Upload
steemwheel
View
218
Download
0
Embed Size (px)
Citation preview
8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
1/10
Catalytic growth of carbon nanotubes throughCHNO explosive detonation
Yi Lu, Zhenping Zhu *, Zhenyu Liu
State Key Laboratory of Coal Conversion, Institute of Coal Chemistry, Chinese Academy of Sciences, Taiyuan 030001, PR China
Received 7 April 2003; accepted 4 November 2003
Abstract
Multi-walled carbon nanotubes (CNTs) have been efficiently synthesized by a self-heating detonation process, operated at low
loading densities of picric acid (PA), which acts as the explosive to provide needed high temperatures and parts of carbon sources.
Paraffin or benzene provides additional carbon source for tube assembling and hydrogen source to capture oxygen in PA to form
H2O and thus to survive some carbon species from oxidation. Cobalt nanoparticles, in situ formed from a detonation-assisted
decomposition and reduction of cobalt acetate, show good catalytic activity for nanotube nucleation and growth and for dispro-
portionation reaction of CO generated from the PA detonation. The nanotubes and catalyst particles are characterized by SEM,
TEM, EDX, SAED, XRD, and Raman spectroscopy techniques. Some tubes are well crystallized but others have lots of structural
defects, especially for the tubes with thin walls and bamboo-like shapes. The catalyst particles show conical shapes and exhibit a fcc
crystalline structure of parent cobalt. These data also experimentally show that tube growth is at a very high rate and suggest that it
is possible for a large-scale synthesis of CNTs under high-density and high-pressure conditions.
2003 Elsevier Ltd. All rights reserved.
Keywords: A. Carbon nanotubes; B. Catalyst; C. Electron microscopy; Raman spectroscopy; X-ray diffraction
1. Introduction
Carbon nanotubes (CNTs) can be thought as
graphite sheets built from the hexagonal lattice of sp 2
bonded carbon wrapped into seamless cylinders. CNTs
are intriguing one-dimensional systems with nano-scale
diameters and micron-scale or greater lengths, and are
ideally suitable for fundamental studies of atomically
well-defined materials. Strong in-plane CC bond, un-
ique dimension, and hollow channel endow CNTs with
excellent electronic, mechanical, gas-storing and cata-
lytic properties. It has been suggested that CNTs can
find important applications in numerous fields, such as
enforced polymer composites, electron field emitters,
quantum nanowires, nano-scale gas store devices, cata-
lysts and catalyst supports [1,2].
Since the synthesis of CNTs by Iijima [3], many
methods have been developed and proven effective for
CNTs growth, such as arc-discharge of graphite elec-
trode [4,5], laser ablation of graphite [6], catalytic
decomposition of hydrocarbon [79], and catalytic dis-
proportionation of CO [10,11]. Although there are many
differences among these methods in technique as well as
in chemical and physical principles, one common
ground is that CNTs always grow under high-tempera-
ture conditions, from several hundreds to a few thou-
sands. At high temperatures, carbon precursors are
decomposed or evaporated and then condensed to build
the sp2 graphite networks of CNTs. The needed high
temperatures are normally obtained from external
heating, which is highly energy-consumed. Theoreti-
cally, such a problem can be solved by employing hugely
exothermic reaction systems. The detonation of explo-
sives represents such a reaction system. During the
detonation, an enormous amount of heat is released by
transformation of chemical energy involved in explosive
molecules to thermal energy, which results in very high
temperatures. In addition, most of explosives are car-
bon-rich organic compounds, which can fully or par-
tially provide carbon building blocks for the CNTs
construction. Therefore, the detonation of carbon-rich
explosives can be designed to provide a self-heating
method for CNTs synthesis.
* Corresponding author. Tel.: +86-351-404-8310; fax: +86-351-404-
1153.
E-mail address: [email protected] (Z. Zhu).
0008-6223/$ - see front matter 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carbon.2003.11.001
Carbon 42 (2004) 361370
www.elsevier.com/locate/carbon
http://mail%20to:%[email protected]/http://mail%20to:%[email protected]/8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
2/10
It is well known that the detonation of nitro-hydro-
carbon (CHNO) explosives can produce nano-sized pure
carbon particles with both sp3 and sp2 bonded structures
[12]. Detonation synthesis of sp3 diamond nanoparticles
has been studied extensively and industrialized success-
fully [1216]. The synthesis generally employs high-en-
ergy explosives and operates at very high loadingdensities (23 g/cm3) to reach a ChapmanJouguet (CJ)
state 1 with extremely high pressures and temperatures,
typically a few tens GPa and thousands degree [1216].
