In situ imaging and spectroscopy of single-wall carbon nanotube synthesis by laser vaporization

Article abstract of DOI:10.1063/1.125696

The synthesis of single-wall carbon nanotubes by Nd:YAG laser vaporization of a graphite/(Ni, Co) target is investigated by laser-induced luminescence imaging and spectroscopy of Co atoms, C₂ and C₃ molecules, and clusters at 1000 °C

Full text of DOI:10.1063/1.125696

In situ imaging and spectroscopy of single-wall carbon nanotube synthesis by laser
vaporization
A. A. Puretzky, D. B. Geohegan, X. Fan, and S. J. Pennycook
Citation: Applied Physics Letters 76, 182 (2000); doi: 10.1063/1.125696
View online: http://dx.doi.org/10.1063/1.125696
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/76/2?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Limited Number of Possible Mean Diameters in the Evaporation Synthesis of SingleWalled Carbon Nanotubes
AIP Conf. Proc. 723, 57 (2004); 10.1063/1.1812042
Ultrafast carrier dynamics in single-walled carbon nanotubes probed by femtosecond spectroscopy
J. Chem. Phys. 120, 3368 (2004); 10.1063/1.1640339
Covalent Functionalization of Arc Discharge, Laser Ablation and HiPCO SingleWalled Carbon Nanotubes
AIP Conf. Proc. 685, 291 (2003); 10.1063/1.1628037
NearInfrared Photoluminescence of SingleWalled Carbon Nanotubes Obtained by the Pulsed Laser Vaporization
Method
AIP Conf. Proc. 685, 148 (2003); 10.1063/1.1628006
Diameter grouping in bulk samples of single-walled carbon nanotubes from optical absorption spectroscopy
Appl. Phys. Lett. 75, 2217 (1999); 10.1063/1.124969
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Tue, 02 Dec 2014 10:59:01

APPLIED PHYSICS LETTERS
VOLUME 76, NUMBER 2
10 JANUARY 2000
In situ imaging and spectroscopy of single-wall carbon nanotube synthesis
by laser vaporization
A. A. Puretzky, D. B. Geohegan,a) X. Fan, and S. J. Pennycook
Solid State Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831
͑Received 9 September 1999; accepted for publication 10 November 1999͒
The synthesis of single-wall carbon nanotubes by Nd:YAG laser vaporization of a graphite/͑Ni, Co͒
target is investigated by laser-induced luminescence imaging and spectroscopy of Co atoms, C2 and
C3 molecules, and clusters at 1000 °C in flowing 500 Torr Ar. These laser-induced emission images
under typical synthesis conditions show that the plume of vaporized material is segregated and
confined within a vortex ring which maintains a ϳ1 cm3 volume for several seconds. Using
time-resolved spectroscopy and spectroscopic imaging, the time for conversion of atomic and
molecular species to clusters was measured for both carbon ͑200 ␮s͒ and cobalt ͑2 ms͒. This rapid
conversion of carbon to nanoparticles, combined with transmission electron microscopy analysis of
the collected deposits, indicate that nanotube growth occurs over several seconds in a plume of
mixed nanoparticles. By adjusting the time spent by the plume within the high-temperature zone
using these in situ diagnostics, single-walled nanotubes of controlled length were grown at an
estimated rate of 0.2 ␮m/s. © 2000 American Institute of Physics. ͓S0003-6951͑00͒02102-1͔
The remarkable electronic and structural properties of
carbon nanotubes promise to revolutionize application areas
from nanoscale electronics to ultralightweight structural
materials.1 Since electronic transport in single-wall carbon
nanotubes ͑SWNT͒ is directly related to their atomic struc-
ture, it is essential to understand what controls nanotube size
and structure during synthesis. Furthermore, methods for
large-scale production of SWNT have not been developed
because so little is understood about the synthesis process.
