See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/6183760

FTIR Spectroscopy of Multiwalled Carbon Nanotubes: A Simple

Approachto Study the Nitrogen Doping

Article in Journal of Nanoscience and Nanotechnology · July 2007

DOI: 10.1166/jnn.2007.723 · Source: PubMed

CITATIONS READS

94 11,414

4 authors, including:

Pawan K Tyagi Padmnabh Rai

Delhi Technological University UM-DAE Centre for Excellence in Basic Sciences
76 PUBLICATIONS 1,219 CITATIONS 34 PUBLICATIONS 577 CITATIONS

SEE PROFILE SEE PROFILE

Devi Misra
Indian Institute of Technology Bombay and IIa Technology Singapore
125 PUBLICATIONS 2,878 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Nanoscale Electronics & Photonics View project

conjugate heat transfer within cavity View project

All content following this page was uploaded by Padmnabh Rai on 26 May 2014.

The user has requested enhancement of the downloaded file.

 Copyright © 2007 American Scientific Publishers Journal of
All rights reserved Nanoscience and Nanotechnology
Printed in the United States of America Vol. 7, 1820–1823, 2007

FTIR Spectroscopy of Multiwalled Carbon Nanotubes:

A Simple Approach to Study the Nitrogen Doping
Abha Misra, Pawan K. Tyagi, Padmnabh Rai, and D. S. Misra∗
Department of Physics, Indian Institute of Technology Bombay, Mumbai 400076, India

The nitrogen doped multiwalled carbon nanotubes (MWNTs) were synthesized by microwave
plasma chemical vapor deposition (MPCVD) technique. In this paper, we report the results of FTIR,
Raman, and TGA studies to confirm the presence of N-doping inside carbon nanotubes. Fourier
transform infrared (FTIR) studies were carried out in the range 400–4000 cm−1 to study the attach-
ment of nitrogen impurities on carbon nanotubes. FTIR spectra of the virgin sample of MWNTs show
dominant peaks which are corresponding to Si–O,
Delivered byC–N, N–CHto:
Ingenta 3 , CNT, C–O, and C–Hx , respectively.
The Si–O peak has its origin inRensselaer
silicon substrate
Polytechnic Institutepeaks are due to the precursor
whereas the other
gases present in the gas mixture. The peaks are sharp and highly intense showing the chemisorp-
IP : 128.113.111.89
tion nature of the dipole bond. The intensity of the peaks due to N–CH3 , C–N, and C–H reduces
Tue, 17 Jul 2007 14:18:21
after annealing. It is interesting to note that these peaks vanish on annealing at high temperature

(900 C). The presence of C–N peak may imply the doping of the MWNTs with N in substitution
mode. The position of this intense peak is in agreement with the reported peak in carbon nitride
samples prepared by plasma CVD process, since the Raman modes are also expected to be delo-
calized over both carbon and nitrogen sites it was found that the intensity ratio of the D and G
peaks, I(D)/I(G), varies as a function of ammonia concentration. The TGA measurements, carried
out under argon flow, show that the dominant weight loss of the sample occurs in the temperature
RESEARCH ARTICLE

range 400–600  C corresponding to the removal of the impurities and amorphous carbon.
Keywords: MWNTs, FTIR, Raman Spectroscopy, TGA/DTA.

