Nickel antigorite synthesis and carbon tubular

nanostructures formation on antigorite-based nickel

particles by acetylene decomposition

P. Leroi a, F.J. Cadete Santos Aires a,*, T. Goislard de Monsabert b,H. Le Poche b, J. Dijon b, J.L. Rousset a, J.C. Bertolini a

a Institut de Recherches sur la Catalyse (UPR 5401—CNRS), 2, Avenue Albert Einstein, 69626 Villeurbanne Cedex, FrancebCEA Grenoble—LETI-DOPT/STCO/LTCV, 17, Rue des Martyrs, 38054 Grenoble Cedex 9, France

Received 24 January 2005; received in revised form 20 June 2005; accepted 20 June 2005

Abstract

We have synthesized nickel antigorite in order to obtain, by subsequent heat treatment and reduction, supported nickel particles more or

less anchored in the antigorite structure and rather well oriented depending on the temperature. The sample prepared at low temperature

(450 8C) lead to mainly anchored Ni particles on rather well structured antigorite. For the samples prepared at intermediate temperatures (530,

600 8C) increasing destructuring of antigorite with temperature is observed. Finally, for the sample prepared at the highest temperature

(700 8C) weakly bound Ni particles on a support resembling amorphous SiO2 are mainly obtained.

Acetylene decomposition performed around 600 8C on these samples yield small amounts of carbon tubular nanostructures (nanotubes and

nanofilaments). However, we have found that, whatever the temperature, carbon tubular nanostructures growth (tip growth mechanism) is

favoured on the destructured regions of the material where the Ni particles are weakly bound to the support. Higher density of tubular

nanostructures is consequently observed for the higher temperature samples. We attributed this behaviour to the fact that the strongly bound

anchored Ni particles mainly exhibit (1 1 1)-type faces, which are known to promote carbon precipitation. This would lead to coking when

particles are not able to rearrange (as it is the case for the anchored particles) whereas carbon tubular nanostructures may grow on weakly

bound particles that are able to rearrange.

# 2005 Elsevier B.V. All rights reserved.

Keywords: Nickel antigorite; Carbon nanotubes and nanofilaments growth; Acetylene decomposition

www.elsevier.com/locate/apcata

Applied Catalysis A: General 294 (2005) 131–140

1. Introduction

Hydrocarbon decomposition on nickel catalysts is known

to generate carbonaceous deposits. Depending upon the

carburizing mixture and the synthesis temperature, carbon

structures of different nature and shapes can be obtained:

deposited graphite, amorphous carbon, graphitic carbon

nanotubes (CNT) and nanofibers (GNF) [1]. Amorphous

carbon and graphitic adlayers are often responsible for

catalyst deactivation in high-temperature reactions implying

molecules containing carbon atoms, such as hydrocarbons or

CO. Nevertheless, CNTs and GNFs may be of great interest

* Corresponding author. Tel.: +33 4 72 44 53 03; fax: +33 4 72 44 53 99.

E-mail address: aires@catalyse.cnrs.fr (F.J. Cadete Santos Aires).

0926-860X/$ – see front matter # 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.apcata.2005.06.006

in a wide range of applications. For instance, they can be

used as catalysts supports thanks to their extremely high

surface area and chemical stability. Moreover, CNTs have

exceptional mechanical properties that makes them suitable

for reinforcement of composite materials. They have also

many potential applications for nanotechnologies in

electronic devices; for instance, their aspect ratio make

them very good candidates for electron field emitters [2–5].

Two mechanisms have been proposed for the carbon

tubular nanostructures (nanotubes and nanofilaments)

growth on transition metals [6]. The most commonly

observed is the so-called ‘‘tip growth’’. The growth of CNTs

and GNFs would then proceed via a carbon dissolution step

followed by an expulsion of carbon atoms generating tubes

with nickel particles located at the end tip of these tubes

P. Leroi et al. / Applied Catalysis A: General 294 (2005) 131–140132

[7–10]. One can think that the graphite-covered nickel

particles are unstable for particles having a large curvature

and therefore transforms rapidly into carbon tubular

nanostructures [11]. In the less frequently observed ‘‘root

growth’’ mechanism the metal particle remains anchored on

the support [7,12]. The shape of the graphitic layers will thus

depend on the metal particle size [11,13], its structure

[11,14] and/or morphology [15] and the metal-support

interactions [16].