In theory [17,18], there is a large difference between the
surface energies of bulk graphite (40 Kcal/mol) and
diamond (70 Kcal/mol) phases. A size reduction of
carbon particles to nano-scale results in a further
development of the difference in surface energy and
enlarges the differences of graphite and diamond phases
in surface tension and in Gibbs free energy and thus
significantly raise the pressure level of the graphite
diamond boundary line in phase diagram compared
with that of bulk carbon. Diamond nanoparticles are
formed at a CJ point that is above the raised graphite
diamond boundary line, at which sp3 diamond structure
is thermodynamically stable. On the other hand, the
raised graphitediamond boundary line of the size-re-
duced carbon widens the graphite phase zone, that is,
sp2 graphite structures can be well formed under mild
detonation conditions. If proper conditions are selected,
it is possible for the detonation-resulted small carbon
species to assemble into CNTs.
Following this idea, we recently synthesized CNTs
successfully using the detonation of explosives under
much mild conditions [19,20]. Before our reports,tubular carbon products have also been observed in the
soot from the detonation of special explosives,
1,2:5,6:11,12:15,16-tetrabenzo-3,7,9,13,17,19-hexadehy-
dro annulene [21] and 2,4,6-triazido-s-triazine [22], with
2% CNTs in the products [22]. We employed some
common CHNO explosive such as m-dinitrobenzene
(DNB) [19] and picric acid (PA) [20], introduced cata-
lysts into the detonation system and operated at very
low pressure and temperature, typically a few tens MPa
and several hundreds degree. Under such conditions,
CNTs can well grow, with 3040% and 8090% contents
for the cases of DNB and PA, respectively. Transition
metal catalysts such as Fe, Co and Ni were found to be
indispensable for the CNTs growth under the mild
detonation conditions employed, and the activity of Co
is much higher than Ni and Fe. Moreover, the intro-
duction of hydrocarbons such as paraffin into PACo
system further increases CNTs yield.
The present paper focuses the attention on some re-
cent experimental results for PACo detonation system
and provides outlined descriptions on the chemistry of
the catalysis-assisted detonation process, the relation-
ship between the experimental conditions and tubesgrowth, structure and morphologies as well as on the
formation, structure and roles of the catalyst particles.
2. Experimental
All of the chemical reagents used, including PA,
benzene, and cobalt acetate (Co(AC)2 4H2O) were of
analytical grade without further pre-treatment. The li-
quid paraffin employed has a boiling point of 300 C,
and carbon and hydrogen contents of 82 wt% and 18
wt%, respectively. Before the detonation experiments,
the starting materials were mixed mechanically in de-
sired ratios.
The detonation was performed in a sealed stainless
steel pressure vessel (10.8 cm3) connected with a pressure
gauge, and induced by external heating (20 C/min) to
310 C (ignition temperature of PA). The experiments
conditions are summarized in Table 1. When the deto-
nation occurs, pressure and temperature peaks of 2040
MPa and several hundreds to one thousand degree were
generated inside the vessel, varied with the loading
density of PA (the amount of PA loaded in unit volume
of reactor, g/cm3). After the detonation, the vessel was
cooled in air to ambient temperature. The gaseousproducts were vented and the solid products were col-
lected for further characterizations. In some cases, the
gaseous products were also collected and analyzed by
gas chromatography.
The as-synthesized samples were examined using a
Hitachi H-600 transmission electron microscopy (TEM)
operated at 75 kV. High-resolution TEM imaging were
performed at a Philips CM 200FEG high resolution
TEM (HRTEM) operated at 200 kV, which is equipped
with energy dispersive X-ray spectrometer (EDX) for
elemental analysis. Scanning electron microscopy
(SEM) was carried out on a Philips XL 30-FEG oper-
ated at 30 kV. For TEM analyses, the samples were
prepared by sonicating in ethanol for 10 min, followed
by depositing one drop of the resulted suspension on
thin carbon film supported on holey copper grid. X-ray
diffraction (XRD) patterns of the products were re-
corded using a Rigaku D/max 2500X X-ray diffrac-
tometer operated at a step size of 0.02 with Cu Ka
radiation (k 154:178 pm) and a Ni filter at 40 kV and100 mA anode current. Diffraction profiles were ob-
tained by the usual h2h scan. Raman scattering was
conducted using Renishaw Micro-Raman 2000 equip-
ped with a 20 objective to give an illuminated spot.
1 CJ theory assumes that thermodynamic equilibrium of the
detonation products is reached instantaneously and that all products
are consumed completely. In the CJ state, the shock velocity of the
detonation wave equals to the sound speed behind the detonation
wave. CJ pressure is an intrinsic property of an explosive and is an
indicator of explosive performance.
362 Y. Lu et al. / Carbon 42 (2004) 361370
8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
3/10
The 632.8 nm line of heliumneon laser is used as
excitation source. The CNTs content is estimated as the
volume ratio of CNTs to the whole carbonaceous
products based on systematic TEM and SEM investi-
gations (random sampling and repeated observations).
