Growth of Vertically Aligned Nitrogen-Doped Carbon Nanotubes: Control of the Nitrogen Content over the Temperature Range 900-1100 °C

Article abstract of DOI:10.1021/jp0274536

Nitrogen-doped carbon nanotubes were grown vertically aligned on the iron nanoparticles deposited on silicon substrates, by thermal chemical vapor deposition of methane/ammonia and acetylene/ammonia mixtures in the temperature range 900-1100 °C. The concentration of the nitrogen atoms has been controlled in the range 2-6 atomic percent, by the flow rate of ammonia. All nanotubes exhibit a bamboo-like structure over this temperature range. The growth rate is insensitive to this nitrogen content, but the structure is strongly dependent on it. As the nitrogen content increases, the thicker compartment layers form uniformly at a regular distance and the relative amount of crystalline graphitic sheets is notably reduced. Electron energy-loss spectroscopy reveals the higher nitrogen concentration and the lower crystallinity for the compartment layers compared to the wall. The growth of nitrogen-doped carbon nanotubes has been explained using a base growth mechanism proposed for carbon nanotubes. We suggest that the nitrogen doping would produce more flexible compartment layers connecting the wall under a less strain.

Full text of DOI:10.1021/jp0274536

12958
J. Phys. Chem. B 2003, 107, 12958-12963
Growth of Vertically Aligned Nitrogen-Doped Carbon Nanotubes: Control of the Nitrogen
Content over the Temperature Range 900-1100 °C
Yun Tack Lee, Nam Seo Kim, Seung Yong Bae, and Jeunghee Park*
Department of Chemistry, Korea UniVersity, Jochiwon 339-700, South Korea
Soo-Chang Yu
Department of Chemistry, Kunsan National UniVersity, Chonbuk 573-701, South Korea
Hyun Ryu and Hwack Joo Lee
Korea Research Institute of Standards and Science, Daejeon 305-600, South Korea
ReceiVed: NoVember 13, 2002; In Final Form: June 9, 2003
Nitrogen-doped carbon nanotubes were grown vertically aligned on the iron nanoparticles deposited on silicon
substrates, by thermal chemical vapor deposition of methane/ammonia and acetylene/ammonia mixtures in
the temperature range 900-1100 °C. The concentration of the nitrogen atoms has been controlled in the
range 2-6 atomic %, by the flow rate of ammonia. All nanotubes exhibit a bamboo-like structure over this
temperature range. The growth rate is insensitive to this nitrogen content, but the structure is strongly dependent
on it. As the nitrogen content increases, the thicker compartment layers form uniformly at a regular distance
and the relative amount of crystalline graphitic sheets is notably reduced. Electron energy-loss spectroscopy
reveals the higher nitrogen concentration and the lower crystallinity for the compartment layers compared to
the wall. The growth of nitrogen-doped carbon nanotubes has been explained using a base growth mechanism
proposed for carbon nanotubes. We suggest that the nitrogen doping would produce more flexible compartment
layers connecting the wall under a less strain.
1
. Introduction
system. Hot filament CVD of methane/nitrogen/ammonia and
pyrolysis of acetylene/ammonia/iron pentacarbonyl mixtures
Carbon nanotubes (CNTs) are an important focus of current
2
9-31
were also used.
Recently, Zhang and co-workers reported
research because of many potential applications including field
the vertically aligned N-doped CNTs using the pyrolysis of iron
1
-3
emission displays and radiation source, conductive and high-
3
2
phthalocyanine under the ammonia atmosphere. Those works
showed that the N-doped CNTs have a bamboo-like structure
with distinctive compartment layers. However, the dependence
of the growth properties on the N concentration has not been
thoroughly examined yet.
strength composites,4 nanoscale semiconductor devices,
,5
6-8
9
10
energy storage devices, and hydrogen storage media. The
CNTs are usually made by arc discharge, laser vaporization,
pyrolysis, or chemical vapor deposition (CVD).1
1-16
Synthesis of nitrogen (N)-doped CNTs has recently been
considered as a possible method to control the electronic
properties of CNTs in a well-defined way. An enhancement of
conductivity is expected, because the additional electrons
contributed by the nitrogen atom provide electron carriers for
the conduction band.17 The advantage of such nanotubes is that
their electronic properties are primarily determined by the
composition and are thus relatively easy to control. Many
In this study, we report a systematic approach to the influence
of the N doping on the growth rate and structure of CNTs over
the temperature range 900-1100 °C. It would lead eventually
to an understanding of the growth mechanism of N-doped CNTs,
which is prerequisite for various potential applications. The
vertically aligned N-doped CNTs on the iron (Fe) nanopar-
ticles-deposited silicon (Si) substrates were synthesized by
thermal CVD of methane (CH4)/ammonia (NH3) and acetylene
1
7-32
research groups reported the synthesis of the N-doped CNTs.