1. INTRODUCTION Despite the possibility of various synthesis routes, the

The presence of defects and impurities that are electron-
ically and chemically active can change effectively the
properties of carbon nanotubes (CNTs). Nitrogen (N) dop-
ing into carbon layers is one of the effective approaches to
modify these properties, as the controlled electrical proper-
ties are very important in electronic devices at nanoscale.
There are several recent reports1–4 on the synthesis of
N-doped CNTs using different approaches. For exam-
ple Terrones et al.5 prepared N-doped carbon nanotubes
by pyrolyzing various C/N sources including triazine,
melamine, ferrocene, melamine mixture, and ferrocene/C60
mixtures in ammonia atmosphere. Ma et al.6 obtained N-
doped carbon nanofibers on large scale using microwave
plasma-assisted CVD with a mixture of methane and nitro-
gen. Han et al.7 concluded that pyrolysis in the presence of
NH3 proves an efficient route to achieve N-doped carbon
nanotubes. From these results, it is clear that the presence
of N2 during the CNTs growth process is proven to be an
efficient method of N doping.8
∗
Author to whom correspondence should be addressed.
mechanism of the N-doping inside carbon layers is still
an open question. Many groups have proposed that the
electronic structure of N doped CNTs are as pyridine-
like or graphite like with triple-bonded CN.9–11 Another
group has proposed that N atoms might exist mainly
inside the inner core of CNTs.12 In this paper, we inves-
tigate the nitrogen doped multiwalled carbon nanotubes
(MWNTs) using Fourier transform infrared (FTIR) spec-
troscopy and Raman spectroscopy supported by thermo
gravimetric analysis (TGA).
2. EXPERIMENTAL DETAILS
MWNTs encapsulated with nickel nanorods used in this
study were synthesized by microwave plasma chemical
vapor deposition (MPCVD) technique, the details of which
are reported elsewhere.13 The nickel layers having a thick-
ness of 5–10
m were treated in the plasma of NH3 gas
for 10 seconds in the temperature range 300  C–750  C
at 10 torr for the formation of nano-sized nickel catalyst,
which was the precursor for subsequent CNTs growth
using chemical vapor deposition (CVD). The flow rate

1820 J. Nanosci. Nanotechnol. 2007, Vol. 7, No. 6 1533-4880/2007/7/1820/004 doi:10.1166/jnn.2007.723

Misra et al. FTIR Spectroscopy of Multiwalled Carbon Nanotubes: A Simple Approach to Study the Nitrogen Doping

of ammonia ranged from 180–250 sccm. Methane (CH4 )

and hydrogen (H2 ) were used with respective flow rates
of 6 sccm and 40 sccm. The temperature of the substrate
was maintained at 820  C ± 20  C at deposition pressure
40 torr. The MWNTs were treated with isopropyl alco-
hol in an ultrasonic bath and were dispersed onto a car-
bon coated copper micro grid for transmission electron
microscopy (TEM). The lattice imaging of the walls of
CNTs as well as of the nickel nanorods is recorded using
the transmission electron microscopy (TEM) in a JEOL
TEM 2010 (UHR model) operated at 200 kV.
The FTIR spectra were recorded on Nicolet Magna 550
FT-IR Spectrometer. The samples for FTIR studies were
prepared by suspending approximately 6 mg of MWNTs
materials in ∼15 ml isopropyl alcohol by sonication with
an ultra sonic probe for several minutes. One drop of this
solution was sprayed onto silicon wafer and a uniform thin
MWNT film on the IR transparent silicon substrate was
thus obtained. FTIR studies were carried out in the range
Delivered by Ingenta to:
400–4000 cm−1 in the absorbance mode. The FTIR results
Rensselaer
along with the TGA give a reasonably good picture of the
Polytechnic Institute
IP : 128.113.111.89
nitrogen attachment on the carbon nanotubes. Raman spec-
tra were recorded in backscattering geometry Tue, 17 Jul
for 514.5 nm 2007 14:18:21
+
excitation from an Ar ion laser using Jobin-Yvon HR800.
Thermal Gravimetric Analysis (TGA) were carried out on
Mettler-Toledo (TGA/SDTA851) thermal analyzer under
argon flow at the heating rate of 5  C/min to obtain infor-
mation on the decomposition and the burning properties of
carbon nanotubes and impurities present in it. The temper-

ature of the sample was varied from room temperature to

900 C. The bonding was studied by infrared and Raman
spectroscopy.
3. RESULTS AND DISCUSSION
Figure 1(a)–(b) shows the SEM and TEM micrographs of
high density MWNTs grown on nickel coated copper sub-
strate after ammonia plasma treatment, respectively. The
average length and diameter of the tubes are 100 micron
and 20 nm, respectively.
Using NH3 gas is an efficient way to incorporate nitro-
RESEARCH ARTICLE
Fig. 1. (a) SEM micrograph of a sample of multi-walled carbon nano-
tubes on a copper substrate. The high density of tubes is noteworthy.
(b) TEM image of the MWNTs.
Out of the above identified bonds, the presence of C–N and
N–CH3 bonds at 1250 and 1372 cm−1 is most interesting.
The strong peaks at 1250 and 1372 cm−1 are consistent
with C–N and N–CH3 stretching vibrations attributed by
(d)
(c)
Absorbance (arb. units)

