Catalytic growth of carbon nanoballs with and without cobalt encapsulation

Article abstract of DOI:10.1016/S0009-2614(00)01080-0

Carbon nanoballs encapsulated with cobalt were prepared by decomposition of methane. After acid treatment, the encapsulated cobalt can be removed to produce hollow carbon nanoballs. The content of hollow carbon nanoballs as high as 90 vol% in acid-treated

Full text of DOI:10.1016/S0009-2614(00)01080-0

3
November 2000
Chemical Physics Letters 330 (2000) 41±47
www.elsevier.nl/locate/cplett
Catalytic growth of carbon nanoballs with and without cobalt
encapsulation
Ziyi Zhong , Huayi Chen , Songbei Tang , Jun Ding , Jianyi Lin a,*,
a
a
a
b
a
Kuang Lee Tan
a
Department of Physics, Surface Science Laboratory, National University of Singapore, 10 Kent Ridge Cresent,
Singapore 119260, Singapore
Department of Materials Science, National University of Singapore, 10 Kent Ridge Cresent, Singapore 119260, Singapore
b
Received 31 July 2000; in ®nal form 19 September 2000
Abstract
Carbon nanoballs encapsulated with cobalt were prepared by decomposition of methane. After acid treatment, the
encapsulated cobalt can be removed to produce hollow carbon nanoballs. The content of hollow carbon nanoballs as
high as 90 vol% in acid-treated carbon product can be obtained by increasing the content of cobalt from 50 to 75 mol%
in the catalysts. However, under the same catalytic conditions, the main carbon products are multi walled carbon tubes
(
2 3 2 3 2 4
MWNTs) over Co=Al O , Co=La O , Co=CeO and Ni/MgO catalysts by decomposition of CH , or by decomposition
of CO over Co/MgO catalysts. These carbon nanoballs encapsulated with cobalt are novel magnetic materials. Ó 2000
Elsevier Science B.V. All rights reserved.
1
. Introduction
ducts can be roughly divided into two types of
methods. The ®rst is mainly based on the sub-
limation of carbon in an inert atmosphere, such as
electric arc-discharge process, laser ablation or
sublimation by solar energy, and the second in-
cludes catalytic decomposition of suitable organic
precursors, electrolysis in a molten ionic salt, heat
treatment of polymers and low-temperature solid
pyrolysis, etc. [9,10]. In most cases, the structures
and morphologies of carbon products vary with
the di€erent preparation methods and conditions.
It has been reported that single- and multi-walled
carbon nanotubes (SWNT and MWNT), fullere-
rene molecules, carbon nanopolyhedra, or their
mixture, have been prepared by arc-discharge and
laser ablation methods [11±14], and by catalytic
chemistry methods, SWNTs and MWNTs or their
mixture can also be obtained [15±18]. Moreover,
Since the discovery of buckminsterfullerene
molecules in 1985 [1] and carbon nanotubes [2] in
991, there has been a worldwide research e€ort on
1
novel carbon products due to their unique physical
properties that could potentially impact on broad
areas of science and technology. For example, it
has been reported that nanotubes have potential
applications as superconductors [3], single-mole-
cular transistors [4,5] and carbon nanotubes or
particles ®lled with metal or metal oxides have
potential applications in magnetic recording tech-
nology [6±8]. The synthesis of these carbon pro-
*
Corresponding author. Fax: +65-874-6126.
E-mail address: phylinjy@nus.edu.sg (J. Lin).
0009-2614/00/$ - see front matter Ó 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 9 - 2 6 1 4 ( 0 0 ) 0 1 0 8 0 - 0

