Preparation of aligned carbon nanotubes catalysed by laser-etched cobalt thin films

Article abstract of DOI:10.1016/S0009-2614(98)00093-1

Pyrolysis of 2-amino-4,6-dichloro-s-triazine over laser-etched thin films of cobalt deposited on an inverted silica substrate generates aligned bundles and films of carbon nanotubes of uniform dimensions. Scanning electron microscopy, high resolution transmission electron microscopy, energy dispersive X-ray analyses, electron energy loss spectroscopy and electron diffraction studies reveal that the aligned tubes, which usually grow perpendicular to the substrate surface, are mainly straight (length ≤100 μm; 30-50 nm OD), contain traces of nitrogen (<5%) and exhibit a preferred helicity (e.g. armchair 25-30%). Other graphitic structures (e.g. polyhedral particles, encapsulated particles and amorphous carbon) are absent. Energised cobalt clusters/particles (≤50 nm), ablated during laser-etching, may condense and recrystallise evenly on the surface of the substrate as cobalt and/or cobalt oxide. These crystals apparently play a key role in nanotube production.

Full text of DOI:10.1016/S0009-2614(98)00093-1

2
7 March 1998
Chemical Physics Letters 285 Ž1998. 299–305
Preparation of aligned carbon nanotubes catalysed by laser-etched
cobalt thin films
a,b
a
b
c
a
a
M. Terrones , N. Grobert , J.P. Zhang , H. Terrones , J. Olivares , W.K. Hsu ,
a
b
a
a
J.P. Hare , A.K. Cheetham , H.W. Kroto , D.R.M. Walton
a
School of Chemistry, Physics and EnÕironmental Science, UniÕersity of Sussex, Brighton BN1 9QJ, UK
b
Materials Research Laboratory, UniÕersity of California, Santa Barbara, CA 93106, USA
c
Instituto de F ´ı sica, UNAM, Apartado Postal 20-364, M e´ xico D.F. 01000, M e´ xico
Received 18 December 1997; in final form 21 January 1998
Abstract
Pyrolysis of 2-amino-4,6-dichloro-s-triazine over laser-etched thin films of cobalt deposited on an inverted silica substrate
generates aligned bundles and films of carbon nanotubes of uniform dimensions. Scanning electron microscopy, high
resolution transmission electron microscopy, energy dispersive X-ray analyses, electron energy loss spectroscopy and
electron diffraction studies reveal that the aligned tubes, which usually grow perpendicular to the substrate surface, are
mainly straight Žlength (100 mm; 30–50 nm OD., contain traces of nitrogen Ž-5%. and exhibit a preferred helicity Že.g.
armchair 25–30%.. Other graphitic structures Že.g. polyhedral particles, encapsulated particles and amorphous carbon. are
absent. Energised cobalt clustersrparticles Ž(50 nm., ablated during laser-etching, may condense and recrystallise evenly
on the surface of the substrate as cobalt andror cobalt oxide. These crystals apparently play a key role in nanotube
production. q 1998 Elsevier Science B.V.
1
. Introduction
lus measurements also show that multi-layered nan-
otubes are mechanically much stronger than conven-
tional carbon fibres w17,18x and are extraordinarily
flexible when subjected to substantial strain w19x.
Most methods used, to date, to generate nan-
otubes suffer from the drawback that considerable
quantities Ž(40%. of unwanted particles are also
formed, although such particles can be eliminated by
oxidation w20x. Unfortunately more than 95% of the
nanotubes are also destroyed in this process and, to
make matters worse, the remaining nanotubes lose
their protective end-caps. As a result, further high
temperature annealing Že.g. 3000 K. is necessary in
Soon after the identification of carbon nanotubes
w1x in soot residues generated in the Kratschmerr
Huffman fullerene reactor, bulk quantities were suc-
cessfully prepared w2x. Nanotubes can now be pro-
¨
duced from graphite by various techniques, including
arc discharge w2–4x, laser vaporisation w5–7x and
electrolysis w8,9x and by hydrocarbon pyrolysis over
catalysts w10–12x.
Recent measurements on nanotubes w13,14x, indi-
vidual multi-layered tubes w15,16x and ropes of sin-
gle-walled tubules w6x have revealed that their con-
ducting properties depend markedly on the degree of
graphitisation, helicity and diameter. Young’s modu-
order to eliminate dangling bonds located at the
edges of the open nanotubes. Another general prob-
0
009-2614r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.
PII S0009-2614Ž98.00093-1