(C2H2)/NH3 mixtures at 900-1100 °C. The Fe catalytic nano-
Terrones et al. reported the generation of aligned N-doped
carbon nanofibers and CNTs by pyrolyzing various C/N sources,
i.e., triazine, melamine, ferrocene/melamine mixture, and fer-
rocene/C60 mixtures, in ammonia atmosphere.14,17-20 Wang and
particles were formed using a simple method: a thin film of
ethanol solution containing Fe salt is deposited on the Si sub-
strate, followed by the NH3 pretreatment at the growth tem-
perature. Configuration and structural characteristics of CNTs
have been investigated using electron microscopy and Raman
spectroscopy. The composition of nanotubes was determined
using X-ray photoelectron spectroscopy (XPS), Auger spec-
troscopy, and electron energy-loss spectroscopy (EELS).
co-workers adopted a microwave plasma-assisted CVD of
methane/nitrogen mixture.2
1-24
The Rao group employed the
25,26
pyrolysis of pyridine over Fe or Co catalyst.
The pyrolysis
of nickel phthalocyanine was used by Yudasaka et al. and
Suenaga et al.2
7,28
Sung et al. synthesized the N-doped CNTs
on a porous alumina template in a microwave-excited plasma
of acetylene/nitrogen using an electron cyclotron resonance CVD
2
. Experimental Section
Some 10 mm × 15 mm size n-type Si(100) substrates were
*
Corresponding author. E-mail: parkjh@korea.ac.kr.
coated with a drop of 0.01 M FeCl2‚4H2O (Aldrich, 99.995%)
1
0.1021/jp0274536 CCC: $25.00 © 2003 American Chemical Society
Published on Web 11/05/2003

Vertically Aligned N-Doped Carbon Nanotubes
J. Phys. Chem. B, Vol. 107, No. 47, 2003 12959
respectively. The 35 µm long CNTs are grown for 10 min at
1
000 °C (Figure 1b). The flow rate of CH4 and NH3 is 160 and
4
0 sccm, respectively. The length of the CNTs grown during 5
min at 1100 °C is 160 µm (Figure 1c). The flow rate of CH4
and NH3 is 150 and 50 sccm, respectively. From several
experimental runs, the average length of the CNTs grown at
9
2
00, 1000, and 1100 °C has been measured as 10 ( 10, 30 (
0, and 150 ( 25 µm, respectively. Dividing the length by the
growth time yields the average growth rate of CNTs, which is
( 1, 3 ( 2, and 30 ( 5 µm/min, respectively, for 900, 1000,
1
and 1100 °C. The growth rate at 900 and 1000 °C is constant
during the first 10 min and then decreases as the carbonaceous
particles cover the surface of substrate. At 1100 °C the growth
rate decreases after 5 min. The number density of CNTs on the
1
0
2
substrate is ∼10 /cm . The pretreatment condition has been
adjusted to form the size and density of nanoparticles suitable
to grow the high-density CNTs.
XPS and Auger analyses show that these CNTs are doped
with the same N concentration, 4.5 atomic % (at. % or %). We
synthesized the vertically aligned CNTs using the approximately
same total flow rate (200 sccm) of CH4 and NH3 mixture. The
N concentration of CNTs depends on the flow rate of NH3 as
shown in Table 1. The CNTs grown without the flow of NH3
are doped with 2% N atoms. The N concentration can reach up
to 6% for the CNTs grown at 1100 °C. It is noteworthy that
the growth rate increases dramatically with the temperature,
while it is insensitive to the flow rate of NH3.
The N-doped CNTs were also synthesized by the thermal
CVD of C2H2/NH3 mixture at temperatures in the range of
9
00-1100 °C. The N content was adjustable by the flow rate
of NH3, as shown in Table 1. The CNTs are doped with 2% N
atoms under no NH3 flow. The concentration of N atoms
increases to 5% when the flow rate of NH3 is 40 sccm. The
growth rate is independent of the N content in the range 2-5%.
The growth rate is 7 ( 2, 14 ( 2, and 29 ( 5 µm/min,
respectively, for 900, 1000, and 1100 °C, showing the 40-times
enhancement of growth rate with the temperature. The growth
rate was reproducible within the experimental error.