Up until now, three principal methods have been used to
synthesize carbon nanotubes. These include laser vaporiza-
tion ͑LV͒,2 dc-arc vaporization ͑AV͒,3,4 and chemical vapor
deposition ͑CVD͒.5,6 Unlike CVD, both LV and AV now
produce nearly exclusively SWNT. However, very little is
known about where and when the SWNT are formed in LV
or AV growth chambers, the identity of the precursor spe-
cies, and the growth rates.
that used by Guo et al.2 It consists of a quartz tube ͑2 in.
diameter, 24 in. length͒ mounted inside a hinged tube fur-
nace ͑12 in. length͒ operating at 1000 °C. The quartz tube
was O-ring sealed to standard 4.5-in.-Conflat vacuum com-
ponents. The ablation and probe laser beams entered through
the same Suprasil window which was mounted in a vacuum
flange. Argon gas was introduced around this window, con-
trolled at 100 sccm to maintain a 500 Torr pressure, and
pumped out through a needle valve downstream of a brass,
water-cooled collector which was inserted as shown.
A 1-in.-diameter graphite target containing 1 at. % each
of Ni ͑Alfa, 2.2–3.0 ␮m, 99.9%͒ and Co ͑Alfa, 1–6 ␮m,
99.8%͒ powders was prepared with the carbon cement ͑Dy-
lon GC͒ procedure described in Ref. 2. The target was
screwed onto a 0.25-in.-diameter graphite rod and was ro-
tated during operation. The ablation laser beam ͑single 8 ns
pulses, 140 mJ of combined 1.06 ␮m and 532 nm outputs͒
was focused to a 1.6-mm-diameter donut-shaped spot on the
target. An unfocused, time-delayed XeCl-laser ͑308 nm, 30
ns pulse, 20 mJ/cm2͒ was used to induce luminescence in the
ablation plume.
Pulsed LV is especially amenable to diagnostic investi-
gations. The vaporizing pulse lasts only ϳ10 ns and SWNT
growth then can occur undisturbed from further excitation,
even for single laser ablation events.7 Recent spectroscopic
measurements of the luminous laser plasma have been made
for pulsed LV at early times after Nd:YAG laser ablation
The furnace was opened for ϳ3 s to permit imaging of
the plume with a gated ICCD-camera system ͑Princeton In-
struments, 5-ns-minimum gate, 200–820 nm spectral range͒.
8
͑Ͻ200 ␮s͒ and after long-pulse CO2 laser ablation at room
temperature.9 However, these measurements were limited to
times while the ablated material is still quite hot.
Recently, laser induced luminescence ͑LIL͒, gated inten-
sified charge-coupled device ͑ICCD͒ imaging, and optical
emission spectroscopy were applied to understand how nano-
particles form and grow in pulsed laser ablation plumes at
room temperature.10–12 In this study these techniques are ap-
plied to investigate SWNT growth under typical ablation
conditions at high temperatures.
FIG. 1. Schematic of the 2-in.-diameter quartz tube and hot furnace used for
laser vaporization growth of SWNT with in situ LIL imaging and spectros-
copy diagnostics. Beam geometries and imageable area are indicated. The
black dots and the numbers show the collection points of the ablated mate-
rial: ͑1͒ upstream; ͑2͒ collector. The C/Ni/Co target was positioned at dif-
ferent distances, d, from the front of the furnace.
The SWNT growth setup in Fig. 1 was fashioned after
a͒
Electronic mail: odg@ornl.gov
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
0003-6951/2000/76(2)/182/3/$17.00 182 © 2000 American Institute of Physics
132.174.255.116 On: Tue, 02 Dec 2014 10:59:01

Appl. Phys. Lett., Vol. 76, No. 2, 10 January 2000
Puretzky et al.
183
FIG. 2. TEM images of the raw soot collected: ͑a͒ downstream on the
collector ͑point 2, for dϭ21 cm in Fig. 1͒, showing a very high fraction of
SWNT bundles along with metal nanoparticles ͑black dots͒; ͑b͒ upstream
͑point 1, for dϭ12.5 cm in Fig. 1͒ showing short ͑ϳ100 nm͒ SWNT in the
early stages of growth, controlled by limiting the growth time to 0.5 s.
At each plume location, the plasma emission and/or laser-
induced luminescence was collected for optical emission
spectroscopy using a 0.3 m spectrometer ͑Acton VM-503,
resolution either 10 or 1.3 nm͒ with an intensified, gated
diode array ͑Princeton Instrument IRY-700RB͒.
