Studies of LaNiO3 used as a precursor for catalytic carbon nanotubes growth

Article abstract of DOI:10.1016/j.molcata.2006.10.014

The growth of the carbon nanotubes is carried out on nickel containing catalysts. A fine distribution of metal nanoparticles is obtained by the reduction in situ of a perovskite phase. The LaNiO₃ perovskite is synthesized by three different methods in order to compare the nature of the tubes obtained after CH₄-CVD. These structures are analyzed by DRX, H₂-TPR and SEM before catalytic test. After CVD, the internal diameter of the tubes is constant no matter the method of preparation of the perovskite and the external diameter depends on the distribution of the metallic particles. The homogeneous distribution of the CNT is related to the homogeneous distribution of the metallic particles obtained after reduction under methane.

Full text of DOI:10.1016/j.molcata.2006.10.014

Journal of Molecular Catalysis A: Chemical 265 (2007) 209–217
Studies of LaNiO used as a precursor for
3
catalytic carbon nanotubes growth
∗
M. Kuras, R. Roucou, C. Petit
LMSPC, Laboratoire des Mat e´ riaux, Surfaces et Proc e´ d e´ s pour la Catalyse, UMR 7515 du CNRS,
ECPM, 25 rue Becquerel, 67087 Strasbourg C e´ dex 2, France
Received 17 May 2006; received in revised form 4 October 2006; accepted 10 October 2006
Available online 14 October 2006
Abstract
The growth of the carbon nanotubes is carried out on nickel containing catalysts. A fine distribution of metal nanoparticles is obtained by the
reduction in situ of a perovskite phase. The LaNiO perovskite is synthesized by three different methods in order to compare the nature of the
tubes obtained after CH –CVD. These structures are analyzed by DRX, H -TPR and SEM before catalytic test. After CVD, the internal diameter
3
4
2
of the tubes is constant no matter the method of preparation of the perovskite and the external diameter depends on the distribution of the metallic
particles. The homogeneous distribution of the CNT is related to the homogeneous distribution of the metallic particles obtained after reduction
under methane.
©
2006 Elsevier B.V. All rights reserved.
Keywords: Perovskite; MWNT; CVD; SEM; Mass spectrometry
1
. Introduction
Mixed metal oxides as perovskites have been recently inves-
tigated as the precursor of catalyst [7]. The size of the active
transition metallic particles may be controlled by changing the
reduction conditions. Liang et al. [8] and Li et al. [9] have used
La2NiO4 as catalyst precursor for CNTs growth. The growth of
nanotubes was obtained by catalytic decomposition of CH4 over
Ni-nanoparticles from the pre-reduced oxide. Different meth-
ods of preparation of perovskite are known in solid chemistry
as grinding–milling of solids, co-precipitation, sol–gel and resin
methods.
The objectives of this work are the synthesis and characterisa-
tionsofLaNiO3 preparedbythreedifferentresinmethodscalled:
citrate, Pechiniandpropionate. ThenatureoftheCNTsproduced
by CH4–CVD are investigated as a function of the method of
perovskite’s preparation. The studied parameters measured by
HR-MET will be the internal and external diameters, the length
and the variety of tubes.
Carbon nanotubes have been the center of interest since their
discovery by Iijima in 1991 [1] for many reasons due to their
one-dimensional carbon nanostructures [2,3]. Due to the persis-
tent and resilient nature of sp C network, CNTs possess high
2
mechanical strength as well as flexibility [4]. The morpholo-
gies of nanotubes depend on the specific growth conditions.
Catalytic growth as a method of synthesis of nanotubes pro-
vides a large variety of forms: straight filaments, as well as bent
and helically wound tubes [5]. In most cases, the production of
nanotubes is based on the catalytic decomposition of carbon-
containing gases on metal nanoparticles dispersed on a support.