Nickel catalysts prepared by reduction of nickel

antigorite as precursor seems to be a promising way to

produce nickel particles strongly anchored on the support

and exhibiting preferential (1 1 1), and to a less extent

(1 1 0), surface planes [17–19]. Therefore, it could be a way

to obtain metal particles having original properties able to

grow carbon overlayers with peculiar structures.

Among the hydrocarbon molecules, acetylene is by far

the most reactive with Ni catalysts, allowing the possible

decomposition at relatively low temperatures. Furthermore,

the presence of a reducing agent like hydrogen or ammonia

[20] is known to influence the graphite deposit and hence the

nature of the CNTs generated.

In this paper, we first report on the preparation and

characterisation by X-ray diffraction (XRD) and transmis-

sion electron microscopy (TEM), of nickel antigorite

synthesized from nickel hydroxide and silica, and of the

nickel particles obtained after its calcinations–reduction at

different temperatures. Then, the nature of carbon deposits

obtained by CVD decomposition of acetylene in different

experimental conditions on these nickel catalysts is

analysed.

2. Experimental

2.1. Material and catalyst

Nickel antigorite Ni3(OH)4Si2O5 was prepared by

hydrothermal treatments of stoichiometric mixtures of silica

(0.63 g, Aerosil 200 Degussa) and nickel hydroxide (1.53 g,

Ni(OH)2, Ni 61%, Alfa Aesar) in deionised water (25 ml).

The mixture was placed in a stainless steel autoclave of

about 350 ml volume. The autoclave was then kept in an

oven for 120 h at temperatures of 200 and 250 8C. Furtherincrease of the temperature (up to 350 8C) was obtained by

covering the autoclave with a heating ribbon. The collected

sample was further dried at 120 8C.In order to obtain the final catalyst containing Ni

supported particles, the nickel antigorite was then heated

at the desired temperature T under vacuum (heating rate:

1 8C/min) during 8 h. After cooling, the as-obtained

precursor was reduced during 16 h under 250 mbar of

hydrogen at the same temperature T, the heating rate being

limited to 1 8C/min. Finally, after cooling at RT the

reduced catalyst was held under argon flow for 0.5 h,

removed and stored in air.

2.2. Acetylene decomposition

The experiments were carried out in two separate

reactors. A fixed-bed flow-type quartz reactor designed to

work at atmospheric pressure was used at IRC. A

PLASSYS1 hot walls vertical CVD reactor designed to

work at low pressure was used at LETI.

At IRC, the growth process consisted in reducing the

catalyst in situ at 500 8C for 1 h in an hydrogen flow

(50 ml min�1) followed by an increase in temperature up to

600 8C under nitrogen/hydrogen (95/50 ml min�1) flow

and finally introducing C2H2 to obtain a C2H2/H2 (1/10)

mixture (total flow of 150 ml min�1 measured at room

temperature and atmospheric pressure) for a short time

(2 min). The experiment was completed by cooling the

material from the synthesis temperature to room tempera-

ture under neutral gas (N2). The choice for using in our

study short acetylene decomposition times is based on

previous works [21,22] and where it was observed that after

a very short period tubular carbon growth stops. Namely,

Baker et al. [22] have observed that for decomposition of

acetylene at 600 8C on Ni (similar to our working

conditions) growth stops after 12–90 s for Ni particles

treated in argon and hydrogen, respectively.

At LETI, the growth process consisted in pumping the

chamber down to 10�6 mbar, reducing the catalyst in situ at

630 8C for 30 min under 0.2 mbar and 100 ml min�1 of

hydrogen, stopping the hydrogen supply and pumping down

for 1 min, introducing acetylene for 2 min at 0.2 mbar and

80 ml min�1 and finally cooling to RT under vacuum.