3. Results and discussion
3.1. Detonation reaction systems and CNTs growth
The detonation of PA alone (sample 1) results in very
little solid carbon, about 5 wt% of carbon in fed PA. The
obtained solid product shows no tubular structure, and
contains only spherical carbon nanoparticles with sizes
of 1070 nm (Fig. 1a). Its XRD pattern (Fig. 1b)
exhibits single weak and broad diffraction peak, cen-
tered at 24.5 with interplanar distances of 0.360.37
nm, indicating that the sample is highly disordered
amorphous carbon. Analyses of the gaseous products
show that CO is dominant, followed by N2. Minority of
the gases are CO2, CH4 and H2O (noted that H2O
should be at high concentration but, during the cooling,
it heavily condenses on reactor wall). NH3 is also found
but with very low concentration. Therefore, the low
solid carbon yield is clearly associated with the high O/C
atomic ratio (7:6) and low hydrogen content (H/C 1:2)in PA molecule. These make most of the produced small
carbon species to be oxidized into gaseous CO because
the O/C atomic ratio is still high (0.92) after complete
oxidation of all H atoms into H2O. In the situation of
diamond synthesis, the detonation occurs at highFig. 1. TEM image (a) and XRD pattern (b) of the sample prepared
from the detonation of PA alone (sample 1).
Table 1
Experimental parameters and the related data about the detonation products
Sample Loading density
of PAa (g/cm3)
Co(AC)2/PA mass
ratio
Paraffin/PA mass
ratio
Benzene/PA mass
ratio
Solid carbon
yieldb (%)
CNTs contentc
(%)
1 0.2 0 0 0 5.12 0
2 0.2 1/5 0 0 11.18 40
3 0.2 1/10 0 0 9.17 50
4 0.2 1/20 0 0 8.58 905 0.2 1/30 0 0 5.44 80
6 0.2 1/20 1/2 0 50.25 40
7 0.2 1/20 1/4 0 33.31 80
8 0.2 1/20 1/6 0 21.38 80
9 0.2 1/20 1/8 0 16.67 90
10 0.2 1/20 0 1/2 40.21 30
11 0.2 1/20 0 1/4 26.60 50
12 0.2 1/20 0 1/6 18.78 80
13 0.2 1/20 0 1/8 14.78 90
14 0.1 1/20 1/4 0 26.0 50
15 0.05 1/20 1/4 0 20.6 20
16 0.02 1/20 1/4 0 14.4 5
a The amount of PA loaded in unit volume of the reactor. For the calculation, the total volume of the reactor was roughly taken as 10 cm3.b
Calculated as follows: Carbon yield M
P M
Co=M
C, whereM
P is the total amount of the detonation product;M
Co is the amount of metalcobalt involved in the product and is nearly equal to the amount of metal cobalt in starting material; MC is the amount of carbon involved in the
starting materials.c Estimated from the TEM and SEM observations.
Y. Lu et al. / Carbon 42 (2004) 361370 363
8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
4/10
pressure and temperature, and thermodynamics favors
the reaction of 2COfiCO2 + C (solid) [12]. Under the
experimental conditions used in this work, however, CO
is kinetically stable. The disordered spherical carbon
structures are likely associated with the relatively low
temperatures (several hundreds degree) generated from
the mild detonation, at which the carbon species aredifficult to spontaneously develop into organized sp2
networks and tubular structures without inducement of
a catalyst.
To enhance the solid carbon yield and development
of tubular graphite structure, Co(AC)2 is mixed with PA
at Co(AC)2PA mass ratios of 1/51/30 before the det-
onation experiment. The introduction of Co(AC)2 sig-
nificantly enhances the solid carbon yield and CNTs
growth (Table 1). In the products obtained (samples 4
and 5), tubular structural carbon with contents up to
8090% is dominant. Correspondingly, the amount of
CO is obviously decreased in the gaseous detonation
products whereas the concentration of CO2 is increased.
These results indicate that cobalt nanoparticles, in situ
formed from a detonation-assisted decomposition and
reduction of cobalt acetate (see below), play important
catalytic roles in the disproportionation reaction of CO
and in the nucleation and growth of CNTs. Both the
direct PA decomposition and the subsequent CO dis-
proportionation reaction on the surface of cobalt par-
ticles provide carbon building blocks for CNTs
assembling.
Considering the consumption of part of carbon
atoms from PA into carbon oxides, carbon- and
hydrogen-rich hydrocarbons such as benzene and liquidparaffin are introduced into the PACo(AC)2 system,
with hydrocarbonPA mass ratio from 1/8 to 1/2. The
heat produced from the detonation quickly decomposes
the hydrocarbon (may be also catalyzed by the cobalt
particles [7]) into small carbon species and hydrogen
radicals. The carbon species directly participate in CNTs
assembling, and the hydrogen radicals capture oxygen
to form H2O, which effectively depresses the oxidation
of carbon species and enhances the solid carbon for-
mation. The addition of hydrocarbons greatly increases
the product yield (Table 1).