The nanotube diameter appears to be governed by the size of the
catalytically active particles. The length of the tubes depends
mainly on the reaction time. Various hypothetical mechanisms
have been proposed to explain the growth of nanotubes produced
by decomposition of hydrocarbons on active metals [6]. For the
catalytic method, it is crucial to select an effective catalyst with
the appropriate size of active metal.
2. Experimental
2.1. Preparation of the LaNiO3
Two types of sources were used to introduce the metallic
∗
elements: nitrate salts Ni(NO ) ·6H O and La(NO ) ·9H O.
Corresponding author. Tel.: +33 3 90 24 27 69; fax: +33 3 90 24 27 68.
E-mail address: petitc@ecpm.u-strasbg.fr (C. Petit).
3 2
2
3 3
2
The initial compounds are more than 99% pure (Strem).
1
381-1169/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2006.10.014

2
10
M. Kuras et al. / Journal of Molecular Catalysis A: Chemical 265 (2007) 209–217
◦ ◦ −1
5 to 650 C with a slope of 15 C min . The reducing mix-
In the citrate method the aqueous solution of nitrates were
2
−1
mixed with an aqueous solution of citric acid in order to obtain
metallic citrates. Ammonia is introduced to favour the trans-
formation of citric acid to citrate. The solvent is removed by
evaporation until the formation of resin. The resin undergoes a
ture was CH4 in helium (with a flow of 2.85 and 35 mL min
,
respectively). The gases coming out were analyzed by mass
spectrometry. Prior to any measurement, the mass spectrom-
eter is calibrated by introducing a known amount of carbon
monoxide, carbon dioxide, methane and hydrogen. A typical
measurement consists in the following steps: (i) the catalyst is
positioned in the central hot zone of the reactor; (ii) the reaction
mixture, methane in helium, is introduced; (iii) once the flow
is stabilised, the temperature is increased and the record of the
◦
thermal treatment at 400 C to break up the intermediate citrate
◦
complexes and then a treatment at 900 C for one night in order
to form LaNiO3:
La(NO3)3·9H2O + Ni(NO3)2·6H2O + citric acid + NH3
◦
→
LaNiO3
exhaust gas composition begins; (iv) after 1 h at 600 C the CH4
is stopped and the reactor is ventilated with helium. The SEM
analyses have been performed on a JEOL JSM-6700 F scanning
microscope. A very thin carbon deposit was used to improve the
conductivity of the sample. Transmission Electron Microscopy
The Pechini’s method follows the same steps of preparation
as the citrate method apart from the ethylene glycol addition.
The aqueous solutions of the two nitrate salts were mixed with
an aqueous solution of citric acid and ethylene glycol. The solu-
tion is heated and ammonia is added. The resin undergoes a
(
TEM) analyses were performed using a TOPCON EM 002-B
apparatus. BET surface area has been determined on SA 3100
Coulter sorptometer at 77 K. The visualisation of the structure
was made by the CaRIne software, which allows rebuilding the
elementary cell and the crystal starting from the space group,
the cell parameters and the position of the atoms. This software
provided by Divergent S.A. is a tool for research and teaching.
◦
thermal treatment at 400 C to break up the intermediate citrate
◦
complexes and then a treatment at 900 C for one night in order
to form LaNiO3:
La(NO3)3·9H2O + Ni(NO3)2·6H2O + citric acid
+
ethyleneglycol + NH3 → citrate + ester of La and of Ni
LaNiO3
3. Results and discussion
→
In the propionate method the salts were dissolved in hot pro-
pionic acid [10]. The solvent was evaporated until formation of
resin. Nitrates species are eliminated as reddish-brown nitrous
oxide gases during the elimination of propionic acid. The ther-
Perovskites, of the general stoichiometry ABO3 where A is
a trivalent rare earth metal ion and B is a trivalent first row tran-
sition metal ion [11]. Norman and Morris [12] reported diffrac-
tograms were Rietveld refined using the trigonal space group,
R-3C, and crystallographic parameters for the LaNiO3 system.