As shown in Fig. 2͑a͒, ϳ10-␮m-long SWNT bundles
comprised very high fractions of the deposit found down-
stream on the collector ͑Ͼ70%–90% as estimated from
field-emission SEM analysis͒ when the target was positioned
dϭ21 cm from the front of the furnace. Metal catalyst ͑Ͻ20
nm diameter͒ and small amounts of amorphous carbon nano-
particles accounted for the rest of the deposit ͑as described in
the literature͒.2 EELS investigation of individual metal-
catalyst nanoparticles ͑using a 0.3-nm-diameter electron
beam͒ revealed that their compositions were reproducibly
NixCoy(xϷyϷ0.5), with both Ni and Co uniformly distrib-
uted inside each nanoparticle to within sub-nm dimensions.
To determine where and when SWNT growth occurred,
LIL-ICCD imaging and spectroscopy of the C/Ni/Co plume
were performed at different times after laser vaporization,
⌬t, spanning 20 nsр⌬tр3 s. For ⌬tр0.2 ms, a series of
shocks within the plume13,14 results in segregation of the
FIG. 3. Laser-induced luminescence ͑LIL͒ images of the C/Ni/Co plume
during SWNT synthesis. A Nd:YAG laser vaporizes a C/Co/Ni target ͑right
edge of figure͒ inside a 2-in.-quartz tube at 1000 °C in 500 Torr Ar ͑flowing
to the right at 100 sccm, ϳ0.6 cm/s at the center of the tube͒. Each image
represents a different ablation event. ͑a͒ Unfiltered images show the location
of all C/Co/Ni species during nanotube growth ͑100 ns gate width, opened
⌬␶ϭ100 ns after the XeCl-laser pulse͒; ͑b͒ selective imaging in the 320–
380 nm spectral region to locate ground-state atomic Co in the plume ͑100
ns gate width, ⌬␶ϭ0͒; ͑c͒ temporal history of the total number density of
ground-state atomic Co, determined from integrating the LIL from a com-
plete set of atomic-Co images shown in ͑b͒.
10
ablated material into the vortex ring15,16 ͑or ‘‘smoke ring’’͒
shown at ⌬tϭ0.2 ms in Figs. 3͑a͒ and 3͑b͒. As shown in Fig.
3͑a͒, the vortex motion efficiently traps aggregated nanopar-
ticles in a confined volume for long times ͑ϳ3 s within ϳ1
cm3 in this study͒.
The leading edge of the plume propagates with velocities
of: 103 cm/s between 200 ␮sϽ⌬tϽ2 ms; 50 cm/s for
10 msϽ⌬tϽ50 ms; and 6 cm/s at 30 msϽ⌬tϽ200 ms. After
⌬tϭ2 s the plume stops moving upstream, and the plane of
the vortex ring tilts toward the tube axis, possibly due to flow
currents or thermophoretic forces. The plume is then dragged
by the gas flow back to the collector with an estimated flow
velocity of 0.6 cm/s. Finally, nanotubes, metal catalyst par-
ticles and unconverted carbon soot deposit on the cool col-
lector surface by thermophoresis.12 LIL-ICCD imaging en-
sured that all deposits and images were collected without
interference from previous laser shots using laser repetition
rates less than 0.016 Hz.
FIG. 4. Plasma emission ͑dashed͒ and laser-induced luminescence ͑solid͒
spectra measured at different time delays after the ablation laser pulse, ⌬t,
and distances, x, from the target: ͑a͒ ⌬tϭ20 ␮s, xϭ0.2 cm; ͑b͒ ⌬t
ϭ100 ␮s, xϭ0.5 cm; ͑c͒ ⌬tϭ1 ms, xϭ2 cm; and ͑d͒ ⌬tϭ20 ms,
x
ϭ5 cm. Acquisition times of 100 ns ͓͑a͒, ͑b͔͒ and 3.5 ␮s ͓͑c͒, ͑d͔͒ began 50
ns after the XeCl-laser pulse. The inset in ͑c͒ shows a 1.3-nm-resolution
spectrum of induced fluorescence from the following transitions: ͑1͒
b4F7/2– y2G09/2 at 341.23 nm; ͑2͒ b4F9/2– y4G011/2 at 345.35 nm; ͑3͒
b4F5/2– y4G0 at 350.98 nm; ͑4͒ b4F7/2– y4G09/2 at 352.98 nm; and ͑5͒
Optical emission spectra were obtained for each image.