Finally, an explosive mixture can be formulated,
including explosive PA, additive hydrocarbon (paraffin
or benzene), and catalyst precursor Co(AC)2, from
which CNTs can be effectively synthesized. In this sys-
tem, the introduction of paraffin or benzene results in
similar products (see below) despite significant differ-
ences in their molecular structures and in hydrogen
contents. Three parameters mainly influence the CNTs
growth. One is Co(AC)2PA mass ratio, with an optimal
value of about 1/20 by mass. Higher ratio would result
in the formation of large-sized cobalt particles, which
show low catalytic activity for CNTs growth (see below
for a detailed description). In this case, the obtained
products contain less CNTs and the produced CNTs are
obviously short, thick and poorly crystallized. Another
factor is the hydrocarbonPA mass ratio. Low ratio
(less than or equal to 1/4) leads to low carbon yield but
does not impact the content of CNTs, and about 80%
tube content can be obtained (Table 1). Higher hydro-
carbonPA mass ratios bring high carbon yields butheavily decrease tube contents. The other factor is the
loading density of PA, which determines the system
temperature and thus is crucial for CNTs growth and
the subsequent annealing process. A systematic TEM
examination on CNTs synthesized at different PA
loading densities (from 0.02 to 0.2 g/cm3) shows that the
CNTs content in the obtained products nearly increases
linearly from 5% to 80% (Table 1). At a low loading
density, the obtained CNTs obviously have more
structural defects, although the defects are still evident
in the cases of 0.2 g/cm3 (see below). Based on this
information, further tailoring of the structural defects is
possible by raising the PA loading density to generate
higher temperatures.
3.2. Morphologies and basic aspects of CNTs
The detonation products are very voluminous, uni-
form, and fine black powders. Elemental analyses by
EDX show that the products mainly consist of carbon
and cobalt. In addition, very little amounts of oxygen
and nitrogen are also detectable, possibly from adsorp-
tion of H2O, N2 and/or other oxygen- and nitrogen-
containing gases. HNO3 and HClO4 treatments [23] ofthe detonation products were performed to remove co-
balt and sp2 bonded carbon. The result shows no resi-
due, suggesting that no sp3 diamond carbon is
produced.
Figs. 2 and 3 are typical SEM and TEM images of the
samples obtained from the detonations of PA
Co(AC)2paraffin and PACo(AC)2 benzene mixtures,
respectively. These images indicate that these two dif-
ferent mixtures result in similar detonation products.
CNTs are dominant in the products with about 7090%
in content. Carbon-encapsulated cobalt nanoparticles,
some short and thick carbon fibers, and amorphous
carbon particles are also evident in the samples.
The inner and outer diameters of the CNTs are in the
range of 825 and 1250 nm, peaked at about 20 and 35
nm, respectively. The lengths of the CNTs are up to 40
lm, peaked at about 20 lm. No single-walled tubules
are observed under the used preparation conditions. The
tube ends are normally closed on one side and capped by
catalyst particles on the other side. Some open ends are
also observed, which may be as grown naturally or
generated during the subsequent sample preparation for
TEM analyses. All tubes are separated from each other
and frequently show tangled morphologies.
364 Y. Lu et al. / Carbon 42 (2004) 361370
8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
5/10
The CNTs normally show cylindrical morphologies,
with hollow channel along the full length (Figs. 2b and
3b), but some bamboo-like tubes, constructed by many
jointed hollow compartments are also present in small
amount for the samples obtained under optimal condi-
tions. The bamboo-like tubes (Fig. 4) are primarily
found at high hydrocarbon/PA ratios (e.g. higher than 1/
3 by mass) or at low PA loading densities (e.g. less than
0.1 g/cm3), which seem mean that the formation of
bamboo-like tubes is associated with relatively high
carbon density and low temperature environments since
the decomposition of the massive hydrocarbon mole-
cules engenders considerable carbon species and con-
sumes large amount of the heat generated from the PA
detonation. This suggestion is supported by other
observations that the amounts of amorphous carbon
particles and thick carbon fibers with poor graphite
structures are visibly high under these conditions.
3.3. Structures of CNTs
Fig. 5 shows typical XRD patterns of the as-synthe-
sized samples obtained from the detonations of PA
Co(AC)2paraffin and PACo(AC)2 benzene mixtures.
Unlike the XRD pattern of the sample without catalyst
(Fig. 1b), both of the two patterns exhibit one strong
and narrow diffraction peak at about 26.2, which is
corresponding to an interlayer spacing of 0.34 nm
associated with the (0 0 2) planes of hexagonal graphite
structure, suggesting the formation of sp2 bonded
graphite crystals. However, the peak is rather asymme-
try and broader than that of the well-crystallized
graphite, indicating a relatively low stacking order (Lc)
and crystalline dimension (La) [24] in the tube struc-
tures. The other three diffraction peaks are generated by
the cobalt catalyst particles and will be described below
in detail.
Raman spectra taken from the samples obtained
from the two different mixtures are depicted in Fig. 6.