For the modelling of the structure, the rhombohedric structure
was used with the cell parameters (a = 5.4412, b = 5.4389 and
◦
mal treatment of 900 C for one night is also preceded by a step
◦
of thermal treatment at 400 C to break up the metal propionates:
La(NO3)3·9H2O + Ni(NO3)2·9H2O + propionic acid
◦
◦
c = 13.1715, α = β = 90 and γ = 119.8573 ) and the atomic posi-
tions of the ions in the unit cell given in [13]. The corresponding
representation is given in Fig. 1. The ions Ni3 are in a per-
→
LaNiO3
+
2
.2. Characterisation of the LaNiO3 and synthesis of
carbon nanotubes
The X-ray diffraction (XRD) experiments were performed
on a D5000 Siemens diffractometer using the Cu K␣ radiation
(λ = 0.15406 nm). The scattering intensities were measured over
an angular range of 10 < 2θ < 90 for all the samples with a step-
◦
size (2θ) of 0.02 and a step time of 2 s. The diffraction patterns
have been indexed by comparison with the JCPDS files (joint
committee on powder diffraction standards). The mean particle
size was calculated by the Debye–Scherrer’s relation. Temper-
ature programmed reduction (TPR) under H2 was performed
on a 50 mg sample placed in a U-shaped quartz tube (6.6 mm
◦
i.d.), the temperature was increased from 25 to 650 C with a
◦
−1
slope of 15 C min . The reducing mixture was 3% hydrogen
−
1
in helium (50 mL min ). Hydrogen consumption was quan-
tified by a TC detector after scavenging of the obtained water.
Temperature programmed reduction (TPR) under CH4 and CVD
was performed to reduce the oxide and to produce carbon tubes
in the same time. A sample of 50 mg was placed in a U-shaped
quartz tube (6.6 mm i.d.), the temperature was increased from
Fig. 1. The structure of LaNiO3 obtained with the parameters proposed by Nor-
man and Morris [12].

M. Kuras et al. / Journal of Molecular Catalysis A: Chemical 265 (2007) 209–217
211
SEM-micrographs of all the samples prepared by the three
methods are shown in Fig. 4a–d. The mean particle size was
estimated by visual inspection of the SEM-micrographs. The
estimated mean particle sizes were in the range of 50–100 nm
for citrate method and from 30 to 100 nm for the two other meth-
ods. The primary crystal sizes for LaNiO3 are in accordance with
XRD results, however, there is a tendency to agglomerate the
initial nanoparticles to larger particles. For example the citrate
sample present larger areas of cubic crystals stuck one to the
other by several faces. The statistical estimation of this sample
results in one or two free faces of six possible faces. The SEM
images of LaNiO3 sample prepared by propionate and Pechini’s
methods show small, rather monodispersed particles (Fig. 4c
and d). Propionate presents more or less the same morpholog-
ical aspect as the citrate one. However the catalyst grains are
less geometrical compared to the citrate and Pechini’s method
samples.
Fig. 2. XRD of LaNiO3 prepared by (a) citrate method; (b) Pechini’s method;
◦
(
c) propionate method after a treatment at 900 C.
3+
fect hexagonal environment, the La ions are in a deformed
environment of 12 O2
−
3.2. H -TPR and DRX after partial and total TPR
.
2
The reductibility of prepared samples of LaNiO3 is investi-
gated by H2-TPR and the results are presented in Fig. 5. Under
these H2-TPR conditions the La3 ions are non-reducible, so
the observed H2 consumption is ascribed only to the reduc-
3
.1. Characterizations of the perovskite
+
The XRD diagrams of LaNiO3 prepared by the three different
◦
2+
3+
methods were compared after calcination at 900 C in Fig. 2. In
each case, mainly the perovskite phase is obtained. In the case
of the Pechini’s method a low amount of NiO is detected. The
average size of crystallites of free NiO is estimated at 11.3 nm.