3
00
M. Terrones et al.rChemical Physics Letters 285 (1998) 299–305
lem arises in that the dimensions of the nanotubes
are highly variable. This Letter describes the produc-
tion of aligned carbon nanotubes in high yield Žun-
contaminated by particles. of controlled length Ž(
1
00 mm. and of fairly uniform external diameter
Ž30–50 nm. by pyrolysis of 2-amino-4,6-dichloro-s-
triazine over cobalt thin films, deposited on an in-
verted silica substrate, and etched using laser tech-
niques. This method fosters the generation of small
cobaltrcobalt oxide clusters of uniform size, which
encourages the formation of the aligned tubes, cru-
cial if the mechanical properties of the nanotubes as
advanced composite materials are to be harnessed.
2
. Experimental
2
.1. Catalyst substrate.
A thin film of cobalt Žf10–100 nm. was de-
posited on a silica plate Ž1 mm thick, 5 mm wide and
2
0 mm long. using laser ablation w21x andror subli-
mation techniques. The plate was then exposed to air
and etched with single laser pulses ŽNd:YAG 266
nm; 5 mJrpulse. using cylindrical lenses Ž65 mm
focal length; Fig. 1a., thus creating linear tracks
Žwidth 1–20 mm; length (5 mm; Fig. 1b. where
Fig. 1. Ža. Etching technique using a Nd:YAG 266 nm laser and
cylindrical lens Ž65 mm focal length.. Žb. SEM image from an
etched nanotrack showing how uniform the channel is made. It is
believed that along this channelrtrack uniform Co nanoparticles
are deposited evenly by the laser striking the thin film.
the cobalt had been ablated.
2
.2. Triazine preparation and pyrolysis
-Amino-4,6-dichloro-s-triazine was prepared by
Using a heated funnel, the solution was quickly
2
filtered to remove precipitated NH Cl and then set
4
modifying an existing procedure w22x.
aside to cool. Dioxane was removed under reduced
pressure and the solid residue, 2-amino-4,6-dichloro-
s-triazine, was recrystallised from hot water to con-
stant mp 2358C dec ŽRef. w22x, 2358C., after remov-
1
ing final traces of water under high vacuum. H-NMR
q
ŽCDCl . d 5.88 Žs, 2H.; MS Žmre. 164 ŽM ..
3
2
-Amino-4,6-dichloro-s-triazine Ž0.02–0.15 g.
was introduced into one end of a silica tube Ž6 mm
ID: length 60 cm., and the silica substrate coated
with etched cobalt was placed facing downwards at
the other end. The tube was then positioned in a two
furnace system fitted with temperature controllers
under an Ar flow Ž40 cm rmin.. Details of this
pyrolytic process carried out at 9508C have been
described elsewhere w21,23x.
A solution of NH Ž0.05 mol. in 1,4-dioxane Ž100
3
3
cm ; Aldrich., was added dropwise to ice-cooled
3
cyanuric chloride Ž4.6 g, 0.025 mol.. The mixture
was allowed to attain room temperature during 12 h
Žmagnetic. stirring and was then heated to 958C.

M. Terrones et al.rChemical Physics Letters 285 (1998) 299–305
301
Following pyrolysis, the silica plate was covered
with dark tracks observable to the naked eye; those
areas where the cobalt film had not been laser-etched
were transparent in most cases Žsee Section 2.1..
at 100 keV., high-resolution transmission electron
microscopy ŽHRTEM; JEOL JEM 2010 at 200 keV
and JEM 4000 at 400 keV., electron energy-loss
spectroscopy ŽEELS; beam size 3 nm, 0.3 eV for
Plates were analysed directly by scanning electron
microscopy ŽSEM; Leo 5420 operated at 5–20 keV..
The black deposits thereon were removed by scrap-
ing, dispersed in acetone, and analysed by transmis-
sion electron microscopy ŽTEM; JEOL JEM 100CX
fine dispersion study and 0.5eV for larger energy
range with a collecting angle of 5 mrad for elemental
ratio calculation. and energy dispersive X-ray ŽEDX;
NORAN Instruments detector attached to the JEM
2010..
Fig. 2. SEM images of aligned nanotubes. Ža. Double bundles at low magnification in which nanotubes appear aligned. Žb–d. Higher
magnification of these bundles showing aligned nanotubes exhibiting uniform lengths Ž20 mm. and diameters Ž30–50 nm.. Že,f. Aligned
nanotube films also generated in other substrates etched with different laser defocus conditions.