Figure 4 presents a short summary. At early times in the
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the te70r/m2 s at: http://scitation.aip.org/termsconditions. Downloaded to IP:
plume expansion, emission from excited states in the plasma
b4F5/2– y4D at 357.50 nm.
132.174.255.116 On: Tue, 02 Dec 20154/210:59:01

184
Appl. Phys. Lett., Vol. 76, No. 2, 10 January 2000
Puretzky et al.
dominates any laser-induced luminescence. As in Fig. 4͑a͒,
the d3⌸g→a3⌸u Swan system of C2 and the A1⌸u
→X1⌺g band C3 are prominent at these times. Later when
the plasma expands, cools, and recombines, the 308-nm-
XeCl laser can induce emission by pumping transitions from
the ground states: of C3 ͑via A1⌸u X1⌺g͒; of atomic Co
͑via a4FJ y4G0J͒; and of atomic Ni ͑via a3D1 y3D02͒.
In addition, blackbody radiation can be induced from
carbon clusters,17 nanoparticles,18 and nanotubes. The inten-
sity of this blackbody emission, I, is defined by IϭAr3(T0
ϩ⌬T)5 where T0 is the initial temperature of the cluster, ⌬T
is the temperature increase due to laser-heating, A is a con-
stant, and r is the cluster radius. This radiation becomes ob-
servable coincident with the disappearance of the C2 and C3
bands in both the plasma emission and the laser-induced
spectra for ⌬tϾ200 ␮s at 1000 °C ͑and ⌬tϾ100 ␮s at room
temperature͒. We conclude that nearly all of the carbon in
the plume has converted into clusters or larger aggregates by
these times.
However, the Co in the plume remains in atomic form
until much later. As shown in Figs. 3͑c͒ and 4͑c͒, the
ground-state Co population peaks at ⌬tϭ0.8 ms and drops
by an order of magnitude by ⌬tϭ2 ms, permitting estimates
of the Co clustering time of ϳ2 ms at 1000 °C ͑ϳ1 ms at
room temperature͒. Blackbody radiation remains the only
feature of the spectra taken at all later times ͓Fig. 4͑d͔͒.
Further evidence of the sequential condensation of car-
bon and cobalt into clusters is the relatively uniform spatial
distribution of atomic Co in the plume for ⌬tϽ2 ms com-
pared to the vorticity of the clustered carbon material ͓com-
pare Figs. 3͑a͒ and 3͑b͔͒. We believe that the higher diffu-
sivity of the atomic Co effectively competes with the
hydrodynamic trapping during this time. Only the leading
edge of the atomic-Co plume overlaps the carbon-cluster
vortex ring during the condensation of the Co atoms. The
NixCoy alloy particles indicate similar condensation times
for Ni and Co. It is quite possible that the carbon clusters
serve as condensation centers for metal cluster growth.
Both imaging and spectroscopy indicate that within the
first few milliseconds after laser ablation, atoms and mol-
ecules of both carbon and metal catalyst disappear due to
condensation into nanoparticles. Unless SWNTs grow very
rapidly from atoms and molecules within these first few mil-
liseconds, the majority of growth appears to occur from a
feedstock of mixed nanoparticles over seconds of annealing
time.
average growth rate of 0.2 ␮m/s at 1000 °C. The relative
yield of the carbon particles is much larger than the yield of
the carbon nanotubes which clearly shows that the time spent
by the plume in the hot zone ͑ϳ0.5 s͒ was not sufficient to
convert all of the carbon material into nanotubes.
In summary, these results confirm the conclusions of the
time-resolved imaging and spectroscopy measurements
͑along with ex situ TEM, EELS, and FESEM͒. Although
SWNT may initiate growth during the first few milliseconds
after laser ablation, the majority of growth occurs inside the
oven from a mixture of gas-suspended carbon and metal-
catalyst nanoparticles which are hydrodynamically trapped
in a vortex ring within a ϳ1 cm3 volume for several seconds.
The spectroscopy at early times after laser ablation indicates
that the plume initially consists of atomic and molecular spe-
cies, with no evidence of hot molten particulates which were
recently suggested as the primary ejecta.19 Condensation of
carbon occurs within 0.2 ms after ablation, while atomic Co
condenses much later ͑between 1.5 and 2 ms͒. Through these
images of the growth process, we conclude that long ͑ϳ10
␮m͒ SWNT can form from the small amount of material
vaporized in a single-laser shot, a remarkable feat of self-
assembly.