These two spectra are rather similar and exhibit mainly
two bands at 1328 cm1 (D-band) and 1593 cm1 (G-
band). The origin of D-band has been explained as
disorder-induced features due to the finite particle size
effect, lattice distortion [25], or amorphous carbon
background signals [26]. The G-band has been explained
as the stretching vibration mode of graphite crystals,
Fig. 2. Typical SEM (a) and TEM (b) images of the product obtained
from the detonation of PACo(AC)2paraffin (sample 7).
Fig. 3. Typical SEM (a) and TEM (b) images of the product obtained
from the detonation of PACo(AC)2benzene (sample 12).
Y. Lu et al. / Carbon 42 (2004) 361370 365
8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
6/10
which is also a characteristic of nanotubes [27]. It is well
known that the ratio (ID=IG) of the D-band intensity tothe G-band intensity is inversely proportional to La
value of graphite crystalline [25]. For our samples, ID is
higher than IG, indicating the faulty long-range ordered
crystalline and the existence of disordered carbon
structures, in agreement with the XRD results. The
disordered carbon structures are not only associated
with some defect-rich nanotubes but also with the
impurities including amorphous carbon particles, short
carbon fibers, and amorphous carbon layers on the
outer tube surfaces (see below). However, the position
and the well-defined shape of the D and G bands point
to a well-ordered graphene-based texture of the ob-
tained CNTs.
HRTEM analyses provide more clear and detailed
information for the CNTs wall structures. Fig. 7a is a
representative image of the walls of most tubes, which
reveals the multi-walled structure of the obtained CNTs.
The tube walls are organized by several to tens carbon
layers, which are clearly resolved and are parallel to the
tube axis. The distance between the planes is about 0.34
nm, which is consistent with XRD results and associated
with the (0 0 2) planes of well-crystallized graphite
structures. Selected-area EDX analyses confirm that
carbon is the exclusive element of the tube walls. No
nitrogen is incorporated into the graphite structures,
although the used PA contains lots of nitro groups,which are reduced to N2 and even NH3 during the
detonation.
Fig. 7a also reveals that the inner surfaces of the tubes
are clean, but the outer surfaces are covered with thin
amorphous carbon layers. Similar phenomena were also
observed frequently for the CNTs prepared from cata-
lytic decomposition of hydrocarbon and explained as
the rate of hydrocarbon feedstock pyrolysis over-
whelming the rate of tubes growth [11], in agreement
with the situation in our detonation process, in which
the generation of all the carbon species for tube con-
struction are extremely fast, on a microsecond time scale
[28]. In addition, the formation of the amorphous car-
bon layers is also possibly caused by the great fluctua-
tion in temperature, which is mainly derived from the
energy emitted from the exothermic decomposition of
explosive and drops continuously after the detonation
due to the lack of external heating (during the experi-
ments, the reactor was heated externally to 310 C to
induce the detonation, after which the heating was
stopped). Under such conditions, it is easy to imagine
that some un-reacted carbon species would deposit on
the already-formed tube surfaces as disordered struc-
tures at decreased temperatures. In the sense of appli-
Fig. 4. Bamboo-like tubes produced from the detonation of PA
Co(AC)2 paraffin mixture with a mass ratio of 2:0.11:1 and at a
loading density of 0.2 g/cm3
for PA (sample 6).
Fig. 5. XRD patterns of the detonation products prepared from (a)
PACo(AC)2 paraffin (sample 9) and (b) PACo(AC)2benzene
(sample 12).
Fig. 6. Raman shift spectra of the samples obtained from the deto-
nation of (a) PACo(AC)2paraffin (sample 9) and (b) PACo(AC)2
benzene (sample 12).
366 Y. Lu et al. / Carbon 42 (2004) 361370
8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
7/10
cation, the covered amorphous carbon layers are trou-
blesome for electronic purposes but may be helpful for
chemical surface modification and for catalysis purposes
since they are more reactive than graphite crystals.
The structure of the tube walls seems to have a rela-
tionship with their thickness. Well-crystallized structures
normally appear on the tubes with relatively thick walls,
for example, the tubule shown in Fig. 7a has a wall with
40 carbon layers. Fig. 7b shows a tubule with 7 carbon
layers, which appear wavy and discontinued, suggesting
that lots of structural defects are involved and may be
associated with the jumps in spacing at points where the
helix angle changes [29]. In addition, defects are also
frequently observed on the walls of the bamboo-like
tubes (Fig. 7c), especially at the joints of the compart-
ments. The greatly fluctuated temperature as mentioned
above and the high rate of carbon deposition (derived
from both the decreased temperature and the high
density of carbon species as shown below) are likely
responsible for the formation of the structural defects.
Such defects and the covering amorphous carbon layers
are believed to be tailorable by controlling the reactor
temperature to a proper level or by further annealing the
as-synthesized samples in inert atmosphere at high
temperatures [30].