The mean crystallite diameters for the perovskite, calculated by
Scherrer’s equation, are found to be 21, 16 and 18 nm for the
citrate, Pechini and propionate methods, respectively. The XRD
analysis seems to indicate that the three methods of preparation
allow a nearly identical formation of the perovskite. It could be
considered that the three used methods are able to produce the
catalytic solids. It must be noted that the production of a unique
crystalline phase is one of the necessary conditions to obtain a
controlled size of carbon nanotubes.
These structures were compared to the modelised structure
and the rebuilt diffractogram using the software is given in Fig. 3.
It is found that the experimental diffractograms are in good
agreement with the XRD data. According to [13] the Rietveld
analysis indicates a deficiency in lanthanum and oxygen in the
structure.
tion of Ni or Ni ions. For all the samples, two reduction
zones can be distinguished at low and high temperature. The
low temperature zone presents a large asymmetrical peak of H2
consumption, a multiple step Ni2 or Ni reduction more com-
plex than that proposed by Provendier [10]. The reduction of
the citrate sample results in the most definite profile at lowest
+
3+
◦
temperature—an intense peak centered at 415 C followed by
◦
a small peak at 450 C. This kind of reduction profile has been
found for LaCoO3 by Hansteen et al. [14]. The sample prepared
by Pechini’s method exhibits only one large reduction peak in
◦
the low temperature zone centered at 435 C. It seems that the
small quantity of free NiO found for this sample by XRD does
not influence the reductibility of perovskite, which is confirmed
◦
also by the high temperature peak (575 C) at exactly the same
temperature as for citrate sample. The sample prepared by pro-
◦
pionate method presents wide low temperature peak at 440 C
which is closer to the Pechini’s method sample. However this
◦
sample shows a high temperature reduction at 555 C, slightly
lower temperature than the others. For all samples the second
Fig. 3. Corresponding diffractogram obtained by the modelling of the structure LaNiO3 with the space group R-3C.

2
12
M. Kuras et al. / Journal of Molecular Catalysis A: Chemical 265 (2007) 209–217
Fig. 4. SEM of LaNiO3 prepared by the citrate method (a) fairly dense zone, (b) very dense zone in mode LEI, (c) prepared by the Pechini’s method and (d) propionate
method with an enlarging of 50,000 in mode SEI.
0
high temperature zone presents very symmetrical peak of H2
consumption suggesting a quick one-step reduction.
–
for the second peak: reduction of Ni2+ to Ni . The size of
metallic particles calculated by XRD of the three samples after
total reduction at 650 C and partial reduction at 480 C are
presented in Table 1.
In all the cases, after total reduction the LaNiO3 structure was
destroyed and nickel metal deposited on lanthanum oxide was
obtained. After partial reduction, the formation of La2Ni2O5
is observed by XRD, which confirms that the reduction goes
through two distinct stages. In the case of the Pechini’s sam-
ple, the presence of nickel metal observed even before H2-TPR;
by XRD becomes larger. The mean crystallite diameter for
◦
◦
◦
The amount of H2 consumed in the range 415–440 C for
LaNiO3 sample can be attributed to the reduction of Ni3 to Ni
+
2+
occurring at the surface as well as in the bulk. Subsequently, the
◦
reduction peak at around 570 C was assigned to the reduction
2
+
of Ni to metallic nickel. The relative ratio of second peak to
the first one was calculated as 1.7, 2.0 and 2.2 for the citrate,
Pechini’s and propionate methods, respectively. These values
are close to 2 the quantitative analysis correlates well with the
H2 consumption – for the first peak: reduction of Ni3 to Ni
+
2+
La2Ni2O , calculated by Scherrer’s equation is found to be close
5
to those obtained for LaNiO3. The average crystallite diameter
for La2O3 remains in the same order of magnitude for the three
methods and comparable with that of the starting structure. It
Table 1
Phases and mean crystallite diameters of La2Ni2O5, La2O3 and Ni obtained by
XRD after partial or total reduction
Method
Citrate
Pechini
Propionate
La2O3 + Ni
Phases obtained after
total reduction
La2O3 + Ni
La2O3 + Ni
Size of La2O3
Size of Ni
13 nm
–
17 nm
16 nm
16 nm
14 nm
Phase obtained after
partial reduction
Size of La2Ni2O5
Ni and size of Ni
La2Ni2O5
La2Ni2O5
La2Ni2O5
16 nm
No
16 nm
Yes, 21 nm
19 nm
No
◦
Fig. 5. H2-TPR of LaNiO3 after a treatment at 900 C: (a) citrate method; (b)
Pechini’s method; (c) propionate method.