gen in carbon nanotubes lattice during growth. NH3 gas

not only plays role in MWNTs growth but also increases (b)

nitrogen concentration in the carbon nanotubes. FTIR is CNT

C-N
used to characterize the functional elements absorbed by N-CH3
C-Hx
carbon nanotubes. Figures 2(a–d) shows the FTIR spec- Si-O (a)
−1
tra in the range 400–4000 cm (a) of as-grown tubes,
(b) of the tubes collected after TGA at 400  C, (c) of the CNT
tubes annealed at 600  C, and (d) of the tubes collected
after TGA at 900  C. FTIR of (a) shows dominant peaks
at 1026, 1250, 1372, 1445, 1736, 2362, 2851, 2925 cm−1
which corresponds to Si–O, C–N, N–CH3 , CNT, C–O, and 500 1000 1500 2000 2500 3000 3500
C–Hx , respectively. The infra-red absorbance at 1026 cm−1 –1
consistent with Si–O stretching vibrations due to slightly Wave number (cm )
different concentrations in the native oxide layer of the Fig. 2. FTIR spectra of (a) as grown MWNTs, (b) after TGA at 400  C,
silicon before and after coating with MWNTs film.14 (c) after annealing at 600  C, (d) after TGA at 900  C.

J. Nanosci. Nanotechnol. 7, 1820–1823, 2007 1821

 FTIR Spectroscopy of Multiwalled Carbon Nanotubes: A Simple Approach to Study the Nitrogen Doping Misra et al.

Choi et al.12 to the presence of intercalated N atoms Table I. The peak position, full width at half maxima (FWHM) and
relative intensities I(D)/I(G) of D and G bands with increasing N
between the graphite layers at the inner part of the nano-
concentration.
tube walls. The intercalated nitrogen atoms in between the
graphite walls, however, may not be strongly IR active. D band G band
It has been pointed out9 15–17 that substituting an N atom Position FWHM Position FWHM
in place of a C atom in a sp2 bonded carbon network will No. (cm−1 ) (cm−1 ) (cm−1 ) (cm−1 ) I(D)/I(G)
induce strong IR activity due to a net dipole moment; con-
1 1342 4935 1569 34 0.42
sequently the absorption in the 1200–1600 cm−1 region is 2 1347 5877 1576 51 0.55
expected if the N atoms are bonded into the carbon net- 3 1347 617 1578 42 1.32
work. This suggests that the N-doping in carbon nanotubes
can take place in two ways; either N can substitute for
carbon in benzene ring which is pyridine like ring or N full width at half maxima (FWHM) and relative intensities
can be located between graphite planes which give rise to I(D)/I(G) of D and G bands with increasing N concen-
a C–N  bond or both. The half widths of the C–N and tration are shown in Table I. Since N doping is very
−1
N–CH3 peaks are 36.72 and 18.53 cm , respectively, and effective in introducing defects inside graphite network,
suggest that the chemisorptions process may be dominant results upshift of both D and G band with the N concen-
implying chemical bonding between carbon and nitrogen tration. The substantial shifts of Raman bands give clear
atoms. From their relative intensities we therefore strongly evidence of modification of phonon and electron struc-
believe that N doping in graphene sheets of theDelivered tures of carbon layers due to N doping. The FWHM of the
inner walls by Ingenta to:
G band has increased with N concentration as shown in
is more favorable than N cross-linking between graphitic
Rensselaer Polytechnic Institute
planes of outer planes. Features at 1445 and 1736 −1 Table I. However the presence of a relatively high inten-
IP : cm
128.113.111.89
sity G band provides evidence for high crystalline behav-
may be attributed to MWNT vibrational modes Tue,are
17 also
Jul 2007 14:18:21
18 −1
apparent. The peak at 2362 cm corresponds to the ior in the CNTs. The increase in the I(D)/I(G) ratio as
C–O bonds and the features between 2851–2925 cm−1 are a function of N concentration is attributed to bong angle
consistent with C–Hx stretching vibrations of chemisorbed and bond length disorder induced by introduction of N.
hydrogen of various types presents in all carbon A small peak at 1120 cm−1 may corresponds to single
films.19 20 C–N bonds arises due to N bonding to outer graphitic
Raman spectra as a function of NH3 concentration i.e., layers.21 22
N concentration are presented in Figure 3. All these spec- The intensity of the peaks due to N–CH3 , C–N, and C–H
RESEARCH ARTICLE