3
02
M. Terrones et al.rChemical Physics Letters 285 (1998) 299–305
3
. Results and discussion
3
.1. Electron microscopy
SEM images show the presence of nanotube bun-
dles closely aligned with the nanotracks Žlength 1–5
mm. created by laser etching ŽFig. 2a–c,e.. The
tubes within these bundles were of uniform length
Ž(100 mm. and external diameter Žf30–50 nm;
Fig. 2d.. Probably as a result of the action of HCl,
the cobalt thin film gradually disappeared from the
plate during the experiment and appeared as cobalt
dichloride Žrevealed by mass spectrometric analyses
of the residues at the outlet bubblers. w21x. Interest-
ingly, aligned nanotube films also were observed in
other laser etched substrates using lower pulse ener-
gies out of focus ŽFig. 2e,f.. Related results have
been observed previously w24–26x.
TEM ŽFig. 3a. and HRTEM observations confirm
the presence of multi-layered graphitic tubes Ž30–50
nm OD 60–80 layers; Fig. 4.. In most cases, cobalt
Ž(50 nm OD particles. was detected by EDX analy-
sis within the nanotube tips ŽFig. 3b.. However,
occasionally substantial sections of tubes were filled
with cobalt ŽFig. 4..
Fig. 3. Ža. TEM image of a typical region filled with pure
EELS analyses show that the nanotubes consist of
carbon accompanied by traces of nitrogen Ž-5%;
see Fig. 5a.. In this context, it is possible that
nanotubes. Polyhedral particles and other graphitic nanostructures
are absent. Žb. Low-resolution TEM image showing a group of
parallel tubes together with three metal particles Ž-40 nm OD. at
the end of the bundle. These particles are lined up with the tube
growth direction and are thought to be intimately responsible for
the tubule formation. It is noteworthy that some nanotube tips do
not contain a metal particle. This may because catalytic particles
are located at a distance from the tips or, alternatively, are absent
due to sonication Že.g. 30 min. damage of the structures.
nitrogen, liberated during the cobalt catalysed tria-
zine decomposition, is somehow trapped inside the
hollow tubules. Alternatively, small C–N domains
2
Žsp hybridised C–N regions -5 nm OD. may have
been generated within the tube walls Žsee Fig. 5b..
2
However, no clear evidence for sp hybridised nitro-
gen was found, except a broad and weak peak corre-
sponding to the K-shell ionisation edge of nitrogen.
Both EELS and EDX measurements indicate the
absence of chlorine within the tubules.
3.2. Growth mechanism
Laser-etching of the cobalt thin film in air gener-
ates areas on the silica substrate free of cobalt.
During this process, the energised cobalt clusters
may condense and recrystalise as cobalt andror
cobalt oxide. It is possible that these clusters deposit
evenly Žwith a preferred crystal orientation. along
the edges of the eroded tracks. The clusters Ž(50
nm. probably possess a large surfacervolume ratio
Electron diffraction patterns w21x reveal a rela-
tively well-ordered graphitic arrangement within the
nanotubes, and that 25–30% of the tubes possess an
armchair configuration Žprobably metallic w6x.. This
phenomenon may be due to preferential orientation
of the cobalt or cobalt oxide clusters deposited along
the laser-etched track. As mentioned previously w21x,
these clusters appear to be essential for tube growth,
and subsequent axial carbon agglomeration, since
tubules are not formed in the absence of the catalyst.
with excellent catalytic potential for nanotube growth
ŽFig. 3b..
A possible mechanism whereby aligned nanotubes

M. Terrones et al.rChemical Physics Letters 285 (1998) 299–305
303
sults in axial nanotube growth. Temperature gradi-
ents occurring across the catalyst due to exothermic
decomposition of the CN units may be a key factor.
Nanotubes will continue to grow until the leading
catalytic particle is deactivated or removed by the
HCl.
Ž3. It is not clear at this stage why the nanotubes
are aligned. It may be that, due to overcrowding,
tubes grow simultaneously from the catalyst surface.
Aligned nanotube growth only occurred when the
experiment was conducted with the cobalt catalyst
deposited on the lower Žinverted. silica surface.
Fig. 4. Ža. HRTEM image of a cobalt-filled nanotube, in which a
cobalt–graphite interface from the nanotube shown in Žb.. The
cobalt planes correspond to the Ž002. reflections separated by 2.35
A˚ . The simulation of the diffraction pattern with the Ž120. zone
axis reproduces our observations of the cobalt planes.
grow over these etched regions may be rationalised
as follows:
Ž1. 2-Amino-4,6-dichloro-s-triazine decomposes
above 2358C liberating, in the first instance, HCl and
triazine rings, which are transported in the Ar stream
to the pyrolysis zone containing the inverted cat-
alytic film at 9508C. At this point, HCl reacts with
Fig. 5. Ža. EELS spectra of a typical nanotube showing ionisation
edges at f284 eV corresponding to the characteristic K-shell for
carbon. The sharp peak at f284 eV is due to transitions from the
the smooth Co thin film, producing CoCl in the gas
2
1
s core level to the p) band and the intensity starting at f291
phase. Almost simultaneously, the triazine rings,
possibly fragment on the surface of the hot catalyst
creating CN species.
eV arises from transitions to the p-orbital of the s) band. Addi-
tionally, a weak and broad peak corresponds to the K-shell
ionisation edge of nitrogen. The structure for this feature indicates
the presence of traces of nitrogen Ž-5%., possibly trapped within
Ž2. These CN units may fragment further, liberat-
the tube during its formation. The dispersion per channel is 0.3
ing gaseous N and carbon. The carbon and possibly
2
Ž .
eV. b HRTEM image of a defecteive nanotube containing traces
the nitrogen then precipitate on or diffuse through
the metal clusters w27x. This process eventually re-
2
of nitrogen. The defective edges may consist of small C–N sp
regions within the hexagonal carbon network.