M. Kuras et al. / Journal of Molecular Catalysis A: Chemical 265 (2007) 209–217
213
could be supposed that nickel was reduced without modifying
construction completely.
Table 2
Ratios of calibrated gases obtained by MS and compared with the data obtained
according to the equation of reduction
Hansteen et al. [14] showed that La2Co2O oxidizes via
5
La3Co3O8 to LaCoO3. The reported study of intermediate cobalt
structures suggests the presence of anion vacancies which allows
the diffusion and fast reaction with oxygen through the volume.
The anion vacancy ordering is well known for the oxygen defi-
Method
Given by the
equation
Citrate
Pechini
Propionate
CO/CO ratio
H2/(CO + CO2) ratio
H2/CH4 ratio
CH4/(CO + CO2) ratio
2
1.36
1.85
1.85
1.00
1.50
4.10
0.84
4.90
1.20
4.36
1.17
3.73
1.33
1.86
0.43
4.28
cientperovskites(AMO3 )andresultsinsuperstructureswhere
−␦
the cationic sublattice of the perovskite is essentially unchanged.
The co-ordination of few or all metal atoms decreases the sym-
metry from octahedral to typically tetrahedral or square planar,
product is CO2 which presents the narrowest peak. CO and H2
are formed in two steps. This indicates the decrease of available
oxygen during the reduction. The reduction proceeds in two
stages and the carbonaceous deposit is formed as soon as the
active sites are available, confirmed by accompanied production
of H2 in time on stream.
e.g., La2Co2O and La2Ni2O [13]. The type of oxygen vacancy
5
5
ordering strongly depends on the nature of transition element.
Reduction of perovskite type oxides leads to loss of oxygen,
which is connected with the formation of oxygen vacancies.
Strong similarities exist between the structure LaCoO3 and
By considering the nature of produced a degree of reduction
LaNiO3 on the one hand, and La2Co2O and La2Ni2O on
5
5
of Ni3 from the perovskite structure could be established. The
+
the other hand. The TPR results can be seen as the reverse
mechanism of the oxidation of La2Ni2O and the reduction via
3+
0
total transformation (Ni to Ni ) is given by the equation as
follows:
5
La2Ni2O and perhaps by La3Ni3O8 until the metallic nickel on
5
La2O3 are proposed.
8LaNiO3 + 7CH4
→
3CO + 4CO2 + H2O + 13H2 + 8Ni + 4La2O3
3
.3. CH4–CVD and DRX after test: evolution of LaNiO3
Gommes et al. [15] presented a kinetic study of carbon nan-
otubes production by the use of a mass spectrometer. In situ
monitoring of the different species in effluent gas during the
CNTs growth process could provide precious information for
understanding the reduction of LaNiO3 under methane and the
CNTs growth in detail. Fig. 6 represents the profiles of varia-
tions of relative intensity for 5 m/z selected corresponding to the
consumption of CH4 and the formation of H2, H2O, CO, CO2
in time on stream for the citrate sample. All the experiments
are undertaken under the same conditions. It is observed that
◦
the temperature of 600 C is necessary in all the cases to have a
reaction between methane and perovskite. For all samples, the
reduction begins with a CH4 consumption which is converted
over the oxide into CO2, H2O, CO and H2. The first formed
Fig. 7. SEM of LaNiO3 prepared by the citrate method after CH4–CVD with
an enlarging of (a) 20,000 and (b) 50,000 in mode SEI.
Fig. 6. CH4–CVD of LaNiO3 prepared by citrate method.