The authors gratefully acknowledge research assistance
by M. L. Simpson, M. Guillorn, and helpful conversations
with D. J. Rader. This work was sponsored by the Division
of Material Science, U.S. Department of Energy under Con-
tract No. DE-AC05-96-R22464 with Lockheed Martin En-
ergy Research Corp.
1 B. Yakobson and R. E. Smalley, Am. Sci. 85, 324 ͑1997͒.
2 T. Guo, P. Nikolaev, A. Thess, D. T. Colbert, and R. E. Smalley, Chem.
Phys. Lett. 236, 419 ͑1995͒.
3 T. S. Iijima and T. Ichihashi, Nature ͑London͒ 363, 603 ͑1993͒.
4 C. Journet, W. K. Maser, P. Bernier, A. Loiseau, M. Lamy de la Chapelle,
S. Lefrant, P. Deniard, R. Lee, and J. E. Fisher, Nature ͑London͒ 388, 756
͑1997͒.
5 H. M. Cheng, F. Li, G. Su, H. Y. Pan, L. L. He, X. Sun, and M. S.
Dresselhaus, Appl. Phys. Lett. 72, 3282 ͑1998͒.
6 B. C. Satiskumar, A. Govindaraj, R. Sen, and C. N. R. Rao, Chem. Phys.
Lett. 293, 47 ͑1998͒.
7 M. Yudasaka, T. Ichihashi, T. Komatsu, and S. Iijima, Chem. Phys. Lett.
299, 91 ͑1999͒.
8 S. Arepalli and C. D. Scott, Chem. Phys. Lett. 302, 139 ͑1999͒.
9 F. Kokai, K. Takahashi, M. Yudasaka, R. Yamada, T. Ichihashi, and S.
Iijima, J. Phys. Chem. B103, 4346 ͑1999͒.
10 D. B. Geohegan, A. A. Puretzky, G. Duscher, and S. J. Pennycook, Appl.
Phys. Lett. 72, 2987 ͑1998͒.
11 D. B. Geohagan, A. A. Puretzky, G. Duscher, and S. J. Pennycook, Appl.
Phys. Lett. 73, 438 ͑1998͒.
In order to check this conclusion and estimate the
SWNT growth rate, the target was positioned at several dis-
tances from the front furnace edge to adjust the time spent
within the hot zone before exiting the furnace in the up-
stream direction. LIL-ICCD imaging recorded the plume dy-
namics from ablation to deposition and the collected deposits
were examined by TEM.
Figure 2͑b͒ shows a TEM image of a deposit collected
for a target position of dϭ12.5 cm ͑see Fig. 1͒, where
LIL-ICCD imaging measured a time of 0.5–0.7 s spent by
the plume inside the furnace. The collected material consists
of aggregated carbon and metal-catalyst nanoparticles, and
thin SWNT bundles of only ϳ100 nm length, yielding an
12 D. B. Geohegan, A. A. Puretzky, and D. J. Rader, Appl. Phys. Lett. 74,
3788 ͑1999͒.
13 D. B. Geohegan, A. A. Puretzky, R. L. Hettich, X.-Y. Zheng, R. E. Hau-
fler, and R. N. Compton, Trans. Mater. Res. Soc. Jpn. 17, 349 ͑1994͒.
14 Images also available online at www.ornl.gov/ϳodg.
15 A. V. Bulgakov and N. M. Bulgakova, J. Phys. D: Appl. Phys. 31, 693
͑1998͒.
16 F. Garrelie, C. Champeaux, and A. Catherinot, Appl. Phys. A: Mater. Sci.
Process. 69, 45 ͑1999͒.
17 D. B. Geohegan and A. A. Puretzky, Mater. Res. Soc. Symp. Proc. 397, 55
͑1996͒.
18 E. A. Rohlfing, J. Chem. Phys. 89, 6103 ͑1988͒.
19 M. Yudasaka, R. Yamada, N. Sensui, T. Wilkins, T. Ichihashi, and S.
Iijima, J. Phys. Chem. B 103, 6224 ͑1999͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.255.116 On: Tue, 02 Dec 2014 10:59:01