tra are characterized by mainly two bands D band/G band.
The G band is assigned to E2g symmetric vibrational mode
of graphite layers and the D band is originated from the
level of disordered carbons. Both the bands shift towards
higher wave number as we increased the NH3 concen-
tration from 180 sccm to 250 sccm. The peak position,
are gradually suppressed, after annealing at 400  C. The
suppression of the peaks would be related to the gradual
removal of nitrogen and hydrogen bonded species as impu-
rity. The bonding of N with carbon atom in graphene sheet
will give rise to a defect in structure, which would get
annealed, at high temperature. This results in the reduction
of the intensity of the IR peaks with annealing. Further
annealing of the sample at 600  C does not induce any

120
TGA
Raman intensity (A.U.)

100

80
Weight percent

60
3
40 DTA
2

1 20
D G

1100 1200 1300 1400 1500 1600

– 20
Wave number (cm–1) – 200 0 200 400 600 800 1000 1200
Temperature (˚C)
Fig. 3. Raman spectra of MWNTs samples prepared at different NH3
concentration. Fig. 4. TGA/DTA results for MWNTs.

1822 J. Nanosci. Nanotechnol. 7, 1820–1823, 2007

 Misra et al. FTIR Spectroscopy of Multiwalled Carbon Nanotubes: A Simple Approach to Study the Nitrogen Doping

