ropes, their low-temperature, high-yield synthesis would be
of great scienti"c and technological importance. Carbon
nanotubes have been synthesized by electrolysis in molten
salts (15). It has also been shown that amorphous carbon
(16) and even diamond (17) can be prepared under hy-
drothermal conditions. Hydrothermal synthesis of carbon
"laments, including well-aligned arrays on a substrate (18),
demonstrated the potential of the hydrothermal technique
for growing complex carbon structures. However, these
"laments contained periodic conical cavities and were
thicker than typical carbon nanotubes (18). Taking into
account the advantages of hydrothermal synthesis in mater-
ials processing (19, 20) and the fact that various forms of
hydrothermal carbon can be prepared in the temperature
range of 300}8003C (100}140 MPa) (16, 17), this method
seems to be potentially very promising for synthesis of the
new carbon allotropes.
1. INTRODUCTION
Fullerenes (1) and fullerene-related materials, such as
carbon nanotubes (2), have many potential applications and
thus have attracted great interest of researchers all over the
world. Fullerenes can be synthesized under extreme condi-
tions, which rarely occur in nature, for example the electric
arc method (3), carbon vaporization by pulsed lasers or in
It is also worth mentioning that interactions between
fullerenes/nanotubes and water have recently attracted at-
tention due to e!orts to increase the e$ciency of extrac-
tion/puri"cation of these carbon allotropes. Therefore,
ꢀTo whom correspondence should be addressed. Present address: De-
partment of Ceramic and Materials Engineering, Rutgers University, 607 a study of the hydrothermal behavior of the fullerenes and
Taylor Rd., Piscataway, NJ 08854. Fax: (732) 445-6264. E-mail:
suchanek@rci.rutgers.edu.
ꢁPresent address: Department of Materials Science and Engineering,
Northwestern University, Evanston, IL 60208.
the carbon nanotubes could give unique information about
their formation/stability/transformation conditions, which
is of great importance for mastering puri"cation and syn-
thesis techniques of these materials. Moreover, such a study
could be very useful for geologists looking for natural
ꢂPresent address: Department of Materials Engineering, Drexel Univer-
sity, Philadelphia, PA 19104.
184
0022-4596/01 $35.00
Copyright ( 2001 by Academic Press
All rights of reproduction in any form reserved.

HYDROTHERMAL FORMATION OF CARBON NANOTUBES FROM C
185
ꢄꢅ
deposits of fullerenes and carbon nanotubes, because nat-
urally occurring hydrothermal systems are widespread in
the Earth's crust.
2. EXPERIMENTAL
Fullerene powder (C , purity 99.95%, Science Laborat-
ꢄꢅ
ories Co., Japan) was used in all experiments. Small samples
of this powder (+0.020 g) were inserted into golden cap-
sules of 3 mm diameter (volume of +0.1}0.2 cmꢂ), which
were subsequently "lled with double-distilled water. The
capsules were then sealed, placed into autoclave tubes
(Tuttle-Roy type), and heated at 200, 400, 500, 600, 650, 700,
750, and 8003C for periods between 20 min and 48 h under
100 MPa pressure. Water was used as the pressure transfer
medium. The nickel powder (purity 99.8%, Nilaco Co.,
Japan) was added to some capsules in the amount of 3 wt%.
The Ni-containing capsules, where amounts of water were
between 30%}100% (by weight), were treated at 400, 500,
600, and 7003C for 168 h under 100 MPa pressure. After the
synthesis, the autoclave tubes were rapidly quenched in
water. All the hydrothermal experiments were conducted at
the Tokyo Institute of Technology. Materials were charac-
FIG. 1. Selected Raman spectra of the C powder after hydrothermal
ꢄꢅ
terized at the Tokyo Institute of Technology by X-ray treatments in water with or without Ni, under 100 MPa pressure. The
di!raction (XRD, 40 kV/40 mA, CuKa radiation, MAC
Science Co. Ltd., Tokyo, Japan) and by Raman spectro-
scopy. A Raman spectrometer (T64000, Atago-Jobin Yvon,
temperature and duration of the treatments are marked. Arrows indicate
weak C bands. Amorphous carbon bands at 1608 cm\ꢀ and 1335 cm\ꢀ,
ꢄꢅ
as well as a shoulder band at 1191 cm\ꢀ, are marked. All other bands
originate from C
.