4
2
Z. Zhong et al. / Chemical Physics Letters 330 (2000) 41±47
even the same carbon product can possess various
structure and shape. For example, the SWNTs
have theoretically three conformations, namely
armchair, zig-zag and chiral, and a particular
carbon nanotube can further exhibit various
shapes in straight, curved, helical and planarspiral
eter of 1.5 cm, and prior to the reaction, the cat-
alyst was pre-reduced at 700°C for 40 min in the
¯ow of hydrogen. The reaction was then con-
ducted at 600°C for 1 h in the ¯ow of CH
4
with a
¯ow-rate of 30 ml/min. After reaction, the mixture
was cooled in hydrogen and subjected to treatment
with concentrated sulfuric acid to dissolve metallic
catalyst. The microstructure and the morphology
of the carbon product were observed by low and
high resolution transmission electron microscope
(TEM and HRTEM) on a JEM-100CX and a
Philips FEG CM300 electron microscope, respec-
tively, and SEM and EDX measurements were
carried out on a XL-FEG scanning electron mi-
croscopy. Magnetic measurements were carried
out using a superconducting vibration sample
magnetometer (Oxford Instruments) at room
temperature. The saturation magnetization was
measured at the maximum ®eld of 3 T. Prior to the
magnetic measurement, the powder was shaped
into small pellet.
[
11,18,19]. So we believe the carbon production
reaction should be an ideal model catalytic reac-
tion to study the relationship between the nature
of the catalysts, reaction kinetics and the struc-
tures as well as the morphologies of carbon
products. However, we have found that, though
several reaction models have been proposed
[
11,20±22], the catalytic mechanism and reaction
kinetics are still not very clear, and in most cases,
the control of the composition and structure of
carbon product still poses a big challenge. Our
strategy is to study systematically the nature of
catalyst and the carbon product formation mech-
anism, and if possible, to develop technologies to
control the structure and morphology of carbon
products. Here we demonstrate a simple catalytic
approach to the growth of carbon nanoballs under
controlled conditions.
3. Results and discussion
Table 1 lists the composition of catalysts, cat-
alytic reaction conditions and reaction results.
Figs. 1±3 show some of TEM and SEM images of
carbon products. We found that typical MWNTs
were obtained from the decomposition of methane
on Ni/MgO catalyst (Fig. 1a). These carbon tubes
are slightly curved in shape and have a hollow
structure inside. The diameters of tubes range
from 10 to 40 nm. In the carbon product produced
from NiCo/MgO catalyst, we observed few ring-
like carbon balls besides the curved MWNTs. We
guessed that the growth of carbon nanoballs
should be related to the existence of Co sites in the
catalyst, because we did not observe similar
structure in the carbon product from Ni/MgO
catalyst. The guess was soon con®rmed from our
experiments. Fig. 1b shows a TEM image of car-
bon product from Co/MgO catalyst. It is clearly
seen that most of the carbon products are carbon
ring-like nanoballs. The diameters of these carbon
nanoballs range from 50 to 200 nm. We have also
found that by simply increasing the content of
supported cobalt from 50 to 75 mol% in Co/MgO
2
. Experimental
The preparation of catalysts was carried out
according to the method described in literature
23]. For example, 2.91 g of nitrate salt of cobalt
[
was mixed with 2.56 g of magnesium nitrate (50
mol% Co/MgO) and dissolved in distilled water,
followed by addition of citric acid with a molar
ratio of citric acid: metal ions ˆ 1:1. The aqueous
solution was gelled at 60±70°C with stirring and
dried at 200°C for 1 h. The dried powder was
further heated in air at 400°C for 1 h, and 700°C
for 5 h. For comparison, pure MgO and Co
powder, which were prepared according to the
above procedures using their nitrate as precursors,
were also used as catalysts, respectively, and
several catalysts were calcined at 900°C for 2 h.
All the above chemicals were from Aldrich and in
AR grade.
The catalytic conditions were studied very
carefully. For a typical catalytic reaction, 0.20 g of
catalyst was loaded in a quartz tube with a diam-