appreciable change. It is interesting to note that these References and Notes

peaks vanish on annealing at higher temperature (900  C)
1. Li Fanxing, Yao Wang, Dezheng Wang, and Fei Wei, Carbon 42,
implying that at high temperature the defects inside the 2375 (2004).
tubes are getting strongly mobile leading to the destruc- 2. J. Liuiwen, Scott Webster, and David L. Carroll, J. Phys. Chem. B
tion of the tubes as well as the attachments along with 109, 15769 (2005).
the tubes. 3. Maoshuai He, Shuang Zhou, Jin Zhang, Zhongfan Liu, and Colin
Robinson, J. Phys. Chem. B 109, 9275 (2005).
TGA/DTA curve in Figure 4 reveals a small weight loss
 4. Yang Zhuxian, Yongde Xia, and Robert Mokaya, Chem. Mater. 17,
due to water removal around 80 C. The dominant weight 4502 (2005).
loss steps are due to the removal of carbon materials and 5. M. Terrones, H. Terrones, N. Grobert, and D. R. M. Walton, Appl.
also due to the decomposition of nanotubes23 25 which Phys. Lett. 75, 3932 (1999).
take place in the temperature range 400–600  C. The 6. X. Ma, E. Wang, W. Zhou, D. A. Jefferson, and Yuan, Appl. Phys.
Lett. 75, 3105 (1999).
weight loss starts near 400  C and the process completes
 7. W.-Q. Han, P. Kohler-Radlich, T. Seeger, M. Terrones, and
by 600 C. This result agrees with the FTIR spectra H. Terrones, Appl. Phys. Lett. 77, 1807 (2000).
Figures 2(b–c) discussed above, where we have shown sig- 8. Maoshuai He, Shuang Zhou, and Jin Zhang, J. Phys. Chem. B 109,
nificant decrease in the intensity of IR peaks upon anneal- 9275 (2005).
ing to 400  C. The MWNTs are completely destroyed at 9. J. H. Kaufman, S. Metin, and D. D. Saperstein, Phys. Rev. B 39,
13053 (1989).
900  C, suggesting that at this temperature amorphous car- 10. S. E. Rodil, A. C. Ferrari, J. Robertson, and W. I. Milne, J. Appl.
bon as well as MWNTs converts to gaseous form. The Phys. 89, 5425 (2001).
complete removal of all the dominant peaks from the by Ingenta
Delivered 11. A. C. Ferrari,
to: S. E. Rodil, and J. Robertson, Phys. Rev. B 67, 155306
FTIR spectra (Fig. 2(d)) confirms the burning of MWNTs
Rensselaer Polytechnic (2003).
Institute
at 900  C resulting in the change of the color 12. Hyun Chul Choi, Seung Yong Bae, and Jeunghee Park, Appl. Phys.
IP of the
: 128.113.111.89
Lett. 85, 5742 (2004).
sample. Tue, 17 Jul 2007 13. 14:18:21
Pawan K. Tyagi, M. K. Singh, Abha Misra, Umesh Palnitkar, D. S.
Misra, E. Titus, N. Ali, G. Cabral, J. Gracio, and M. Roy, Thin Solid
Films 469, 127 (2004).
4. CONCLUSIONS 14. S. L. Sung, C. H. Tseng, F. K. Chiang, X. J. Guo, and X. W. Liu,
Thin Solid Films 340, 169 (1999).
The results employing FTIR have revealed nitrogen attach- 15. M. R. Wixom, J. Am. Ceram. Soc. 73, 1973 (1990).
ment to MWNTs as is evident from strong and intense 16. Y. K. Yap, S. Kida, T. Aoyama, Y. Mori, and T. Sasaki, Appl. Phys.
IR peaks at 1250 and 1372 cm−1 . From the intensity Lett. 73, 915 (1998).

17.
and halfwidth of these peaks we conclude that N atom
is incorporated at substitutional sites in the carbon net- 18.
work. Raman spectra as a function of NH3 concentration
i.e., N concentration have been presented. The intensity 19.
ratios (I(D)/I(G)), band positions and bandwidths of D
20.
and G bands are found different for different NH3 con- 21.
centration. TGA/DTA has been used to study the weight
loss of carbon nanotube as a function of temperature. 22.
We found that the dominant weight loss occurs between 23.
400  C–600  C ranges and agrees well with disappearance
24.
of IR peaks corresponding to N-attachment. TGA can be
used to get the information about the weight loss of the 25.
sample.
RESEARCH ARTICLE
S. H. Lai, Y. L. Chen, L. H. Chan, Y. M. Pan, X. W. Liu, and H. C.
Shih, Thin Solid Films 444, 38 (2003).
A. C. Dillon, T. Gennett, J. L. Alleman, K. M. Jones, Proceedings
of the 2000 DOE/NREL Hydrogen Program Review (2000).
J. Ristein, R. T. Stief, L. Ley, and W. Beyer, J. Appl. Phys. 84, 3836
(1998).
G.-Q. Yu et al., Diam. Rel. Mater. 11, 1633 (2002).
J. M. Mendez, A Gaona-Couto, S. Muhl, and S. Jimenez-Sandoval,
J. Phys.: Condens Matter 11, 5225 (1999).
E. B. Wilson, Phys. Rev. 45, 706 (1973).
D. G. McCulloch, E. G. Gerstner, D. R. McKenzie, S. Prawer, and
R. Lalish, Phys. Rev. B 52, 850 (1995).
M. Zhang, M. Yudasaka, and S. Iijima, Chem. Phys. Lett. 364, 42
(2002).
M. Zhang, M. Yudasaka, S. Bandow, and S. Iijima, Chem. Phys.
Lett. 369, 680 (2003).

Received: 16 March 2006. Revised/Accepted: 16 June 2006.

J. Nanosci. Nanotechnol. 7, 1820–1823, 2007 1823

View publication stats