These two processes will be referred to as process A and

process B, respectively, throughout the manuscript.

2.3. Characterisation techniques

The structure of the samples obtained at the various steps

of the preparation was checked by X-ray diffraction (XRD)

using a Brucker D5005 powder goniometer (type u–u),

where the sample is fixed and the X-ray tube (Cu

Ka = 0.1541 nm) and the detector rotate. The diffracto-

grams were acquired in a 2u range of 3–808 with steps of

0.0208. The acquisition time was 1 s step�1. The nature of

the different phases constituting the sample was identified

using the database of the Joint Committee on Powder

Diffraction Standards (JCPDS).

Transmission electron microscopy (TEM) characterisa-

tion was carried out with a JEOL JEM 2010 microscope

working at 200 kV and equipped with a LaB6 tip, a high-

resolution pole piece and a PentafetLink ISIS EDS-X

spectrometer (Oxford Instrument). The samples were

dispersed in ethanol in an ultrasonic bath before deposition

on a holey carbon-coated copper grid for TEM examination.

The samples after acetylene decomposition were only

dispersed in ethanol and not ultrasonicated so that possible

damage to the carbon tubes could be avoided.

P. Leroi et al. / Applied Catalysis A: General 294 (2005) 131–140 133

3. Results and discussion

3.1. Nickel antigorite preparation

The XRD patterns of samples prepared by hydrothermal

treatment at temperatures ranging from 200 to 350 8C are

presented in Fig. 1. The samples treated at 300 and 350 8Cshowed the presence of two crystalline phases, nickel

antigorite and nickel oxide, whereas data given in the

literature [17–19,23] indicate the unique presence of nickel

antigorite phase in these conditions. The presence of nickel

oxide seems to indicate that the temperature of synthesis is

too high, producing some oxidation of the nickel

hydroxide. After preparation at a lower temperature

(250 8C) the XRD pattern of the product obtained exhibited

diffraction lines which can be assigned to the presence of

only antigorite and nickel hydroxide phases. The XRD

pattern of products synthesized at 200 8C indicated the

presence of two phases, as for synthesis at 250 8C.Nevertheless, the nickel antigorite was less perfectly

crystallized (absence of (0 0 1) and (0 0 2) reflections at

2u = 128 and 248, respectively) and the intensity of

diffraction peaks associated to nickel hydroxide increased.

The TEM images of nickel antigorite synthesized at

250 8C (Fig. 2) indicated the presence of layers of several

hundred nanometres. These layers were either rolled around

each other (white arrows in Fig. 2) or as folded sheets (black

arrows in Fig. 2). In the images (Fig. 2b), the distance

between two planes equals 0.723 nm, corresponding to the

distance along the c axis as reported for the nickel antigorite

structure in the JCPDS file (no. 20-0791). This structure is

Fig. 1. XRD pattern of nickel antigorite synthesized at 200, 250, 300 and 3

close to that observed by Martin et al. [23] whereas the

nickel antigorite observed by Maubert [18] was essentially

constituted by platelets.

The antigorite synthesized by hydrothermal treatment at

250 8C was thus retained as the base material used for the

preparation of nickel catalysts.

3.2. Thermal treatments and preparation of nickel

catalysts

From previous works [18], it appears clearly, that

calcination prior to reduction is essential in order to obtain

well-structured Ni particles anchored in the antigorite since

water elimination, more or less fast during the reduction

process, could influence the nickel orientation. We have thus

performed calcination and reduction treatments for each

temperature.

The XRD patterns of the products obtained after

calcination and reduction at 700 8C are given in Fig. 3.