Figure 1. SEM micrographs of the vertically well-aligned N-doped
CNTs grown on the Si substrates. The concentration of the N atoms is
about 4.5%. (a) The 15 µm long CNTs grown for 10 min at 900 °C,
(
b) 35 µm long CNTs grown for 10 min at 1000 °C, and (c) 160 µm
long CNTs grown for 5 min at 1100 °C.
ethanol solution and then dried. The thickness of the film was
adjusted to be ∼300 nm. The coated substrates were loaded
on a quartz boat placed inside a 2.5 cm diameter and 80 cm
long quartz tube reactor. Ar (99.999%) flowed into the reactor
in order to prevent the oxidation of Fe while the temperature
rises. The substrates were pretreated by NH3 (99.999%) with a
flow rate of 20 sccm for 1-10 min, forming the nanometer-
size Fe particles. The CNTs were grown using the mixture of
CH4/NH3 or C2H2/NH3 for 5-10 min at temperatures ranging
between 900 and 1100 °C. The flow rate of CH4 and NH3 was
(2) The Dependence of C/N Composition on the Flow Rate
of NH3. Figure 2a shows the XPS data for the N-doped CNTs
grown at 1000 °C using the CH4/NH3 mixture with a total flow
rate of about 200 sccm. The flow rate of NH3 is selected as 0,
5, and 40 sccm. The C 1s, N 1s, and O 1s peaks appear at 284.5,
400.7, and 532.9 eV, respectively. A strong and sharp peak at
284.5 eV confirms that the major component is carbon. The O
peak arises from the oxygen in air and/or adsorbed on the surface
of the nanotubes. The N content is 2.0, 3.3, and 4.7%,
respectively, for 0, 5, and 40 sccm of the NH3 flow rate.
5
0-300 and 0-150 sccm, respectively. The flow rate of C2H2
and NH3 was 10 and 0-80 sccm, respectively. The reactor was
cooled to room temperature under Ar ambient after the growth.
Scanning electron microscopy (SEM, Hitachi S-4300), transmis-
sion electron microscopy (TEM, Hitachi H9000-NAR, 300 kV),
XPS (Physical electronics PHI 5800), Auger spectroscopy (PHI
Figure 2b shows a plot of the N content as a function of the
NH3 flow rate. Each data point is obtained using the flow rate
of NH3 as 0, 5, 20, 40, 50, 100, and 150 sccm. Total flow rate
is about 200 sccm. As the flow rate of NH3 increases, the N
content increases from 2.0 to 4.7% then it decreases to 2.6%.
The data are summarized in Table 1.We also changed the flow
rate of CH4 from 10 to 300 sccm and found that the nearly
same N content was achieved for a given flow rate of NH3. We
obtained the maximum N content to be 4.4 and 6.0% for 900
and 1100 °C when the NH3 flow rate is 30 and 100 sccm,
respectively.
6
80), EELS (GATAN GIF-2000) attached to a TEM (Philips
CM200, 200 kV), and Raman spectroscopy (Renishaw micro-
Raman 2000) using the excitation wavelength as the 514.5 nm
line of argon ion laser were employed.
3
. Results
1) Morphology and Growth Rate of the Nanotubes. Figure
shows SEM images for the vertically well-aligned nanotubes
(
(3) Structure of the N-Doped CNTs. Figure 3 shows TEM
images for the nanotubes grown at three temperatures, 900,
1000, and 1100 °C. As the temperature increases, the average
diameter of CNTs increases from 40 to 150 nm and the diameter
distribution becomes broader. The diameters of CNTs are
1
homogeneously grown on a large area of the substrates. The
CNTs grown for 10 min at 900 °C exhibit a length of 15 µm
(Figure 1a). The flow rate of CH4 and NH3 is 180 and 20 sccm,

12960 J. Phys. Chem. B, Vol. 107, No. 47, 2003
Lee et al.
TABLE 1: Condition of NH
3
Pretreatment, Growth Rate, Diameter, Concentration of N Atoms, and Degree of Crystallinity of
N-Doped CNTs Grown Using Thermal CVD of CH
4
/NH
3
and C
2
H
2
/NH
3
Mixtures at Temperatures Ranging from 900 to 1000
°
C
3
NH pretreatment
flow
rate
(sccm)
CH
flow rate
(sccm)
4
or C
2
H
2
NH
flow rate
(sccm)
3
growth
rate
(µm/min)
diameter
range
(nm)
average
diameter
(nm)
temp
time
(min)
at. %
of Na
b
(°C)
D G
I /I
CH
4
+ NH
3
900
000
20
20
10
5
180
200
20
0
5
20
40
50
100
150
0
1 ( 1
2 ( 2
3 ( 2
3 ( 2
3 ( 2
3 ( 2
30-70
50
100
4.4 ( 0.4
2.0 ( 0.2
3.3 ( 0.3
4.5 ( 0.4
4.7 ( 0.4
4.5 ( 0.4
3.2 ( 0.3
2.6 ( 0.2
2.0 ( 0.5
4.5 ( 0.5
6.0 ( 1.0
0.75 ( 0.08
0.48 ( 0.05
0.61 ( 0.06
0.66 ( 0.06
0.66 ( 0.06
-
1
50-150
2
1
1
1
1
00
80
60
50
00
c
-
-
c
50
-
0.43 ( 0.05
0.44 ( 0.04
0.54 ( 0.05
0.62 ( 0.06
1
100
20
1
200
30 ( 5
30 ( 5
100-200
150
40
1
1
50
00
50
100
c
-
+ NH
3
900
000
20
20
10
5
10
10
0
7 ( 1d
7 ( 1
30-100
70-120
2 ( 0.5
5 ( 0.5
2 ( 0.5
5 ( 0.5
3.5 ( 0.5
2 ( 0.5
3( 0.5
0.65 ( 0.06
0.86 ( 0.05
0.67 ( 0.07
0.82 ( 0.08
0.70 ( 0.07
0.45 ( 0.05
0.60 ( 0.08
2
C H
2
4
0
0
0
0
0
0
d
1
1
14 ( 3
100
80
80
4
8
14 ( 3
14 ( 3
d
100
20
1
10
28 ( 5
100-200
150
2
28 ( 5
a
Atomic % measured by XPS. b Intensity ratio of D band to G band in Raman spectrum. c Not vertically aligned. d Reference 36.