Z. Zhong et al. / Chemical Physics Letters 330 (2000) 41±47
43
Table 1
Summary of the catalyst composition, reaction conditions and results
Cat.
Transiton
metal
Reaction
temp.
Reaction gas
Flow rate
3
(cm /min)
Reaction
time
(min)
Carbon product
composition
(mol%)
(°C)
NiCo/MgO
Ni/MgO
25 + 25
600
CH
4
30
60
MWNTs + few
nanoballs
MWNTs
50
50
50
600
600
600
CH
CH
CO
4
30
30
30
60
60
60
Co/La
2
O
3
4
MWNTs
Co/MgO
Co/MgO
Co/MgO
MWNTs + few
nanoballs
50
75
600
600
CH
4
4
30
30
60
60
MWNTs + ca. 50%
nanoballs
CH
MWNTs + ca. 90%
nanoballs
MgO
Co
0
100
50
600
600
500
700
600
700
600
700
CH
CH
CH
CH
CH
CH
CH
CH
4
4
4
4
4
4
4
4
30
30
30
30
30
30
15
60
60
60
60
60
3
No carbon product
Carbon sheets
Co/MgO
Co/MgO
Co/MgO
Co/MgO
Co/MgO
CoMgO
MWNTs + nanoballs
MWNTs + nanoballs
MWNTs + nanoballs
MWNTs + nanoballs
MWNTs + nanoballs
MWNTs + nanoballs
50
50
50
50
20
60
60
50
catalysts, the content of carbon nanoballs as high
as 90 vol% can be obtained in acid-treated carbon
product. When the content of Co in the catalysts is
lower than 30 mol%, the carbon product is pre-
dominantly composed of slightly curved MWNTs.
However, in the case of using pure MgO as cata-
lysts, we found there is almost no change in color
for the white MgO catalyst after reaction, indi-
cating that MgO is not catalytically active to the
decomposition of methane at 600°C (Table 1).
When pure Co was used as catalyst, which was
prepared by heating and reducing the mixture of
cobalt nitrate and citric acid, only sheet-like car-
bon was observed (Fig. 1f and Table 1).
the reaction products are determined by the com-
position of catalysts and reaction gases. Obviously
using of Co/MgO catalyst and methane gas favors
the formation of carbon nanoballs.
In¯uence of reaction conditions on carbon
products was also studied for decomposition re-
action of methane on the Co/MgO catalyst. It is
found that there is no obvious di€erence for the
carbon product in morphology as the reaction
temperature was changed from 500°C to 700°C,
or/and the ¯ow-rate of methane gas from 15 to
3
60 cm /min, or/and the reaction time from 3 to
120 min. Our TEM and HRTEM results indicated
that more amorphous carbon was covered on the
carbon nanoballs at higher reaction temperature
or higher methane ¯ow rate and longer reaction
time. Fig. 1d shows a TEM image with reaction
time only 3 min. It clearly shows the ring-like
structure of carbon nanoballs. The formation of
nanoballs proceeds very quickly. An extension
of reaction time only increases the quantity of
product.
The e€ects of various supports and catalytically
active metal elements and reaction gases on
products were also studied under the same reac-
tion conditions (Table 1). MWNTs and minor
component of amorphous carbon were observed in
product from Co=La O , and only amorphous
2
3
carbon was seen in product from Fe/MgO. In the
case of using CO as reaction gas and Co/MgO as
catalyst, the carbon product mainly consisted of
MWNTs with smaller diameter compared to the
product of methane decomposition, but few car-
bon nanoballs could still be observed (Fig. 1c). So
Fig. 2 shows a SEM image of carbon product
from decomposition of methane on Co/MgO cat-
alyst. The carbon product is composed of nano-
spheres with diameters ranging from 50 to 200 nm.

4
4
Z. Zhong et al. / Chemical Physics Letters 330 (2000) 41±47
Fig. 1. TEM images of acid-treated carbon product produced by decomposition of methane on: (a) 50 mol% Ni/MgO; (b) 75 mol%
Co/MgO at 600°C for 1 h; (c) decomposition of CO on 50 mol% Co/MgO at 500°C for 1 h; (d) decomposition of methane on 75 mol%
Co/MgO at 600°C for 3 min; (e) raw carbon product from decomposition of methane on 50 mol% Co/MgO at 600°C for 3 min and
treated by ultrasound for 4 h. The wall of carbon nanoballs encapsulated with catalysts is discernable; (f) sheet-like carbon product
from decomposition of methane on pure Co catalyst at 600°C for 1 h. The pure Co catalyst was prepared using the mixture of cobalt
nitrate and citric acid as precursors.