After calcination one can see the disappearance of the

diffraction lines characteristic of nickel antigorite at

2u = 128 and 248, indicating the destruction of the lamellar

structure. A shift of the 2u values to higher values (dashed

lines in Fig. 3) showed a decrease of the cell parameter. The

nickel hydroxide phase has disappeared and nickel oxide is

observed. This result differs from those reported by

Maubert [18] where important changes in morphology or

structure were observed only when antigorite was heated

above 700 8C. Nevertheless, this decomposition was

reported by Martin et al. [17] to begin above 420 8C. We

must assume that the experimental conditions were not

50 8C: (*) nickel antigorite; (+) nickel hydroxide; (*) nickel oxide.

P. Leroi et al. / Applied Catalysis A: General 294 (2005) 131–140134

Fig. 2. TEM images of nickel antigorite 250 8C used for the preparation of nickel catalysts: (a) large zone showing the rolled layers (white arrow) and folded

sheets (black arrow) of antigorite. (b) Detailed view of the rolled layers (white arrow) and folded sheets (black arrow) of antigorite.

identical in those two works. Our approach seems to be

consistent with the one of Martin et al. [23]. After

reduction (Fig. 3c) only metallic nickel can be evidenced

by XRD. The EDX analysis in the TEM showed the

presence of Ni and SiO2. The TEM images (Fig. 4a) show

essentially the presence of nickel particles with sizes

ranging from 5 to 20 nm beside SiO2 grains. Some very

rare NiO particles were also observed as well as some very

reduced regions where a few antigorite layers prevailed

(Fig. 4b).

In order to avoid a complete destruction of the structure

of the support, thermal treatments were realized at lower

temperatures: 600, 530 and 450 8C.

The XRD patterns of calcinated products (Fig. 5a–c)

showed that (0 0 1) and (0 0 2) reflections related to

antigorite are observed after treatment at 450 8C but not

after treatment at 530 and 600 8C, indicating the

progressive destructuring of the layered structure with

temperature for the latter values. Nickel oxide was

obtained, but with XRD beam intensities lower than after

calcination at 700 8C. After reduction (Fig. 5d–f), metallic

Ni is observed, whose intensity increases with tempera-

ture. In contrast with the results obtained after heating at

700 8C, the XRD plots present peaks associated to

antigorite, whose intensity decreases when the temperature

increases.

P. Leroi et al. / Applied Catalysis A: General 294 (2005) 131–140 135

Fig. 3. XRD patterns of nickel antigorite (a), after calcination at 700 8C (b)

and after calcination and reduction at 700 8C (c). (*) nickel antigorite; (+)

nickel hydroxide; (*) nickel oxide; (~) nickel.

TEM images (Fig. 6) showed gradual decomposition of

antigorite into nickel and silica with temperature increase.

After treatment at 450 8C (Fig. 6a), the sample is

homogeneous with Ni particles of average size of 5 nm

deposited on a support with a layered antigorite structure.

After treatment at 530 8C (Fig. 6b) and 600 8C (Fig. 6c) the

sample was more heterogeneous. Decomposition into nickel

and silica was not as pronounced as at 700 8C, and many

regions of the sample retained the antigorite structure. The

size of the nickel particles ranged then from 5 to 20 nm, but

with some bigger Ni particles (up to 40 nm) after

calcinations–reduction at 600 8C.In conclusion, in the calcinations–reduction temperature

range of 530–600 8C the antigorite precursor was more or

less decomposed with some nickel particles anchored on the

layered support and some on silica. The layered structure of

antigorite with Ni nanoparticles prevailed at 450 8C, whileheating at 700 8C generated only Ni particles beside

amorphous SiO2.

3.3. Acetylene decomposition

Acetylene decomposition was performed both in the flow

reactor working at atmospheric pressure (process A) for the

Fig. 4. TEM images of the calcined and reduced antigorite at 700 8C showing: (

(SiO2) and (b) image of a very rare region where some antigorite layers remain

catalysts thermally treated at 450, 530, 600 and 700 8C (they

will be noted A450-A, A530-A, A600-A and A700-A), and

in the low pressure CVD reactor (process B) on the samples

issued from thermal treatments at 450 and 530 8C (they will

be noted A450-B and A530-B).