3.4. Formation and structures of the catalyst particles
The presence of catalysts is necessary for the CNTs
growth in this method, at least under the experimental
conditions used. Cobalt shows a high catalytic activity.
In the detonation products, the cobalt catalyst exists as
nano-sized particles with various morphologies and lo-
cated at various positions (Fig. 8). Most of the cobalt
nanoparticles are located at the tube ends and show
conical shapes (Fig. 8a), or filled inside of spherical
onion-like carbons (Fig. 8c). Some particles are filled
inside compartments of bamboo-like tubes and located
near the joints of different compartments, with conical
appearances (Figs. 4 and 8b). Some particles are con-
tinuously filled inside of some tubules, forming as metal
wires (Fig. 8c).
In the XRD patterns (Fig. 5), three peaks at 44.3,
51.5 and 75.9 are indexed to the diffraction lines from
the (11 1), (20 0) and (22 0) planes of face-centered
cubic (fcc) cobalt, respectively. Signals for cobalt
Fig. 7. Typical HRTEM images of the normal tubes with thick (a) and thin (b) walls and of bamboo-like tubes (c). These images were taken from the
sample 7.
Y. Lu et al. / Carbon 42 (2004) 361370 367
8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
8/10
carbides or oxides are not detected. Selected-area elec-tron diffraction (SAED) of individual catalyst particles
as shown in the inset of Fig. 9 further confirms that the
catalyst nanoparticles have a fcc single crystal structure
of parent cobalt, independent on their morphologies and
the locations. From the TEM analyses (Figs. 8 and 9), it
has been also observed that nearly all the catalyst par-
ticles are covered with carbon layers (including the tube
walls), which coincides with the fact that treating the
samples with 1 mol/l HNO3 solution at ambient tem-
perature shows no cobalt dissolution.
In the formation of the nano-sized cobalt particles,
the PA detonation plays an important role. Despite theadopted precursor (Co(AC)2 4H2O) can decompose at
298 C before the ignition of PA at 310 C, however, the
decomposition is rather uncompleted because the tem-
perature increase from 298 C to 310 C only needs very
short time at the used high heating rate of 20 C/min. In
addition, the decomposition of cobalt acetate (microm-
eter-sized powders) alone results in the formation of
micrometer-sized cobalt oxide particles, as shown by theTEM and XRD analyses of the product obtained from a
2-h decomposition reaction of cobalt acetate at 310 C.
This result suggests that the nano-sized cobalt particles
in the detonation products are really formed in situ. The
heat and the carbon species produced from the deto-
nation promote the decomposition and the reduction of
cobalt acetate, respectively. The detonation wave likely
helps the size control.
The presences of the coating carbon layers around the
catalyst particles and the conical catalyst particles at the
tube ends suggest that the mechanism of tube growth in
this process is similar to that in the catalytic decompo-
sition of hydrocarbons. For the latter process, it has
been widely accepted [3137] that the tube growth fol-
lows a way in which carbon atoms adsorb, dissolve, and
diffuse into the catalyst particle interior to form a metal
carbon solid state solution. Tube growth occurs when
supersaturation leads to carbon precipitation into a
crystalline tubular form. According to the mechanism,
the sizes of the catalyst particles regulate diameters of
CNTs, which is supported by the fact that the sizes of
the catalyst particles located at the tube ends are nor-
mally close to the tube diameters (left of Fig. 8a).
However, the catalyst particles with sizes larger than
tube diameters are also observed frequently at the endsof some tubules, for example, the size of the metal
particle shown at the right in Fig. 8a is about two times
of the tube diameter. This phenomenon especially ap-
pears in the cases with high Co(AC)2PA mass ratios, at
which both the contents and lengths of CNTs are re-
duced concomitantly. These observations suggest that
during the tubules growth, the catalyst particles grow as
well, possibly due to a continuous sorption of additional
metal species from vapor phase. The catalyst particle
growth is likely responsible, at least partly (combined
with the decreased local temperature) for the termina-
tion of tube growth by lengthening diffusion path and
blocking the diffusion of carbon species in the Co par-
ticles (because the diffusion of carbon through the cat-
alyst particles limits the tube growth [35]), which results
in a complete encapsulation of catalyst particles with
carbon layers and deactivates the catalyst.
It is of importance to emphasize that the physical and
chemical environments generated from the detonation
for the CNTs growth are much different from those of
other processes, the latter are normally at a relatively
clean atmosphere, ambient pressure, and low density
of feedstock. In the present process, the detonation-
generated atmosphere is rather complex. Particularly, all
Fig. 8. Some typical TEM images of Co catalyst particles presented in
various morphologies and at various positions, (a) and (c) were taken
from sample 7, (b) from sample 6.
Fig. 9. HRTEM images of the carbon-covered Co particles (sample 7)
at tube end, the inset shows a SAED pattern of the core.