Figure 2. (a) XPS spectra of the N-doped CNTs grown at 1000 °C,
showing 2.0, 3.3, and 4.7% N atoms. (b) A plot of the N content for
the CNTs grown using various flow rates of NH (sccm).
3
summarized in Table 1. For all temperatures, the nanotubes
exhibit exclusively a bamboo-like structure in which the tube
inside is separated into a series of compartments. The 2 and
5
% N-doped nanotubes grown using C2H2/NH3 mixture at 900
Figure 3. TEM images for (a) 2% and (b) 5% N-doped CNTs grown
at 900 °C; (c) 2.0% and (d) 4.7% N-doped CNTs grown at 1000 °C;
(e) 2.0% and (f) 4.5% N-doped CNTs grown at 1100 °C.
°
C are shown in Figure 3, parts a and b, respectively. The 2.0
and 4.7% N-doped nanotubes grown using a CH4/NH3 mixture
at 1000 °C are shown in Figure 3, parts c and d, respectively.
Figure 3e and 3f correspond to the 2.0 and 4.5% N-doped
nanotubes grown using CH4/NH3 mixture at 1100 °C, respec-
tively. The structures of CNTs grown using C2H2 and CH4 are
similar for a given N % and temperature.
For 2% N-doped CNTs, the graphitic sheets separating the
tube inside (so-called “compartment layers”) are formed more
randomly and frequently as the temperature increases (Figures
3a, 3c, and 3e). A close examination of the CNTs grown at

Vertically Aligned N-Doped Carbon Nanotubes
J. Phys. Chem. B, Vol. 107, No. 47, 2003 12961
Figure 4. (a) HRTEM image for a 2.0% N-doped CNT grown at 1100
°
C and (b) its atomic-resolved image. (c) HRTEM image for a 6.0%
N-doped CNT grown at 1100 °C and (d) its atomic-resolved image.
As the N content increases, the degree of crystalline perfection
decreases.
Figure 5. EELS data for the N-doped CNTs grown at 1100 °C,
showing the K-shell ionization edges for carbon (C-K) and nitrogen
(
N-K). (a) The compartment layers and the wall contain 2% N atoms.
900 °C reveals that the extremely thin compartment layers exist.
(b) The compartment layers and the wall contain 9 and 4% N atoms,
respectively.
At 1100 °C, a number of nanotubes show a nearly closed inside
with the stacked compartment layers. The most distinctive
structural change due to the increase of the N content is the
appearance of the thick and curved compartment layers at a
regular distance. As the concentration of N atoms in the CNTs
grown at 900 °C increases from 2 to 4.7%, the compartment
layers become more obvious (Figure 3, parts a and b). At 1000
Figure 5a shows that the N content is about 2% for both com-
partment and wall parts. The π* peak of the C-K edge is
stronger and sharper in the wall, compared to that of the com-
partment layers. It indicates that the crystallinity of the wall is
higher than that of the compartment layers, even though the
concentration of N atoms is the same.
°
C, the higher concentration of N atoms induces the thicker
compartment layers (Figure 3, parts c and d). At 1100 °C the
highly N-doped CNTs look like springs due to the well-curved
compartment layers at uniform distance (Figure 3f). As the N
concentration increases, the number of compartment layers
decreases while their thickness increases. The average number
per micrometer is ∼20 and ∼10 for 2.0- and 4.5%-doped CNTs
grown at 1100 °C, respectively.