We performed analysis of the total carbon formed by

acetylene decomposition on the above-mentioned samples.

For the acetylene decomposition by process A it appears that

the carbon yield is quite similar (10 � 2%) for all samples

(A450-A, A530-A, A600-A and A700-A). In the case of

acetylene decomposition by process B, the yield is different

for A A450-B (4%) and A530-B (8%) indicating that the low

pressure CVD process is more sensitive to the structural

changes of the antigorite.

The TEM pictures of the tubular nanostructures obtained

on the catalysts A450-A, A530-A and A450-B, A530-B after

acetylene decomposition at 600 8C are given in Figs. 7 and 8,

respectively. In both experimental conditions, despite the

decomposition temperature (about 600 8C), the catalysts

kept the same external appearance after acetylene decom-

position. Only few carbon tubular nanostructures are

observed despite a multitude of nickel particles on the

layered structure. These carbon tubes grow on some spots of

the periphery of the grains. Only rarely Ni particles are

observed at the end of the tubular nanostructures.

Furthermore, coking of the supported particles was observed

together with the formation of the tubular nanostructures.

Nevertheless, no tube growth could be unquestionably

attributed to particles anchored on the antigorite surface.

However, some differences were noticed: while for both

processes graphene planes are rather parallel to the growth

direction CNTs grown during process A are rather short

(several tens of nanometer) not straight and poorly

graphitised (Fig. 7) whereas CNTs grown during process

B seem to be longer and straight and to have a better

definition with graphene planes quite parallel to the tube axis

(Fig. 8). The reaction conditions (under atmospheric

pressure in process A and under low pressure in process

B) and more specifically the presence of hydrogen in process

A can be tentatively invoked to explain these differences.

Indeed, the growth of carbon tubular nanostructures is

a) collection of Ni particles (�5 nm) on a completely destructured support

(arrows).

P. Leroi et al. / Applied Catalysis A: General 294 (2005) 131–140136

Fig. 6. TEM images of the calcined and reduced antigorite at 450 8C (a),

530 8C (b) and 600 8C (c) showing a progressive destructuring of the

antigorite with increasing temperature.

Fig. 5. XRD patterns of the samples of nickel antigorite after calcination at

450, 530, 600 8C (a–c) and after calcinations–reduction at 450, 530, 600 8C(d–f). The dashed lines are characteristic of the nickel antigorite phase; (*)

nickel oxide; (~) nickel.

strongly affected by well/ill-match between the decomposi-

tion rate of the feed gas (i.e. the rate of carbon formation)

and the transfer/diffusion rate of the carbon formed [24] and

by the presence of hydrogen which is known to influence the

graphite deposit and hence the nature of the CNTs generated

[20].

More considerable amounts of CNTs were present on

the catalysts prepared at higher temperatures, A600-A

(Fig. 9) and A700-A (Fig. 10), especially for the latter.

They are often much longer (several hundreds of nanometer

to micrometer) than those previously observed on A450-B

or A530-A or A530-B. Ni particles at the end as well as

within the tubular nanostructures are almost systematically

observed. The tube growth occurs merely in the areas

where the layered structure does not exist, i.e. on areas

looking like nickel on silica, where the nickel particles are

less strongly anchored on the support. This is consistent

with a larger yield of carbon tubular nanostructures for the

A700-A than for the A600-A. Such a behaviour is a priori

surprising since acetylene decomposition, under similar

conditions performed, on a Ni/SiO2 catalyst (prepared by

conventional impregnation of nickel nitrate on silica

(Aerosil 200 Degussa) followed by calcination and

reduction yielding particles with sizes ranging from 3 to

15 nm) showed essentially graphitic carbon on the surface.

In this case, only very few carbon tubular nanostructures

(lengths up to hundred nanometers and diameters of around

10–30 nm), were observed together with the coking of the

catalyst. Probably, destructuring of the antigorite at high

temperatures (especially at 700 8C) creates a somehow

specific sample where Ni particles are located beside silica.