ꢄꢅ
France-Japan) with an Ar> laser with an excitation
wavelength of 514.5 nm has been used. The laser power was
low to avoid transformation of the fullerenes to amorphous
carbon during the measurements. The materials were char-
acterized also by "eld emission scanning electron micros-
copy (FESEM) and transmission electron microscopy
(TEM) at the Research Resources Center of the University
of Illinois at Chicago. The FESEM used was a JSM-6320F
with magni"cation up to 500,000;. High resolutions at low
accelerating voltages are possible with this instrument due
to its objective lens design. The microscope is also "tted
with a Noran Voyager EDX system with a light element
X-ray detector analyzer, which was used for elemental anal-
ysis of nanotubes. The TEM used was a JEOL 3010 (300
kV) with a lattice resolution of 0.14 nm (point resolution
0.17 nm).
C
, indicating that the C was stable under hydrothermal
ꢄꢅ
ꢄꢅ
conditions in this temperature range for these times, as
shown in Fig. 1 (in this context the word &&stable'' is not used
in a thermodynamically rigorous sense). In the Raman
spectra acquired from the fullerenes hydrothermally treated
at 5003C (168 h), or at higher temperatures but for shorter
times, e.g., at 6003C (18 h) or 7003C (0.3 h), new bands
appeared in addition to weakening of C -derived bands.
ꢄꢅ
Two strong and broad bands at about 1335 cm\ꢀ and
1608 cm\ꢀ can be ascribed to amorphous carbon (16).
These positions of bands of amorphous carbon (down-
shifted D-band and up-shifted G-band) are typical for hy-
drothermally formed carbon. A shoulder band at about
1200 cm\ꢀ has been observed in hydrothermally formed
carbon but its origin is not clear. After hydrothermal treat-
ment at 6003C (48 h), 7003C (168 h), or 8003C (0.3 h), the
Raman spectra did not show any fullerenes, demonstrating
3. RESULTS AND DISCUSSION
Selected Raman spectra of the hydrothermally treated
fullerene powders are shown in Fig. 1. The Raman spectrum
denoted &&as received'' corresponds to the fullerene powder,
which did not undergo any hydrothermal treatment. This
that C was completely transformed to amorphous carbon
ꢄꢅ
(Fig. 1). XRD patterns of the hydrothermally treated
fullerene powders supported the Raman spectra. New broad
peaks indicating transformation of the fullerenes were
observed at 24.53 and 43.53 and could be ascribed to graph-
itic carbon. Both Raman spectra and XRD patterns of
spectrum is characteristic of C , showing clearly all 10
ꢄꢅ
Raman-active modes (21). Raman spectra of the C ac-
ꢄꢅ
quired after the hydrothermal treatments in pure water at
the C hydrothermally treated in water under di!erent
200}4003C (0.3}48 h) were almost identical to as-received
ꢄꢅ

186
SUCHANEK ET AL.
C
is stable up to over 40003C (22). In an inert gas atmo-
conditions show that, with increasing temperature and time
ꢄꢅ
sphere C crystals transform into amorphous carbon at
700}9503C (23). Our results indicate that water strongly
accelerates transformation of fullerenes. The process of hy-
drothermal transformation of C to amorphous carbon is
kinetically controlled; thus, longer hydrothermal treatments
will shift the stability range to lower temperatures.
FESEM analysis of the fullerene crystals after hydrother-
mal treatment in the presence of nickel revealed formation
of carbon "laments in the vicinity of the Ni particles
(Fig. 3a). The "laments had a circular cross-section and
their outer diameter was in the range of 30}120 nm. The
wall thickness of the carbon "laments was in the range
of 4}40 nm, which corresponds to a number of graphite
layers of 10}100. Some of the "laments appeared to be
high-quality open-ended multiwalled carbon nanotubes,
of the hydrothermal treatment, the C fullerenes gradually
ꢄꢅ
ꢄꢅ
transform to amorphous carbon. The highest temperature
of their stability was 4003C after 48-h-long treatment under
100 MPa.
ꢄꢅ
The C crystals were transformed during the hydrother-
ꢄꢅ
mal treatment into amorphous carbon, some of which was
graphitized (Fig. 2b), without any signi"cant shape changes
(Fig. 2a), but sometimes with visible etch textures. The
fullerenes, which did not survive any hydrothermal treat-
ment at temperatures higher than 4003C, are known to
exhibit quite high temperature stability without H O. Mo-
ꢁ
lecular dynamic studies show that in the carbon-only system
FIG. 2. (a) Typical FESEM micrograph of amorphous carbon
formed during hydrothermal treatment (3% Ni, 30% water, 7003C,
100 MPa, 168 h), showing that the shape of the transformate is the same
FIG. 3. (a) FESEM photograph showing the carbon "laments in the
as that of the original C crystals. Some cracks and etch textures vicinity of a Ni particle. (b) TEM micrographs of typical carbon nanotubes
ꢄꢅ
were observed. (b) TEM photograph of the amorphous and graphitized produced by hydrothermal treatment of C with 3% Ni, 30% water at
ꢄꢅ
carbon formed in our experiments, revealing the presence of graphite 7003C for 168 h under 100 MPa pressure. The inset shows graphite fringes
layers.
in the nanotube wall.