Z. Zhong et al. / Chemical Physics Letters 330 (2000) 41±47
45
formation of carbon nanoballs without typical
hollow channel in the wall suggests the formation
mechanism may be as follows: the deposition of
carbon from dissociation of methane was initiated
on the surface of the catalyst particles that was
exposed to the reaction gas. The carbon further
di€used through the bulk and other surfaces (these
surfaces were connected the bulk of the catalyst) of
the catalyst particles to form layered graphite
structure. The repulsion among the carbon nano-
balls due to their continuous growth would let
them be lifted from the bulk of the catalysts with
their encapsulated catalyst particles. In this way,
the new catalytically active sites were created. So
after a short reaction time, we still can obtain
carbon nanoballs but in smaller quantity. How-
ever, the catalyst particles would be deactivated
due to encapsulation of amorphous carbon as
observed in our HRTEM measurement, which
indicated the existence of amorphous carbon layer
on the inner and outside surfaces of carbon
nanoballs after longer reaction time or higher re-
action temperature. In all TEM observations, for
samples, before acid treatment, we did not ®nd any
carbon nanoballs without encapsulation of cata-
lyst particle. This suggests that there is no carbon
nucleation taking place in the reaction gas. As
observed in the Fig. 2, the carbon nanoballs have
small `windows' on the surface. We believe these
small windows on the carbon nanoballs originated
from the unclosed surfaces that exposed to the
reaction gas and were used to deposit the carbon.
The encapsulation of catalyst particles in car-
bon nanoballs was con®rmed directly by TEM
measurements. Originally it was very dicult to
discern the edge of the carbon nanoballs from the
encapsulated metal catalyst particles in the raw
reaction mixture, because the particles were cov-
ered with amorphous carbon. We tried to remove
part of amorphous carbon from the particle sur-
face by heat treatment of the mixture at 300°C for
2 h in air or by ultrasound treatment for 4±5 h
before TEM measurement. Similar results were
observed for the samples treated with the two
methods. Fig. 1e shows a TEM image of the re-
action mixture (no acid treatment but with ultra-
sound treatment) from Co/MgO catalyst. Clearly,
the contour of the carbon shell (marked with an
Fig. 2. SEM image of carbon nanoballs from decomposition of
methane at 600°C for 1 h on 75 mol% Co/MgO.
Fig. 3. Image of HRTEM (a) of carbon nanoballs produced
from decomposition of methane on 75 mol% Co/MgO catalyst
at 600°C for 1 h.
These particles are rather regular in shape and
most of them have an open window on the surface.
HRTEM image (Fig. 3) shows that it is void in
the center of carbon nanoballs but, in the wall of
nanoballs, there is no hollow structure or channel
as in the case of MWNTs. At the same time a
multiwalled structure is discerned. The interlayer
distance in the wall is 0.33 nm indicating the
multiwall is composed of graphite sheets. The

4
6
Z. Zhong et al. / Chemical Physics Letters 330 (2000) 41±47
Table 2
catalyst particles were trapped in the center of
nanoballs, but they could be easily removed by
acid treatment. The observation of the catalyst
particles encapsulated in the cage of carbon
nanoballs before acid treatment suggests that the
carbon nanoballs were grown out from the nu-
cleation of carbon on the surfaces of catalyst
particles. Also of interest is the in¯uence of sup-
ports and reaction gas on carbon products. We
Mean composition of several carbon nanoballs determined by
EDX for raw carbon product from decomposition of methane
on 50 mol% Co/MgO at 600°C for 3 min
Element
Co
Mg
wt%)
C
O
(
8
8.30
0.42
10.98
0.30
arrow) is seen along the outer edge of the particles,
and the metal particles are indeed trapped in
the middle of the nanoballs. The EDX analysis
2 3 2 2 3
found the use of Al O ; CeO and La O as the
support of Co catalysts or using CO as reaction
gas resulted in the formation of MWNTs with a
slight curvature. A further study is being con-
ducted to clarify this circumstance.
(
Table 2) showed that in this sample the mean
cobalt content is ca. 88.3 wt% in several carbon
nanoballs encapsulated with cobalt, indicating only
cobalt was ®lled in the cage of the carbon balls.
Magnetic results show that the carbon nano-
balls encapsulated with cobalt particles are mag-
netic materials (Fig. 4). It is found that the
saturation magnetization for Co/MgO decreases
with the deposition of carbon. Adversely, the co-
ercive force increases with the deposition of car-
bon. So their magnetic properties are tunable by
simply controlling deposition of carbon.
In summary, the carbon nanoballs are pre-
pared by decomposition of methane on Co/MgO
catalysts. The content of nanoballs in carbon
product as high as 90 vol% can be obtained by
simply controlling the cobalt content at ca.
Acknowledgements
We thank Dr. Ping Chen, Mr. Qi Hong and Ms.
Lin Ji for their technical assistance and helpful
discussions, Dr. Yunjie Chen in Department of
Materials Science and Ms. Tjiu Wuiwui in De-
partment of Physics for SEM and HRTEM mea-
surements, respectively. The ®nancial support
from National Science and Technology Board of
Singapore (GR6773) is gratefully acknowledged.
References
75 mol% in catalysts. It is also found that the
[1] H.W. Kroto, J.R. Heath, S.C. O'Brien, R.F. Curl,
R.E. Smalley, Nature 318 (1985) 162.
2] S. Iijima, Nature 354 (1991) 56.
[
[
3] A.Yu. Kasumov, R. Deblock, M. Kociak, B. Reulet,
H. Bouchiat, I.I. Khodos, Yu. B. Gorbatov, V.T. Volkov,
C. Journet, M. Burghard, Science 284 (1999) 1508.
4] Sander J. Tans, Alwin R.M. Verschueren, Cees Dekker,
Nature 292 (1998) 49.
[
[
[
5] S. Frank, Z.L. Wang, W.A. de Heer, Science 280 (1998)
1744.
6] M.E. Mchenry, K. Gallagher, S.A. Majetich, E. Kirkpa-
trick, J.H. Scott, M.Q. Huang, in: K.M. Kadish, R.S.
Ruo€ (Eds.), Fullerences: Recent Advances in the Chem-
istry and Physics of Fullerences and Related materials,
Electrochemical Society, Pennington, NJ, 1996, p. 703.
7] V.P. Dravid, J.J. Host, M.H. Teng, B.R. Ellioit, J.H.
Hwang, D.L. Johnson, T.O. Mason, J.R. Weertman,
Nature 374 (1995) 602.
[
Fig. 4. Magnetic curves for catalyst 75 mol% Co/MgO reduced
at 700°C in H and reacted with methane at 600°C for 0, 10 and
0 min, respectively. Inset in the ®gure shows the M±H be-
havior in the lower magnetic ®eld range.
[8] B.K. Pradhan, T. Kyotani, A. Tomita, Chem. Commun.
(1999) 1317.
2
6
[9] K.B. Jirage, J.C. Hulteen, C.R. Martin, Science 278 (1997)
655.