They have thus a very weak interaction with the support

compared to particles on Ni/SiO2 catalyst prepared by

impregnation. Indeed, for the latter their nucleation and

growth is regulated by the dangling bonds at the surface

of silica and so the nickel particles interact with the

support. In the case of the particles obtained through the

calcinations–reduction method of nickel antigorite, one can

reasonably assume that the rather large structured Ni

particles will loose their strong interaction with the support

since the interface is ‘‘broken’’ during the destructuring at

high temperature. This is consistent with: (i) the fact that

strongly interacting particles obtained at lower tempera-

tures (450 and 530 8C) yield very few tubular nanos-

tructures and (ii) the increasing yield of CNTs yield with

temperature (600 and 700 8C) and the consequent loss of

ordering within the support which was at the origin of the

close interaction of the nickel particles with it.

Furthermore, as one can see (Fig. 6) the density of

particles on the structured regions (which compose most

P. Leroi et al. / Applied Catalysis A: General 294 (2005) 131–140 137

Fig. 7. TEM images of the tubular nanostructures observed on A450-A (a) and A530-A (b–d). (c and d) detailed region of the end of a nanotube and of a coked

particle, respectively.

Fig. 8. TEM images of the nanotubes observed for A450-B (a and b) and A530-B (c and d).

P. Leroi et al. / Applied Catalysis A: General 294 (2005) 131–140138

Fig. 9. TEM images of the tubular nanostructures observed on A600-A: (a) large zone with few long nanotubes; (b) detail of the beginning of a nanofilament on

the support surface; (c) detail of the end of the same nanofilament capped with a nickel particle.

of the lower temperature samples) is rather high and thus

there is not very much free space in-between such particles

(as opposed to the destructured regions where the particles

are rather well separated). For tubular nanostructures formed

on higher temperature samples, we can notice (Fig. 9b) that

their base on the surface (were it began to nucleate) is

somewhat larger than its diameter and thus larger than the

particle that generated the nanotube (Fig. 9c). This indicates

that a successful formation of the nanotube might require

some extra space around the particle rather than just the

space occupied by the particle on the support. We thus

suppose that, together with the above-mentioned fact that

the particles are in strong interaction with the surface, the

high density of particles might be another reason for the

poor efficiency of the samples prepared at lower tempera-

ture for the formation of nanotubes, at least within our

working conditions.

It seems that no (or few) carbon production by C2H2

decomposition occurs on the anchored nickel particles.

Different authors [25–28] have proposed or reported a

mechanism for filament formation involving the following

steps: (i) adsorption and dissociation of gas molecules on

specific faces of nickel (these faces might be Ni(1 0 0) and/

or Ni(1 1 0)); (ii) dissolution and diffusion of carbon through

the metal particle; (iii) graphite precipitation at the specific

rear faces of the nickel particles to form CNTs. The

favourable rear faces might be Ni(1 1 1) planes and vicinal

Ni faces such as (2 1 1) and (3 1 1). On the nickel anchored

on layered support, even if they were not systematically

detected, the extended exposed faces are most probably

(1 1 1) type faces (Fig. 11) which favour carbon precipita-

tion. However, since (1 0 0) and (1 1 0) are less extended and

so less available, the growth quickly stops due to coking of

the particle that remains anchored on the layered support.

Furthermore, we did not observe the formation of ‘‘root-

grown’’ single-wall nanotubes on particular regions of the

‘‘extended’’ exposed faces described in some recent works

[29,30]. On the contrary, for the weakly bound Ni particles

(higher temperature samples) they are free to rearrange in

order to favour the growth of tubular nanostructures which is

consistent with elongated particles observed within the

tubular nanostructures (Fig. 10b and c) and with recent

P. Leroi et al. / Applied Catalysis A: General 294 (2005) 131–140 139

Fig. 10. TEM images of the tubular nanostructures observed on A700-A:

(a) large zone with many long nanotubes and nanofilaments; (b) elongated

particle at the end of a nanotube; (c) elongated particle inside a nanotube.