3
04
M. Terrones et al.rChemical Physics Letters 285 (1998) 299–305
Growth of highly convoluted material was observed
when the catalyst surface face upwards. It appears
therefore that gravitational and convection effects
may well be significant.
providing Electron Microscope facilities. We thank
the Royal Society ŽLondon., BOC gases, the DERA
ŽNG., CONACYT-M e´ xico ŽHT., DGAPA-UNAM
IN 107-296 ŽHT., EU-TMR grant ŽJO. and the EP-
SRC for financial support.
It should be noted that the size of the catalytic
particles and the type of carbon source are crucial
factors for nanotube production. In some cases, when
the catalytic clusters are small, nanotube caps with-
out encapsulated metal particles are found w24x. This
phenomenon is uncommon when gaseous hydrocar-
bons are pyrolysed and may be associated with
particle mobility during the tube growth process.
Recent pyrolysis experiments conducted with re-
lated precursors such as tris-aminotriazine Žmela-
mine. and cyanuric chloride over Co, Ni and Fe
laser-etched thin films, also yielded similarly aligned
nanotube bundles and films. The use of Fe films in
conjunction with 2-amino-4,6-dichloro-s-triazine
generated very long aligned nanotube bundles Ž-250
mm, 20–40 nm OD. w24x.
References
w1x S. Iijima, Nature 354 Ž1991. 56.
w2x T.W. Ebbesen, P.M. Ajayan, Nature 358 Ž1992. 220.
3
w x
S. Iijima, P.M. Ajayan, T. Ichihashi, Phys. Rev. Lett. 69
Ž1992. 3100.
w4x T.W. Ebbesen, Annu. Rev. Mater. Sci. 24 Ž1994. 235.
w5x T. Guo, P. Nikoleav, A.G. Rinzler, D. Tom a´ nek, D.T. Col-
bert, R.E. Smalley, J. Phys. Chem. 99 Ž1995. 10694.
w6x A. Thess, R. Lee, P. Nikolaev, H. Dai, P. Petit, J. Robert, C.
Xu, Y.H. Lee, S.G. Kim, A.G. Rinzler, D.T. Colbert, G.E.
Scuseria, D. Tom ´a nek, J.E. Fischer, R.E. Smalley, Science
2
73 Ž1996. 483.
w7x L.C. Qin, S. Iijima, Chem. Phys. Lett. 269 Ž1997. 65.
w8x W.K. Hsu, J.P. Hare, M. Terrones, P.J.F. Harris, H.W.
Kroto, D.R.M. Walton, Nature 377 Ž1995. 687.
w9x W.K. Hsu, M. Terrones, J.P. Hare, H. Terrones, H.W. Kroto,
D.R.M. Walton, Chem. Phys. Lett. 261 Ž1996. 161.
w10x M. Endo, K. Takeuchi, K. Kobori, K. Takahashi, H.W.
Kroto, A. Sarkar, Carbon 33 Ž1995. 873.
4
. Conclusions
w11x S. Amelinckx, X.B. Zhang, D. Bernaerts, X.F. Zhang, V.
Laser etching of cobalt thin films provides a novel
Ivanov, J.B. Nagy, Science 265 Ž1994. 635.
route to catalysts which, in conjunction with the
pyrolysis of organic precursors Že.g. 2-amino-4,6-di-
chloro-s-triazine, etc.. yield to aligned nanotubes.
These tubes are of uniform length Ž(100 mm. and
diameter Ž30–50 nm OD. and grow perpendicularly
to the catalytic substrate only in the etched regions.
The orientation of the catalytic particles ablated dur-
ing etching appears to be crucial, nanotubes with
preferred helicity being favoured. It is possible that
matrices consisting of aligned nanotube bundles may
be useful as novel mechanically strong composite
materials and ultra-fine field emission sources
w13,28–30x.
w12x D. Bernaerts, X.B. Zhang, X.F. Zhang, S. Amelinckx, G.
Van Tendeloo, J. Van Landuyt, V. Ivanov, J.B. Nagy, Phil.
Mag. A 71 Ž1995. 605.
w13x W.A. De Heer, A. Chatelain, D. Ugarte, Science 270 Ž1995.
1
179.
w14x M. Terrones, J.P. Hare, W.K. Hsu, H.W. Kroto, A. Lappas,
W.K. Maser, A.J. Pierik, K. Prassides, R. Taylor, D.R.M.
Walton, in: K.M. Kadish, R.S. Ruoff ŽEds.., Recent Ad-
vances in the Chemistry and Physics of Fullerenes and
Related Materials, Vol. 2 ŽElectrochemical Society, Penning-
ton, NJ, 1995., p. 599.
w15x H.J. Dai, E.W. Wong, C.M. Lieber, Science 272 Ž1996. 523.
w
1
6 T.W. Ebbesen, H.J. Lezec, H. Hiura, J.W. Bennett, H.F.
x
Ghaemi, T. Thio, Nature 382 Ž1996. 54.
w17x M.J. Treacy, T.W. Ebbesen, J.M. Gibson, Nature 381 Ž1996.
78.
w18x E.W. Wong, P.E. Sheehan, C.M. Lieber, Science 277 Ž1997.
971.
6
1
Acknowledgements
w19x
M.R. Falvo, G.J. Clary, R.M. Taylor II, V. Chi, F.P. Brooks
Jr, S. Washburn, R. Superfine, Nature 389 Ž1997. 582.
w20x T.W. Ebbesen, P.M. Ajayan, H. Hiura, K. Tanigaki, Nature
We thank K. Kordatos for the supply of 2-amino-
,6-dichloro-s-triazine. We are grateful to R.
Guardi ´a n, R. Hern a´ ndez, P. Mexia, K. Prassides,
D.P. Randall, L. Rend o´ n, S. Tehuacanero, J. Thorpe,
and P.D. Townsend for useful discussions and for
3
67 Ž1994. 519.
4
w21x M. Terrones, N. Grobert, J. Olivares, J.P. Zhang, H. Ter-
rones, K. Kordatos, W.K. Hsu, J.P. Hare, P.D. Townsend, K.
Prassides, A.K. Cheetham, H.W. Kroto, D.R.M. Walton,
Nature 388 Ž1997. 52.