2
14
M. Kuras et al. / Journal of Molecular Catalysis A: Chemical 265 (2007) 209–217
From this general equation, the quantity of methane neces-
11.5 nm for citrate, Pechini’s and propionate methods, respec-
tively.
3+
0
sary for the full reduction of Ni to Ni could be established. It
−3
is 1.78 × 10 mol. The ratios between the produced gases and
methane (CO/CO2, H2/(CO/CO2), H2/CH4, CH4/(CO/CO2))
are compared to the theoretical values in Table 2.
3.4. SEM and MET after test CH4–CVD: nanotubes growth
After calibrating the known amounts of gasses by SM, it
could be observed that the ratio related to the reduction only,
namely CO/CO2 ratio is close to the value found by the sug-
gested equation for all the samples. However the production of
H2 is larger than the calculated amount and it suggests that the
phenomenon of decomposition of methane to H2 and carbon
also takes place. Compared to the other methods, the propi-
onic method produces more water and less CO (figure not
shown). As the quantity of methane consumed for Ni reduc-
tion is known by the equation [1], the quantity of CH4 trans-
formed to carbon deposits could be also calculated. The val-
ues of 79%, 73% and 77% of CH4 for the citrate, Pechini’s
and propionate samples, respectively, are transformed into the
carbonaceous deposit. For all the samples the XRD show the
destruction of LaNiO3 structure to totally reduced nickel and
For the CNT formation mechanism the following steps have
been proposed [16]: (i) methane decomposition at the surface
of the nickel crystallite; (ii) formation of surface carbide in the
reaction zone; (iii) carbon diffusion into the crystal volume; (iv)
carbon release, after some over-saturation, at the Ni surface.
When using a defined structure oxide, it is necessary to reduce
the nickel at the first place. The advantage is that in LaNiO3 the
initial nickel structure is very homogeneously dispersed in vol-
ume. It could be distinguished by the nanotubes distribution on
thecatalyst, itsdistributioninsizes, lengthandform. Tohavefirst
idea on it, the samples were studied by SEM after CH4–CVD.
Figs. 7–9 show the examples, at two different magnifications of
◦
the nanotubes grown at 600 C for the three samples. Generally
the morphology of the initial system is preserved, especially for
the citrate sample. The obtained Ni on La2O3 shows tendency
to agglomerate but the shape and organisation of the grains is
similar to those of LaNiO3. The formation of the nanotubes is
◦
lanthanum oxide after the CH4–CVD at 600 C. The mean
crystalline diameters of Ni obtained by XRD are 15, 18 and
Fig. 8. SEM of LaNiO3 prepared by the Pechini’s method after CH4–CVD with
an enlarging of (a) 20,000 and (b) 50,000 in mode SEI.
Fig. 9. SEM of LaNiO3 prepared by the propionate method after CH4–CVD
with an enlarging of (a) 20,000 and (b) 50,000 in mode SEI.

M. Kuras et al. / Journal of Molecular Catalysis A: Chemical 265 (2007) 209–217
215
Fig. 10. TEM of LaNiO3 prepared by the citrate method after CH4–CVD with
an enlarging of 23,000.
Fig. 11. TEM of LaNiO3 prepared by the Pechini’s method after CH4–CVD
with an enlarging of 23,000.
The analysis of the samples after CH4–CVD continue with
TEM in order to study particularly the nanotubes (Figs. 10–12a
and b). Three parameters were studied for each method, namely:
the external and internal diameters of the tubes as well as their
length. The results are reported in Tables 3–5.