In Figure 5b, the N content in the compartment layers is
estimated as 9% while that in the wall is 4%. The average value
is 6%, which is consistent with the XPS data. The π* peak of
C-K edge is stronger in the wall compared to the compartment
layers. Therefore the compartment layers consist of the less
crystalline graphitic sheets with the higher concentration of the
N atoms than the wall.
Figure 4a corresponds to the high-resolution TEM (HRTEM)
images for a 2.0% N-doped CNT grown at 1100 °C. The
curvature of the compartment layers is oriented to the tip (top
of the figure). The outer surface of the wall is not straight due
to the randomly stacked compartment layers. The atomic-
resolved image of the wall reveals the well-separated graphitic
sheets by a distance of 0.34 nm (Figure 4b). As the N content
increases to 6.0%, the compartment layers are smoothly curved
to connect with the wall without distortion (Figure 4c). Thus
the wall becomes straight and thicker. The graphitic sheets are
waved over a long range, showing the lower degree of crystalline
perfection than that of 2.0% N-doped CNTs (Figure 4d).
(5) The Influence of the N Doping on the Crystallinity of
Graphitic Sheets. Raman spectra taken from the 2.0, 3.3, and
4.7% N-doped CNTs grown using a CH /NH mixture at 1000
4
3
°C are displayed in Figure 6a. Each spectrum is the average of
many samples. All spectra show mainly two bands at ∼1350
-
1
-1
cm (D band) and ∼1576 cm (G band). The G band
originates from the Raman active E mode due to in-plane
2g
atomic displacements. The origin of D band has been explained
as disorder-induced features due to the finite particle size effect
3
3-35
or lattice distortion.
As the concentration of the N atoms
increases, the D band becomes stronger and broader. The linear
relation between the inverse of the in-plane crystallite dimension
and the intensity ratio of D band to G band (ID/IG) was noted.35
The value of ID/IG increases from 0.48 to 0.66 as the N content
increases from 2.0% to 4.7%. The value of ID/IG for the N-doped
CNTs grown at three temperatures is plotted in Figure 6b. The
data show that the degree of long-range ordered crystalline
perfection increases with the temperature, but decreases by
the N doping. The data of the N-doped CNTs grown using
C2H2/NH3 mixture are listed in Table 1. The value of ID/IG
usually decreases by about 0.2 for the increase of the N content
from 2 to 5%.
(
4) The N Distribution in the N-Doped CNTs. The EELS
data have been measured for the N-doped CNTs grown at 1100
C whose N contents are determined as 2% (Figure 5a) and
% (Figure 5b) using XPS. The EELS spectra are taken from
°
6
two regions; the edge part of the wall and the compartment
part. The spot size of the electron probe is about 2 nm. It shows
two distinct absorption features starting at 284 and 400 eV,
corresponding to the known K-shell ionization edges for carbon
(
C-K) and nitrogen (N-K), respectively. For C-K edge, a
defining π* peak at 300 eV corresponds to the graphitic carbons.

12962 J. Phys. Chem. B, Vol. 107, No. 47, 2003
Lee et al.
atoms, its contribution would be negligible compared to that of
C atom. The growth rate of CNTs can be thus determined mainly
by the bulk diffusion rate of C atoms.
The use of CH4 as the C source has been rarely reported for
the growth of the vertically aligned CNTs using thermal CVD.
It is related with the fact that the thermal decomposition of CH4
is not as efficient as that of other C sources such as C2H2 or
ethylene (C2H4). To our best knowledge, it is first time to report
the vertically aligned CNTs using thermal CVD of CH4 at
temperatures down to 900 °C. The growth rate of N-doped CNTs
at 1100 °C is essentially the same for both CH4/NH3 and
C2H2/NH3 mixtures. However, the growth rate of CNTs using
CH4/NH3 mixture is lower than that using C2H2/NH3 mixture.
To understand the effect of the C source on the growth rate,
we synthesized the vertically aligned CNTs on the Fe-deposited
Si substrates using C2H4 and ethane (C2H6) in the temperature
range 800-1100 °C. It shows the same growth rates as those
of CNTs grown using C2H2. The growth rate depending on the
C source will be rigorously discussed in a separated paper. Here,
we suggest that the decomposition of C2H2, C2H4, and C2H6 at
temperatures in the range 800-1100 °C would generate the
saturated C concentration in gas phase, leading the same growth
rate of CNTs. However, CH4 may provide the saturated
concentration only when the temperature is as high as 1100 °C.
In contrast to the insensitive growth rate on the N content,
the overall structure of N-doped CNTs depends on not only
the temperature but also the N content. The average diameter
and the diameter range usually increase with the temperature,
due to the more efficient agglomeration of the Fe catalytic
nanoparticles. The bamboo-like structure of the 2% N-doped
CNTs has been described thoroughly using the base growth
Figure 6. (a) Raman spectrum for 2.0, 3.3, and 4.7% N-doped CNTs
grown at 1000 °C. (b) Plot of the I /I value vs the N content and
growth temperature. The data in parentheses represent the N content
obtained from the XPS analysis.