Fig. 11. TEM of a Ni particle anchored in the antigorite for the calcined–

reduced sample at 450 8C showing developed (1 1 1) and less developed

(1 0 0) and (1 1 0) type exposed planes (bold, dashed and dotted lines,

respectively). We can also note an exposed (2 1 1) type plane (grey line).

in-situ observations of carbon nanotube growth on Ni

particles were the metallic particles rearrange in a pulsating

mode during growth of the nanotube [31].

On Ni/SiO2 catalyst prepared by classical chemical

methods (impregnation of a silica powder by decomposition

of a nickel precursor in an aqueous solution), the

hydrocarbon must adsorb and decompose on Ni particles

but the carbon atoms seem to accumulate on the initially

exposed site of the catalyst particle and form graphene layers

which encapsulate the nickel particles. The rate of diffusion

could be much weaker than the carbon production rate by

C2H2 decomposition causing this encapsulation.

4. Conclusion

We have prepared nickel antigorite Ni3(OH)4SiO2 by

hydrothermal treatment at 250 8C of stoichiometric mixtures

of silica and nickel hydroxide in deionized water. The

morphology of nickel antigorite obtained was essentially

constituted by rolled and folded plate layers of several

hundred nanometres and used as the base material for the

preparation of nickel catalysts.

The nickel catalysts were obtained by thermal treatments

at 450, 530, 600 and 700 8C of the nickel antigorite which

yielded different samples consisting in more or less

destructured antigorite with nickel particles anchored on

the layered support as well as some on silica. The layered

structure of antigorite with Ni nanoparticles was rather well

conserved after a heating at 450 8C, while at 700 8Cdestructuring of the base material yielded Ni particles beside

amorphous SiO2.

Acetylene decomposition was performed on these

samples in a flow reactor at 600 8C at atmospheric pressure

in presence of hydrogen and in a low pressure CVD reactor

at 630 8C. Carbon tubular nanostructures growth is seldom

observed on the different samples and it only occurs in the

areas where the nickel antigorite layered structure does not

exist. As a consequence, the carbon tubular nanostructures

density increases for the higher temperature samples were

tip growth mechanism seems to occur. A root growth

mechanism on the nickel particles anchored on the support

can be excluded. Indeed, such particles do not seem to be

favourable for carbon tubular nanostructures growth. One

may attribute such a behaviour mainly to the fact that

anchored particles seem to have essentially (1 1 1) type

exposed faces which promote carbon precipitation and

consequent coking if the particles are not allowed to

rearrange (since they are anchored on the support).

P. Leroi et al. / Applied Catalysis A: General 294 (2005) 131–140140

Conversely, tubular nanostructures growth happens on

particles which interact weakly with the support. To a less

extent the high density of particles on the structured material

seems to be noxious to the development of carbon tubular

nanostructures.

References

[1] K.P. De Jong, J.W. Geus, Catal. Rev. Sci. Ing. 42 (2000) 481.

[2] J.M. Bonard, M. Croci, C. Klinke, R. Kurt, O. Noury, N. Weiss,

Carbon 40 (2002) 1715.

[3] G.N. Fursey, D.V. Novikov, G.A. Dyuzhev, A.V. Kotcheryzhenkov,

P.O. Vassiliev, Appl. Surf. Sci. 215 (2003) 135.

[4] J. Dijon, C. Bridoux, A. Fournier, F. Geffraye, T. Goislard De

Monsabert, B. Montmayeul, M. Levis, D. Sarrasin, R. Meyer, K.A.

Dean, B.F. Coll, S.V. Johnson, C. Hagen, J.E. Jaskie, J. Soc. Inform.

Display 12 (2004) 373.

[5] J. Dijon, A. Fournier, T. Goislard De Monsabert, B. Montmayeul, D.

Zanghi, AIP Conference Proceedings 685, Molecular Nanostructures,

XVII International Winterschool/Euroconference on Electronic Prop-

erties of Novel Materials, Kirshberg, Austria, 2003, p. 592.