368 Y. Lu et al. / Carbon 42 (2004) 361370
8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
9/10
the carbon species for tube assembling are supplied on a
microsecond time-scale at a high gas pressure (tens
MPa) [28]. Under such conditions, the density of carbon
species just behind the detonation wave (at this time,
carbon deposition starts) is extremely high, typically
larger than 0.008 mol/cm3 for carbon atom, while the
tube growth is much effective so that 80% tube content isobtainable. This fact implies that the rate of the cata-
lyzed CNTs growth is considerably high, supporting the
theoretical estimation from Gamaly and Ebbesen: a
multi-wall CNT with 5 nm diameter and 1000 lm
length grows in 103104 s [38]. Therefore, the results
obtained in this work also suggest that CNTs can be
effectively synthesized under rather harsh condi-
tions with very high pressure and feedstock density,
which is of importance for production of CNTs in large-
scale.
4. Conclusion
CNTs have been efficiently synthesized by a self-
heating detonation process, using PAcobalt acetate
paraffin/benzene as the explosive mixture. The formed
CNTs show multi-walled and tangled morphologies,
with inner diameters of 825 nm, outer diameters of 12
50 nm and lengths up to 40 lm. Some of the tubes are
well crystallized but others have lots of structural de-
fects, especially for the tubes with thin walls and bam-
boo-like shapes. The nanotubes are normally capped by
catalyst particles at one end. The catalyst particles showconical shapes and exhibit a fcc crystalline structure of
parent cobalt. It is also found that the formed cobalt
nanoparticles have a catalytic role in the dispropor-
tionation reaction of CO generated from the PA deto-
nation.
In this detonation approach, three parameters,
Co(AC)2 PA ratio, hydrocarbonPA ratio, and the
loading density of PA, are of special importance in the
tube growth. The former two parameters exhibit optimal
values of about 1/20 (by mass) and 1/4 (by mass),
respectively, while, for the latter parameter, the best
results appear at 0.2 g/cm3 that is the highest value we
have used. Co(AC)2 PA ratio determines the amount
and the size of catalyst nanoparticles. HydrocarbonPA
ratio greatly influences product yield and also the CNTs
content and morphology. The loading density of PA,
which determines the system temperature, is crucial for
CNTs growth.
Although the samples obtained currently do not
contain single-walled nanotubes and fullerenes, it is
believed that their syntheses are possible by similar
processes through designing the reactor and explo-
sive mixture and/or adjusting the detonation condi-
tions.
Acknowledgements
Authors thank Natural Scientific Foundation of
China for partial financial support (no. 59872047), Mr.
Hermann Sauer at Fritz Haber Institute of MPG, Ger-
many, for his helps in some TEM analyses, and the
referees for their nice suggestions.
References
[1] Ajayan PM. Nanotubes from carbon. Chem Rev 1999;99(7):1787
99.
[2] Subramoney S. Novel nanocarbons-structure, properties and
potential applications. Adv Mater 1998;10(15):115771.
[3] Iijima S. Helical microtubules of graphitic carbon. Nature
1991;354(6348):568.
[4] Ebbesen TW, Ajayan PM. Large-scale synthesis of carbon
nanotuebs. Nature 1992;358(6383):2202.
[5] Bethune DS, Kiang CH, de Vries MS, Gorman G, Savoy R,
Vazquez J, et al. Cobalt-catalysed growth of carbon nanotu-
bes with single-atomic-layer walls. Nature 1993;363(6430):605
7.
[6] Thess A, Lee R, Nikolaev P, Dai H, Petit P, Robert J, et al.
Crystalline ropes of metallic carbon nanotubes. Science
1996;273(5274):4837.
[7] Li WZ, Xie SS, Qian LX, Chang BH, Zou BS, Zhou WY, et al.
Large-scale synthesis of aligned carbon nanotubes. Science
1996;274(5293):17013.
[8] Terrones M, Grobert N, Ollvares J, Zhang JP, Terrones H,
Kordatos K, et al. Controlled production of aligned-nanotube
bundles. Nature 1997;388(6637):525.
[9] Kong J, Cassell AM, Dai H. Chemical vapor deposition of
methane for single-walled carbon nanotubes. Chem Phys Lett
1998;292(46):56774.
[10] Dai H, Rinzler AG, Nikolaev P, Thess A, Colbert DT, SmalleyRE. Single-wall nanotubes produced by metal-catalyzed dispro-
portionation of carbon monoxide. Chem Phys Lett 1996;260(3
4):4715.
[11] Kitiyanan B, Alvarez WE, Harwell JH, Resasco DE. Controlled
production of single-wall carbon nanotubes by catalytic decom-
position of CO on bimetallic CoMo catalysts. Chem Phys Lett
2000;317(35):497503.
[12] Greiner NR, Phillips DS, Johnson JD, Volk F. Diamond in
detonation soot. Nature 1988;333(61716172):4402.
[13] Chen Q, Yun S. Nano-sized diamond obtained from explosive
detonation and its application. Mater Res Bull 2000;35(12):1915
9.