D
G
4. Discussion
We previously reported the vertically aligned CNTs using
3
6-38
mechanism.
The growth process was separated into five
the thermal CVD of C2H2 in the temperature range 800-1100
steps: (i) the formation of the closed tip, (ii) the growth of the
wall, (iii) the production of the compartment layers, (iv) the
formation of the joint between compartment layers and wall,
and (v) the production of the bamboo-like structure owing to
the periodic compartment layers. In each step, the bulk diffusion
of C atom in the catalytic particles plays an important role.
36
°
C. Even if no NH3 flows during the growth, those CNTs were
doped with 2% N atoms as shown in Table 1. Under our
experimental condition, the growth of CNTs is possible only
when NH3 is used as a pretreatment gas. The role of NH3 is the
formation of Fe catalytic nanoparticles via etching. Ar flushes
at least 10 min before the flow of C2H2, to eliminate the residual
NH3 in the reactor. However, the N doping still occurs with
the same concentration, 2%. Therefore we suggest that the NH3
etching would inevitably dissolve the N atoms into the Fe
nanoparticles. These dissolved N atoms would precipitate with
the C atoms during the growth of CNTs, following the base
growth mechanism. We’ll discuss a detailed growth mechanism
in the following paragraph. We were able to increase the N
content of CNTs simply by flowing NH3 with C2H2 (or CH4)
during the growth. But it is not clear to explain why the N
content decreases at the higher flow rate of NH3. The maximum
N content in the present CNTs is about 6%, which is consistent
with the result of other groups that the N content usually did
not exceed 10%.
As the temperature increases, the surface and bulk diffusion
rates of C atoms can increase. If the bulk diffusion rate increases
more significantly compared to that of surface diffusion, the
growth rate of compartment layers would be greatly enhanced.
The compartment layers are therefore formed with a larger
thickness, connected with the wall more frequently.36 As the
growth rate of the compartment layers overwhelms that of the
wall, the compartment layers are eventually nested to construct
the tube. When the compartment layers depart from the catalytic
particle, the motive force could be the stress accumulated under
the graphitic cap formed on the catalytic nanoparticles. If the
degree of the crystalline perfection were high, the thick com-
partment layers would be curved under a larger strain and
seriously distort the wall structure.
The growth rate of N-doped CNTs is nearly the same for the
N content ranging from 2 to 5%. For the vertically aligned CNTs
As the N atoms incorporate into the graphitic sheets, the
(
doped with 2% N atoms) grown using C2H2, the growth rate
crystalline perfection of the graphitic sheets deteriorates. This
result is consistent with the work of other groups.1
9-32
The
as a function of temperature has been discussed in ref 36. The
Arrhenius plot of the growth rate yields an activation energy
that is almost the same as the diffusion energy of C atom in the
γ-phase Fe. Moreover, the growth of nanotubes has been
suitably explained by the assumption that the rate-determining
step is a diffusion-controlled first-order reaction.37 It implies
that the growth rate of nanotubes would be proportional to the
diffusion coefficient of C atom and the saturated C concentration
in the bulk. The diffusion of N atoms probably involves in the
growth of nanotubes. However, for the doping of 2-6% N
pyridine-typed N atoms would reduce the crystalline perfection
1
9
of the graphitic sheets. The less crystalline graphitic sheets
probably exhibit more flexibility. Then the compartment layers
are smoothly bent and the connection with the wall would take
place under less strain. After the joint, the graphitic sheets of
compartment layers and wall would grow at the same rate and
eventually depart from the catalytic particle. If the stress were
release, this joint growth would take longer before departing
from the catalytic particle, resulting in the increase of the wall

Vertically Aligned N-Doped Carbon Nanotubes
J. Phys. Chem. B, Vol. 107, No. 47, 2003 12963
thickness and less frequent formation of the thicker compart-
ment layers.
(4) Biercuk, M. J.; Llaguno, M. C.; Radosavljevic, M.; Hyun, J. K.;
Johnson, A. T.; Fischer, J. E. Appl. Phys. Lett. 2002, 80, 2767.
(5) Glatkowski, P.; Mack, P.; Conroy, J. L.; Piche, J. W.; Winsor, P.