The measurements of the tubes formed on citrate sample
show a majority of fine tubes with external diameter in agree-
ment with the size of the nickel particles found by XDR after
CH4–CVD (15 nm). Sixty-five percent of these tubes were
formed with rather important length (about 20 times their exter-
nal diameter). The larger the tubes are, the longer they are but
the length to external diameter ratio is relatively constant (TEM
possible on this catalyst in big quantity as shown by the nan-
otubes starters on the surface. The distribution of these tubes
observed by SEM seems relatively heterogeneous in spite of the
single formation of the phase perovskite at the beginning. For
the Pechini’s method, the morphology of the catalytic sample
is preserved. A few large tubes are visible and the surface of
the catalyst is not covered by the small tube starters as for the
prepared citrate and propionate samples. This implies that for
citrate and Pechini’s samples, the quantity of CH4 transformed
into C and estimated by SM is mainly used for the growth of
the largest tubes. In the case of propionate sample, the surface
is completely covered by short and fine nanotubes with uniform
sizes and the support is not visible. The catalytic system appears
less geometrical than its LaNiO3 precursor.
1
0).
For the Pechini’s sample, the tubes are larger than those of
the citrate sample. The distribution is centered at external tubes
Fig. 12. (a and b) TEM of LaNiO3 prepared by the propionate method after CH4–CVD with an enlarging of 23,000.

2
16
M. Kuras et al. / Journal of Molecular Catalysis A: Chemical 265 (2007) 209–217
Table 3
Characterisations of tubes obtained by CH4–CVD on LaNiO3 prepared by citrate method
Method
Citrate
Percentage of tubes with different external diameter (%)
External tubes diameter (nm)
Internal tubes diameter (nm)
86
10–15
5.5–6
7
7
16–30
6–6.5
31–50
6–6.5
>500
Percentage of tubes with different length (nm)
33% of 40–80, 5% of 81–200, 62% of 201–400
15% of 200–300, 85% of 301–500
Table 4
Characterisations of tubes obtained by CH4–CVD on LaNiO3 prepared by Pechini’s method
Method
Pechini
Percentage of tubes with different external diameter (%)
External tubes diameter (nm)
Internal tubes diameter (nm)
10
76
31–50
6–6.5
5
9
10–30
5.5–6
50–150
51–100
6–6.5
400–450
>100
6–6.5
400 to >500
Percentage of tubes with different length (nm)
10% of 50–150, 60% of 151–250, 30% of 251–425
Table 5
Characterisations of tubes obtained by CH4–CVD on LaNiO3 prepared by propionate method
Method
Propionate
Percentage of tubes with different external diameter (%)
External tubes diameter (nm)
Internal tubes diameter (nm)
95
7.5–10
5.5–6
2
3
10–15
6–6.5
200 to >500
>30
6–6.5
>500
Percentage of tubes with different length (nm)
45% of 25–50, 45% of 51–100, 10% of 101–200
diameters between 30 and 50 nm. The tubes appear to be rel-
atively longer and with larger external diameters compared to
the CNT obtained by the citrate method (TEM 11). There is no
agreement between the value of the diameters of the tubes and
the average size of nickel (18 nm) measured by XRD probably
because of the presence of small amount of free nickel oxide over
the perovskite which is beneficial to the growth of larger tubes. It
is the only case in which the presence of some large tubes (exter-
nal diameter higher than 100 nm) containing migrated nickel in
the tubes during the growth is observed.
of the produced nanotubes shows that the internal diameter is
constant for all tubes independently of the size and the length
of the tubes as well as the method of preparation of the LaNiO3
precursor. It seems to indicate the minimal size of metallic nickel
obtained after LaNiO3 reduction by CH4–CVD for the forma-
tion of nanotubes as 5.5–6 nm. It could be suggested that tubes
are formed around one tube by building the larger concentric
tubes as larger as nickel particles. By taking into account the
thickness of a tube and by considering a distance between two
plans as a distance of Van der Walls interaction, it could be con-
cluded that the number of concentric tubes in majority formed
For the propionate method, TEM confirms the results of the
SEM namely the narrower distribution of fine carbon nanotubes
◦
by CH4–CVD with 600 C for LaNiO3 propionate sample is
(
TEM 12a). The external diameter of the tubes is equal to the size
about 5.
of nickel crystallites calculated to XDR (11.5 nm). The amor-
phous carbon was also observed (TEM 12b).