[6] R.T.K. Baker, Carbon 27 (1989) 315.

[7] R.T.K. Baker, P.S. Harris, Formation of Filamentous Carbon in

Chemistry and Physics of Carbon, vol. 14, Marcel Dekker, 1978.

[8] P. Pinheiro, M.C. Schouler, P. Gadelle, M. Mermoux, E. Dooryhee,

Carbon 38 (2000) 1469.

[9] V.C.H. Kroll, H.M. Swaan, C. Mirodatos, J. Catal. 161 (1996) 409.

[10] J.W. Snoeck, G.F. Froment, M. Fowles, J. Catal. 169 (1997) 240.

[11] M. Yudasaka, R. Kikuchi, Y. Ohki, E. Ota, S. Yoshimura, Appl. Phys.

Lett. 70 (1997) 1817.

[12] S. Fan, M.G. Chapline, N.R. Franklin, T.W. Tombler, A.M. Cassell,

H.J. Dai, Science 283 (1999) 512.

[13] E.F. Kukovitsky, S.G. L’vov, N.A. Sainov, V.A. Shustov, L.A. Cher-

nozatonskii, Chem. Phys. Lett. 355 (2002) 497.

[14] M.H. Kuang, Z.L. Wang, X.D. Bai, J.D. Guo, E.G. Wang, Appl. Phys.

Lett. 76 (2000) 1225.

[15] N.M. Rodriguez, A. Chambers, R.T.K. Baker, Langmuir 11 (1995)

3862.

[16] C. Park, M.A. Keane, J. Catal. 221 (2004) 386.

[17] G.A. Martin, B. Imelik, M. Prettre, C. R. Acad. Sci. Paris 264 (1967)

1536.

[18] A. Maubert, Ph.D. Thesis No. 262, UCB-Lyon 1, 1973.

[19] G.A. Martin, R. Dutartre, J.A. Dalmon, in: J. Bourdon (Ed.), Growth

and Properties of Metal Clusters, Elsevier, 1980, pp. 467–478.

[20] A.C. Chhowalla, K.B.K. Teo, C. Ducati, N.L. Rupesinghe, G.A.

Amaratunga, A.C. Ferrari, D. Roy, J. Robertson, W.I. Milne, J. Appl.

Phys. 90 (2001) 5308.

[21] Z.Y. Juang, I.P. Chien, J.F. Liai, T.S. Lai, C.H. Tsai Uang, et al.

Diamond Relat. Mater. 13 (2004) 1203.

[22] R.T.K. Baker, M.A. Barber, P.S. Harris, F.S. Feates, J.R. Waite, J.

Catal. 26 (1972) 51.

[23] G.A. Martin, A. Renouprez, G. Dalmai-Imelik, B. Imelik, J. Chim.

Phys. 67 (1970) 1149.

[24] P. Chen, H.B. Zhang, G.D. Lin, Q. Hong, K.R. Tsai, Carbon 35 (1997)

1495.

[25] P.E. Anderson, N.M. Rodriguez, J. Mater. Res. 14 (1999) 2912.

[26] R.T. Yang, J.P. Chen, J. Catal. 115 (1989) 52.

[27] R.T.K. Baker, N.M. Rodriguez, Mater. Res. Soc. Symp. Proc. 349

(1994) 251.

[28] J.L. Figueiredo, C.A. Bernardo, J.J. Chludzinski Jr., R.T.K. Baker, J.

Catal. 110 (1988) 127.

[29] J. Gavillet, A. Loiseau, C. Journet, F. Willaime, F. Ducastelle, J.-C.

Charlier, Phys. Rev. Lett. 87 (2001) 275504.

[30] A. Maiti, C.J. Brabec, J. Bernholc, Phys. Rev. B 55 (1997) R6097.

[31] S. Helveg, C. Lopez-Cartes, J. Sehested, P.L. Hansen, B.S. Clausen,

J.R. Rostrup-Nielsen, F. Abild-Pedersen, J.K. Nørskov, Nature 427

(2004) 426.