HYDROTHERMAL FORMATION OF CARBON NANOTUBES FROM C
187
ꢄꢅ
typically with an outer diameter of 30}40 nm and a wall viscosity is lower, enhancing the mass transport (19). These
thickness of 5 nm (Fig. 3b). The high-resolution TEM shows phenomena usually contribute to a signi"cantly increased
well-ordered but not ideal graphite layers in the wall of the yield of synthesized materials, as compared to deposition
nanotube (Fig. 3b). Generally, there was no obvious or from the gas phase (24). Large-scale, low-temperature, and
consistent connection of the tubes to nickel particles. In thus low-cost synthesis of the carbon nanotubes would
some cases, lattice fringes were inclined relative to the tube certainly be a great step toward serious applications of these
longitudinal axis, causing helicity. It is not clear whether the materials.
nanotubes were formed directly from C or from the
The results of the present study also have geochemical
ꢄꢅ
amorphous carbon, to which C was transformed under implications. Hydrothermal solutions with a temperature
ꢄꢅ
the experimental conditions.
ranging between 50 and well over 5003C are common and
The carbon nanotubes synthesized hydrothermally in this widespread in the Earth's crust (27). The hydrothermal
work are open-ended and slightly thicker and moreover #uids, which circulate below the Earth's surface, participate
have a larger central cavity than typical carbon nanotubes in a variety of geological processes by leaching, transport,
prepared by the carbon arc method (typically 2}25 nm in and precipitation of their mineral constituents or by hy-
diameter with an inner hollow diameter of 1}3 nm) (5). They drothermal alteration of existing minerals (27). Taking into
do not exhibit as high graphitization level as the nanotubes account these facts, the formation of carbon nanotubes
prepared in the electric arc (5) but they have a quality under hydrothermal conditions, demonstrated in our ex-
comparable to that of carbon nanotubes prepared by the periments, implies the presence of natural carbon nanotubes
CVD techniques (10, 12}14). CVD synthesis of carbon in carbon-rich rocks, in the areas of the Earth's crust, which
nanotubes has recently attracted wide attention. This tech- had been in#uenced by the hydrothermal #uids. It is worth
nique has the capability to fabricate at relatively low tem- mentioning that very recently Osawa et al. discovered car-
peratures nanotube arrays with controlled alignment bon nanotubes in coals and carbonaceous rocks (28). Nat-
(10, 13, 14) on patterned catalysts (Ni, Co, Fe, etc.) (7, 13), urally occurring fullerenes (29}32) could also be
which is of great importance for #at panel displays due to signi"cantly a!ected by the geological hydrothermal activ-
excellent "eld emission properties of carbon nanotubes ities. Since the stability and transformation conditions of
(7, 14). While in the present work we do not provide direct
evidence of nanotube alignment, we do demonstrate that a geothermometer. Our hydrothermal experiments with
hydrothermal processing may have potential similar to that were conducted for a very short time on a geological
C
were established in our work, C could be useful as
ꢄꢅ
ꢄꢅ
C
ꢄꢅ
of CVD in this regard. A capability to hydrothermally scale (only hours as compared to years in the case of geo-
synthesize aligned carbon "laments has been con"rmed logical processes) but they suggest that the fullerenes might
elsewhere (18). In addition, deposition and patterning of scarcely occur in nature, partially due to the low hydrother-
layers of Ni as a catalytic material can easily be accomp- mal stability of C . However, while a fullerene mine still
ꢄꢅ
lished by the hydrothermal technique (or electrodeposition) seems to be science "ction, the discovery of large-scale
on a variety of substrates, even at room temperature (24), natural deposits of carbon nanotubes may be closer to
which indicates that hydrothermal processing could be ap- reality.
plied in this case for the entire device fabrication (24).
An important implication of our results is the possibility
of large-scale hydrothermal synthesis of bulk carbon
nanotubes. In the follow-up studies, polyethylene (25) and
amorphous carbon (26) were used as much cheaper carbon
sources under similar hydrothermal conditions. The hy-
drothermal synthesis of carbon nanotubes from amorphous
carbon did not require the presence of metal catalysts. The
quality of the carbon nanotubes prepared in those works
was similar to ours, while the material cost was reduced
signi"cantly. The hydrothermal technique is known to give
products with a much higher homogeneity than solid state
processing and with higher density than gas or vacuum
processing (faster growth rate) (20). Moreover, liquids accel-
erate di!usion, adsorption, reaction rate, and crystallization
(nucleation and growth), especially under hydrothermal
conditions (20). Under supercritical conditions, the #uid has
some of the advantages of both a liquid and a gas. Di!usion
in supercritical #uids is higher than that in liquids and the
ACKNOWLEDGMENTS
This research was supported by the &&Research for the Future'' Program
No. 96R06901 of the Japanese Society for the Promotion of Science (JSPS)
and by the Research Institute of Solvothermal Technology (RIST), Japan.