The compartment layers of the highly doped CNTs exhibit
more N content than the wall. Han et al. also reported the higher
N concentration of the compartment layers compared to the
wall.20 The EELS data reveal that the degree of crystalline
perfection is always lower for the compartment layers compared
to that of the wall, even if the N concentration is the same for
both parts. It must be related to the fact that the graphitic sheets
of the compartment layers intrinsically contain more defects due
to the curved feature. In the case of the CNTs doped with the
higher N atoms, the N doping occurs preferentially in the defect
sites at the curved graphitic sheets of compartment layers. The
N-doped graphitic sheets of the compartment layers are thus
released from the strain when connecting with the wall. The
shape of compartment layers can be notably changed due to
the more significant doping of N atoms. We conclude that the
doping of N atoms plays a significant role in modifying the
structure of bamboo-like CNTs.
U.S. Patent 6,265,466, 2001.
(6) Liang, W.; Bockrath, M.; Bozovic, D.; Hafner, J. H.; Tinkham,
M.; Park, H. Nature 2001, 411, 665.
(7) Tans, S. J.; Verschueren, A. R. M.; Dekker, C. Nature 1998, 393,
49.
(8) Bachtold, A.; Hadley, P.; Nakanishi, T.; Dekker, C. Science 2001,
294, 1317.
(9) An, K. H.; Kim. W. S.; Park, Y. S.; Moon, J.-M.; Bae, D. J.; Lim,
S. C.; Lee, Y. S.; Lee, Y. H. AdV. Funct. Mater. 2001, 11, 387.
(10) Liu, C.; Fan, Y. Y.; Liu, M.; Cong, H. T.; Cheng, H. M.;
Dresselhaus, M. S. Science 1999, 286, 1127.
11) Bethune, D. S.; Kiang, C. H.; deVries, M. S.; Gorman, G.; Savoy,
R.; Vazquez, J.; Beyers, R. Nature 1993, 363, 605.
12) Journet, C.; Maser, W. K.; Bernier, P.; Loiseau, A.; Lamy de la
(
(
Chapelle, M.; Lefrant, S.; Deniard, P.; Lee, R.; Fischer, J. E. Nature 1997,
388, 756.
(13) Thess, A.; Lee, R.; Nikolaev, P.; Dai, H.; Petit, P.; Robert, J.; Xu,
C.; Lee, Y. H.; Kim, S. G.; Rinzler, A. G.; Colbert, D. T.; Scuseria, G. E.;
Tomanek, D.; Fisher, J. E.; Smalley, R. E. Science 1996, 273, 483.
(14) Terrones, M.; Grobert, N.; Olivares, J.; Zhang, J. P.; Terrones, H.;
Kordatos, K.; Hsu, W. K.; Hare, J. P.; Townsend, P. D.; Prassides, K.;
Cheetham, A. K.; Kroto, H. W.; Walton, D. R. M. Nature 1997, 388, 52.
(
15) Ren, Z. F.; Huang, Z. P.; Xu, J. W.; Wang, J. H.; Bush, P.; Siegal,
M. P.; Provencio, P. N. Science 1998, 282, 1105.
16) Fan, S.; Chapline, M. G.; Franklin, N. R.; Tombler, T. W.; Cassell,
5
. Conclusion
(
The vertically well-aligned N-doped CNTs were grown on
the Fe nanoparticles deposited Si substrates using thermal CVD
of CH4/NH3 and C2H2/NH3 mixtures in the temperature range
A. M.; Dai, H. Science 1999, 283, 512.
(17) Terrones, M.; Ajayan, P. M.; Banhart, F.; Blase, X.; Carroll, D.
L.; Charlier, J. C.; Crzerw, R.; Foley, B.; Grobert, N.; Kamalakaran, R.;
Kohler-Redlich, P.; R u¨ hle, M.; Seeger, T.; Terrones, H. Appl. Phys. A 2002,
9
00-1100 °C. The N content in the range 2-6% has been
74, 355.
controlled by the flow rate of NH3. The growth rate is almost
independent of the N content. However, the N doping modifies
remarkably the structure and the degree of crystallinity of CNTs.
The bamboo-like structured CNTs are exclusively produced. As
the N content increases, the more curved and thicker compart-
ment layers appear more regularly at the longer distance. The
HRTEM images and Raman spectra reveal consistently that as
the N content increases the degree of crystalline perfection
decreases. The EELS data indicate the higher N % in the less
crystalline compartment layers compared to the wall.
(
18) Terrones, M.; Redlich, P.; Grobert, N.; Trasobares, S.; Hsu, W.
K.; Terrones, H.; Zhu, Y. Q.; Hare, J. P.; Reeves, C. L.; Cheetham, A. K.;
R u¨ hle, M.; Kroto, H. W.; Walton, D. R. M. AdV. Mater. 1999, 11, 655.