4. Conclusion
If one consider the quantity of deposited carbon on the sam-
ples the yields of the reactions are as follow: 1.2%, 1.8% and
This work concerned the use of a structure rich in nickel
(La/Ni ratio = 1) as a possible catalyst precursor for multiwalls
carbon nanotubes growth. It showed the feasibility of this reac-
tion with narrow distribution of nanotubes and the relationship
between distribution of the nickel particles produced by reduc-
2
.5% for propionate, citrate and Pechini’s sample. The quantity
of methane used in each reaction constitutes approximately 75%
of the total quantity of methane (Table 2). Twenty-five percent
of methane was used for structure reduction. So it means that
the conversion of methane to solid carbon in the CH4–CVD was
approximately equal to 1%, 1.3% and 2% for propionate, citrate
and Pechini’s sample.
All theses results confirm that the size of carbon nanotubes
has a direct relationship with the size of metallic particles. This
study shows the great sensitivity of the nature of carbon tubes to
the preparation conditions even in the case of definite structures
oxides as catalyst precursor. The most interesting method of the
synthesis LaNiO3 perovskite is the propionate method. How-
ever this work requires complementary studies to understand
deeper the reasons of these distinct evolutions. The observation
tion under CH . It is observed that the internal diameter is
4
constant for all the catalysts. The variation of the method of
preparation has a remarkable effect on the quantity and qual-
ity of the nanotubes, which indicates the great sensitivity of the
growth of the MWNT with nickel environment.
Acknowledgements
The authors would like to thank the GDRE 2756 for scientific
organization. They are indebted to Dr. S. Ivanova and R. Blell
for proof reading the manuscript.

M. Kuras et al. / Journal of Molecular Catalysis A: Chemical 265 (2007) 209–217
217
References
[8] Q. Liang, L.Z. Gao, Q. Li, S.H. Tang, B.C. Liu, Z.L. Yu, Carbon 39 (2001)
97.
8
[
9] H. Li, Q. Liang, L.Z. Gao, S.H. Tang, Z.Y. Cheng, B.L. Zhang, Z.L. Yu,
C.F. Ng, C.T. Au, Catal. Lett. 74 (2001) 185.
[
[
[
1] S. Iijima, Nature 354 (1991) 56.
2] C. Du, N. Pan, Mater. Lett. 59 (2005) 1678.
3] Y. Omata, Y. Yamagami, K. Tadano, T. Miyake, S. Saito, Physica E 29
[
[
[
[
10] H. Provendier’s thesis, ULP Strasbourg, 1999.
11] Glazer, Acta Cryst. A 31 (1975) 756.
12] A.K. Norman, M.A. Morris, J. Mater. Process. Technol. 92–93 (1999) 91.
13] K. Vidyasagar, A. Reller, J. Gopalakrishnan, C.N.R. Rao, J. Chem. Soc.
Chem. Commun. 7 (1985).
14] O.H. Hansteen, H. Fjelvag, B.C. Hauback, J. Mater. Chem. 8 (1998) 2081.
15] C. Gommes, S. Blacher, Ch. Bossuot, P. Marchot, J.B. Nagy, J.P. Pirard,
Carbon 42 (2004) 1473.
(
2005) 454.
[
[
4] M.M.J. Treacy, T.W. Ebbesen, J.M. Gibson, Nature 381 (1996) 678.
5] V. Ivanov, A. Fonseca, J.B. Nagy, A. Lucas, P. Lambin, D. Bernaerts, X.B.
Zhang, Carbon 33 (1995) 1727.
6] T.E. M u¨ ller, D.G. Reid, W.K. Hsu, J.P. Hare, H.W. Kroto, D.R.M. Walton,
Carbon 35 (1997) 951.
[
[
[
[
7] B.C. Liu, L.Z. Gao, Q. Liang, S.H. Tang, M.Z. Qu, Z.L. Yu, Catal. Lett. 71
[
16] I. Alstrup, J. Catal. 109 (1988) 241.
(
2001) 225.