The authors are greatly indebted to Charles Oakes (Rutgers University) for
reviewing the manuscript.
REFERENCES
1. H. W. Kroto, J. R. Heath, S. C. O'Brien, R. F. Curl, and R. E. Smalley,
Nature 318, 162 (1985).
2. S. Iijima, Nature 354, 56 (1991).
3. W. Kratschmer, L. D. Lamb, K. Fostiropoulos, and D. R. Hu!man,
Nature 347, 354 (1990).
4. C. M. Lieber and C. C. Chen, Solid State Phys. 48, 109 (1994).
5. T. W. Ebbesen, Annu. Rev. Mater. Sci. 24, 235 (1994).
6. A. Thess et al., Science 273, 483 (1996).
7. S. Fan, M. G. Chapline, N. R. Franklin, T. W. Tombler, A. M. Cassell,
and H. Dai, Science 283, 512 (1999).

188
SUCHANEK ET AL.
8. M. Yudasaka, R. Kikuchi, T. Matsui, Y. Ohki, S. Yoshimura, and 21. M. S. Dresselhaus, G. Dresselhaus, and P. C. Eklund, J. Raman
E. Ota, Appl. Phys. ¸ett. 67, 2477 (1995). Spectrosc. 27, 351 (1996).
9. B. K. Pradhan, T. Toba, T. Kyotani, and A. Tomita, Chem. Mater. 22. B. L. Zhang, C. Z. Wang, C. T. Chan, and K. M. Ho, Phys. Rev. B 48,
10, 2510 (1998). 11381 (1993).
10. Z. F. Ren, Z. P. Huang, J. W. Xu, J. H. Wang, P. Bush, M. P. Siegal, 23. M. R. Stetzer, P. A. Heiney, J. E. Fischer, and A. R. McGhie, Phys.
and P. N. Provencio, Science 282, 1105 (1998). Rev. B 55, 127 (1997).
11. K. Mukhopadhyay, A. Koshio, N. Tanaka, and H. Shinohara, Jpn. 24. M. Yoshimura, W. L. Suchanek, and K. Byrappa, MRS Bull. 25, 17
J. Appl. Phys. 37, L1257 (1998). (2000).
12. H. Araki, H. Kajii, and K. Yoshino, Jpn. J. Appl. Phys. 38, L1351 25. Y. Gogotsi, J. A. Libera, W. L. Suchanek, and M. Yoshimura,
(1999). J. Mater. Res. 15, 2591 (2000).
13. H. Ago, T. Komatsu, S. Ohshima, Y. Kuriki, and M. Yumura, Appl. 26. J. M. Calderon Moreno, S. S. Swamy, T. Fujino, and M. Yoshimura,
Phys. ¸ett. 77, 79 (2000).
14. Y. Chen, D. T. Shaw, and L. Guo, Appl. Phys. ¸ett. 76, 2469 (2000).
15. W. K. Hsu et al., Nature 377, 687 (1995).
Chem. Phys. ¸ett. 329, 317}322 (2000).
27. F. Pirajno, &&Hydrothermal Mineral Deposits.'' Springer-Verlag,
Berlin, 1992.
16. Y. G. Gogotsi and M. Yoshimura, Nature 367, 628 (1994).
28. E. Osawa et al., Carbon, submitted (1999).
17. X. Z. Zhao, R. Roy, K. Cherian, and A. Badzian, Nature 385, 513 29. K. T. Daly, P. R. Buseck, P. Williams, and C. F. Lewis, Science 259,
(1997).
1599 (1993).
18. Y. G. Gogotsi and K. G. Nickel, Carbon 36, 937 (1998).
30. L. Becker et al., Science 265, 642 (1994).
19. C. A. Eckert, B. L. Knutson, and P. G. Debenedetti, Nature 383, 31. D. Heymann, L. P. F. Chibante, R. R. Brooks, W. S. Wolbach, and
313 (1996).
R. E. Smalley, Science 265, 645 (1994).
20. M. Yoshimura and W. Suchanek, Solid State Ionics 98, 197 (1997).
32. P. H. Fang et al., Innov. Mater. Sci. 1, 129 (1996).