(19) Terrones, M.; Terrones, H.; Grobert, N.; Hsu, W. K.; Zhu, Y. Q.;
Hare, J. P.; Kroto, H. W.; Walton, D. R. M.; Redlich, P.; R u¨ hle, M.; Zhang,
J. P.; Cheetham, A. K. Appl. Phys. Lett.1999, 75, 3932.
(20) Han, W. Q.; Redlich, P.; Seeger, T.; Ernst, F.; R u¨ hle, M.; Grobert,
N.; Hsu, W. K.; Chang, B. H.; Zhu, Y. Q.; Kroto, H. W.; Walton, D. R.
M.; Terrones, M.; Terrones, H. Appl. Phys. Lett. 2000, 77, 1807.
(
21) Ma, X.; Wang, E.; Zhou, W.; Jefferson, D. A.; Chen, J.; Deng, S.;
Xu, N.; Yuan. J. Appl. Phys. Lett. 1999, 75, 3105.
22) Zhong, D.; Liu, S.; Zhang, G.; Wang, E. G. J. Appl. Phys. 2001,
9, 5939.
(
8
The dependence of the growth rate, the structure, and the
crystallinity on the N content has been explained by adopting
the base growth mechanism of CNTs. We suggest that the
growth rate would be mainly determined by the bulk diffusion
rate of C atoms. The N doping would enhance the flexibility of
the graphitic sheets. Then the compartment layers are bent and
connect with the wall under less strain. The joint growth of the
wall and compartment layers would also take longer due to the
reduced strain. The present result shows that the N doping is
one of promising ways to control the crystallinity and structure
of nanotubes.
(
23) Ma, X.; Wang, E. G. Appl. Phys. Lett. 2001, 78, 978.
(24) Wang, E. G. J. Am. Ceram. Soc. 2002, 85, 105.
(
25) Nath, M.; Satishkumar, B. C.; Govindaraj, A.; Vinod, C. P.; Rao,
C. N. R. Chem. Phys. Lett. 2000, 322, 333.
26) Sen, R.; Satishkumar, B. C.; Govindaraj, A.; Harikumar, K. R.;
(
Raina, G.; Zhang, J. P.; Cheetham, A. K.; Rao, C. N. R. Chem. Phys. Lett.
1998, 287, 671.
(
27) Yudasaka, M.; Kikuchi, R.; Ohki, Y.; Yoshimura, S. Carbon 1997,
5, 195.
28) Suenaga, K.; Yudasaka, M.; Colliex, C.; Iijima, S. Chem. Phys.
Lett. 2000, 316, 365.
29) Sung, S. L.; Tsai, S. H.; Tseng, C. H.; Chiang, F. K.; Liu, X. W.;
Shin, H. C. Appl. Phys. Lett. 1999, 74, 197.
30) Kurt, R.; Klinke, C.; Bonard, J. M.; Kern, K.; Karimi, A. Carbon
3
(
(
(
2
001, 39, 2163.
Acknowledgment. This work was supported by Direct Basic
Research Program (Korea Science and Engineering foundation
Project No. R04-2002-000-20088-012003). SEM analyses were
partly performed at the Korea Basic Science Institute in Seoul.
(31) Lee, C. J.; Lyu, S. C.; Kim, H. W.; Lee, J. H.; Cho, K. I. Chem.
Phys. Lett. 2002, 359, 115.
32) Wang, X.; Liu, Y.; Zhu, D.; Zhang, L.; Ma, H.; Yao, N.; Zhang,
B. J. Phys. Chem. B 2002, 106, 2186.
(
(33) Tuinstra, F.; Koenig, J. L. J. Chem. Phys. 1970, 53, 1126.
(
34) McCulloch, D. G.; Prawer, S.; Hoffman, A. Phys. ReV. B 1994,
0, 5905.
35) Wilhelm, H.; Lelaurain, M.; McRae, E.; Humbert, B. J. Appl. Phys.
5
References and Notes
(
(1) de Heer, W. A.; Ch aˆ telain, A.; Ugarte, D. Science 1995, 270, 1179.
1998, 84, 6552.
(
2) Rinzler, A. G.; Hafner, J. H.; Nikolaev, P.; Lou, L.; Kim, S. G.;
(36) Lee, Y. T.; Park, J.; Choi, Y. S.; Ryu, H.; Lee, H. J. J. Phys. Chem.
B 2002, 106, 7614.
(37) Kim, N. S.; Lee, Y. T.; Park, J.; Ryu, H.; Lee, H. J.; Choi, S. Y.;
Choo, J. J. Phys. Chem. B 2002, 106, 9286.
(38) Lee, C. J.; Park, J. J. Phys. Chem. B 2001, 105, 2365.
Tomanek, D.; Nordlander, P.; Colbert, D. T.; Smalley, R. E. Science 1995,
69, 1550.
3) Rosen, R.; Simendinger, W.; Debbault, C.; Shimoda, H.; Fleming,
L.; Stoner, B.; Zhou, O. Appl. Phys. Lett. 2000, 76, 1668.
2
(