M. Terrones et al.rChemical Physics Letters 285 (1998) 299–305
305
w22x J.T. Thurston, J.R. Dudley, D.W. Kaiser, I. Hechenbleikner,
w26x Y. Chen, Z.L. Wang, J.S. Yin, D.J. Johnson, R.H. Prince,
Chem. Phys. Lett. 272 Ž1997. 178.
w27x R.T.K. Baker, Carbon 27 Ž1989. 315.
w28x A.G. Rinzler, J.H. Hafner, P. Nikolaev, L. Lou, S.G. Kim, D.
F.C. Schaefer, D. Holm-Hansen, J. Am. Chem. Soc. 73
Ž1951. 2981.
w23x M. Terrones, A.M. Benito, C. Manteca-Diego, W.K. Hsu,
O.I. Osman, J.P. Hare, D.G. Reid, H. Terrones, A.K.
Cheetham, K. Prassides, H.W. Kroto, D.R.M. Walton, Chem.
Phys. Lett. 257 Ž1996. 576.
Tom a` nek, P. Nordlander, D.T. Colbert, R.E. Smalley, Sci-
ence 269 Ž1995. 1550.
w29x P.G. Collins, A. Zettl, Appl. Phys. Lett. 69 Ž1996. 1969.
w30x Y. Saito, K. Hamaguchi, K. Hata, K. Uchida, Y. Tasaka, F.
w24x N. Grobert, S. Trasobares, M. Terrones, H. Terrones, W.K.
Hsu, A.K. Cheetham, J.P. Zhang, H.W. Kroto, D.R.M. Wal-
Ikazaki, M. Yumura, A. Kasuya, Y. Nishima, Nature 389
ton, manuscript in preparation.
w25x W.Z. Li, S.S. Xie, L.X. Qian, B.H. Chang, B.S. Zou, W.Y.
Zhou, R.A. Zhao, G. Wang, Science 274 Ž1996. 1701.
Ž1997. 555.