[14] Saha DK, Koga K, Takeo H. Structural properties of diamond
fine particles and clusters prepared by detonation and decompo-
sition of TNT. Surf Sci 1998;400(13):1349.[15] Kuznetsov VL, Chuvilin AL, Moroz EM, Kolomiichuk VN,
Shaikhutdinov SK, Butenko YV, et al. Effect of explosion
conditions on the structure of detonation soots:ultradisperse
diamond and onion carbon. Carbon 1994;32(5):87382.
[16] Vereschagin AL. Detonation Nanodiamonds. Altai State Techni-
cal University Barnaul, Russian Federation, 2001 (in Russian).
[17] Ree FH, Winter NW, Glosli JN, Viecelli JA. Kinetics and
thermodynamic behavior of carbon clusters under high pressure
and high temperature. Physica B 1999;265(14):2239.
[18] Viecelli JA, Ree FH. Carbon particle phase transformation
kinetics in detonation waves. J Appl Phys 2000;88(2):68390.
[19] Lu Y, Zhu Z, Wu W, Liu Z. Catalytic formation of carbon
nanotubes during detonation of m-dinitrobenzene. Carbon
2003;41(1):1948.
Y. Lu et al. / Carbon 42 (2004) 361370 369
8/3/2019 Yi Lu, Zhenping Zhu and Zhenyu Liu- Catalytic growth of carbon nanotubes through CHNO explosive detonation
10/10
[20] Lu Y, Zhu Z, Wu W, Liu Z. Detonation chemistry of CHNO
explosive:catalytic assembling of carbon nanotubes at low
pressure-temperature state. Chem Commun 2002;22:27401.
[21] Boese R, Matzger AJ, Vollhardt KPC. Synthesis, crystal struc-
ture, and explosive decomposition of 1, 2:5, 6:11, 12:15, 16-
tetrabenzo-3, 7, 9, 13, 17, 19-hexadehydro annulene: formation of
onion- and tube-like closed-shell carbon particles. J Am Chem Soc
1997;119(8):20523.
[22] Kroke E, Schwarz M, Buschmann V, Miehe G, Fuess H, Riedel
R. Nanotubes formed by detonation of C/N precursors. Adv
Mater 1999;11(2):15861.
[23] Lewis RS, Ming T, Wacker JF, Anders E, Steel E. Interstellar
diamonds in meteorites. Nature 1987;326(6109):1602.
[24] Takahashi H, Kuroda H, Akamatu H. Correlation between
stacking order and crystallite dimensions in carbons. Carbon
1965;2(4):4323.
[25] Tuinstra F, Koenig JL. Raman spectrum of graphite. J Chem
Phys 1970;53(3):112630.
[26] Boskovic BO, Stolojan V, Khan RUA, Haq S, Silva SRP. Large-
area synthesis of carbon nanofibres at room temperature. Nature
Mater 2002;1(3):1658.
[27] Tan PH, Hu CY, Li F, Bai S, Hou PX, Cheng HM. Intensity and
profile manifestation of resonant Raman behavior of carbonnanotubes. Carbon 2002;40(7):11314.
[28] Cooper PW, Kurowski SR. Introduction to the technology of
explosives. New York: VCH Publishers; 1997, p. 1995.
[29] Liu M, Cowley JM. Structures of the helical carbon nanotubes.
Carbon 1994;32(3):393403.
[30] Bom D, Andrews R, Jacques D, Anthony J, Chen B, Meier MS,
et al. Thermogravimetric analysis of the oxidation of multiwalled
carbon nanotubes: evidence for the role of defect sites in carbon
nanotube chemistry. Nano Lett 2002;2(6):6159.
[31] Charlier JC, Iijima S. Growth mechanism of carbon nanotubes.
Topics Appl Phys 2001;80:5581.
[32] Amelinckx S, Zhang XB, Bernaerts D, Zhang XF, Ivanov
V, Nagy JB. A formation mechanism for catalytically growth
helix-shaped graphite nanotubes. Science 1994;265(5172):6359.
[33] Dai H. Nanotube growth and characterization. Topics Appl Phys
2001;80:2953.
[34] Tibbetts GG. Why are carbon filaments tubular? J Cryst Growth
1984;66(3):6328.
[35] Snoeck J-W, Froment GF, Fowles M. Filamentous carbon
formation and gasification: thermodynamics, driving force, nucle-
ation and steady-state growth. J Catal 1997;169(1):2409.
[36] Snoeck J-W, Froment GF, Fowles M. Kinetic study of the carbon
filament formation by methane cracking on a nickel catalyst
1997;169(1):25062.
[37] Baker RTK. Catalytic growth of carbon filaments. Carbon
1989;27(3):31523.[38] Gamaly EG, Ebbesen TW. Mechanism of carbon nanotube
formation in the arc discharge. Phys Rev B 1995;52(3):2083
9.
370 Y. Lu et al. / Carbon 42 (2004) 361370