Z. Zhong et al. / Chemical Physics Letters 330 (2000) 41±47
47
[
10] R.H. Bauhman, C. Cui, A. Zakhidov, Z. Iqbal, J.N.
Barisci, G.M. Spinks, G.G. Wallace, A. Mazzoldi, D. De
Rossi, A.G. Rinzler, O. Jaschinski, S. Roth, M. Kertesz,
Science 284 (1999) 1340.
[16] Y. Zhang, H.B. Zhang, G.D. Lin, P. Chen, Y.Z. Yuan,
K.R. Tsai, Appl. Catal. A. 187 (1999) 213.
[17] M. Terrons, N. Grobert, J.P. Zhang, H. Terrones, J.
Olivares, W.K. Hsu, I.P. Hare, A.K. Cheetham, H.W.
Kroto, D.R.M. Walton, Chem. Phys. Lett. 285 (1998) 299.
[18] Gao, Z.L. Wang, S. Fan, J. Phys. Chem. B 104 (2000)
1227.
[
[
11] S. Subramoney, Adv. Mater. 10 (1998) 15, 1157.
12] A. Thess, R. Lee, P. Nikolaev, H. Dai, P. Petit, J. Robert,
C. Xu, Y.H. Lee, S.G. Kim, A.G. Rinzler, D.T. Colbert,
G.E. Scuseria, D. Tomanek, J.E. Fischer, R.E. Smalley,
Science 273 (1996) 483.
[19] W. Ebbesen, H.J. Lezec, H. Hiura, J.W. Bennett, H.
Ghaemi, T. Thio, Nature 382 (1996) 54.
[
13] C. Journet, W.K. Maser, P. Bernier, A. Loiseau, M.
Lamyde la Chapple, S. Lefrant, P. Deniard, R. Lee, J.E.
Fischer, Nature 388 (1997) 756.
[20] N.M. Rodriguez, A. Chambers, R.T.K. Baker, Langmuir
11 (1995) 3862.
[21] M. Yudasaka, T. Ichihashi, S. Iijima, J. Phys. Chem. B 102
(1998) 10201.
[
[
14] Z. Shi, Y. Lian, X. Zhou, Z. Gu, Y. Zhang, S. Iijima, H. Li,
K.T. Yue, S.L. Zhang, J. Phys. Chem. B 103 (1999) 8698.
15] J.-F. Colomer, G. Bister, I. Willems, Z. Konya, A.
Fonseca, G. Van Tendeloo, J.B. Nagy, J. Chem. Soc.
Chem. Commun. (1999) 1343.
[22] M. Yudasaa, R. Yamada, N. Sensui, T. Wilkins,
T. Ichihashi, S. Iijima, J. Phys. Chem. B. 103 (1999) 6224.
[23] P. Chen, H.B. Zhang, G.D. Lin, Q. Hong, K.R. Tsai,
Carbon 35 (1997) 1495.