Sei sulla pagina 1di 6

Fibers and Polymers 2008, Vol.9, No.

4, 410-415

Improved Dispersion of Carbon Nanotubes in Chitosan

Sukrut Ozarkar, Manjeet Jassal*, and Ashwini K. Agrawal**
Department of Textile Technology, Indian Institute of Technology, Hauz Khas, New Delhi-110016, India
(Received November 16, 2007; Revised March 26, 2008; Accepted April 9, 2008)

Abstract: Carbon nanotubes (CNT)/chitosan films and fibers can find use in specialized applications like the artificial mus-
cles and other intelligent switching devices. The dispersion state of the single walled carbon nanotubes (SWCNTs) in chito-
san matrix plays a major role in deciding the ultimate properties of composite. A suitable chemical treatment for purification
and functionalization of SWCNTs is reported. Optimal conditions to prepare water soluble and stable, dispersion of SWCNT
in chitosan are presented. The dispersion behavior of purified and functionalized SWCNT was characterized by visual obser-
vations, transmission electron microscopy (TEM), and Raman spectroscopy. The dispersion obtained using functionalized
SWCNT was stable, while the purified SWCNT dispersion showed limited stability. The better stability of functionalized
SWCNT dispersion in chitosan was evidenced by improved interaction between chitosan and carboxyl functional groups of
Keywords: SWCNT, Dispersion, Chitosan, Carbon nanotube, Raman spectra

Introduction New methods of chemical functionalization have been

developed so as to enhance the dispersion of SWCNTs in the
Single-wall carbon nanotubes (SWCNTs) have been of aqueous medium [9-17]. These functionalized CNTs showed
interest since their discovery because of their unique improved solubility/dispersion in water and various common
architecture and remarkable mechanical, thermal, and organic solvents/polymer matrices [18-21].
electrical properties. Carbon nanotubes have the potential In this manuscript, we report a comparison of the dispersion
applications in a broad range of technologies, in different behavior of SWCNTs and carboxyl functionalized SWCNTs
fields including engineering, biology, chemistry, medicine, in chitosan solution for the potential application of SWCNT/
electronics, and material science [1]. As fillers in polymer chitosan composites in the area smart materials like the bio-
matrix materials, SWCNTs promise an extraordinary opportunity actuators, where the carbon nanotubes provide a conducting
for development of new products with desirable physical reinforcement for the biocompatible chitosan matrix. The
properties [2]. For example, the addition of well dispersed dispersion studies using TEM and Raman spectroscopy
SWCNT in an insulating polymer matrix results in a revealed a better dispersion of functionalized SWCNT in
conductive network that forms a polymer composite with chitosan in comparison to non-functionalized SWCNT.
dramatic improvement in electrical conductivity with low
percolation threshold (0.05-0.3 % of SWCNT loading) [3]. Experimental
The critical challenges related to achieving these superior
properties lie in uniformly dispersing the nanotubes in the Materials
required polymer matrix. The carbon nanotubes tend to The SWCNTs used in this experiment were supplied by
aggregate by virtue of very strong van der Waal’s forces of Carbolex Inc. Their average diameter was 4 nm and purity
attraction. Their inherent chemical inertness makes it difficult to was 50 %. Chitosan polymer (degree of deacetylation~85 %
disperse them in any solvent or polymer matrices [4-7]. and molecular weight~5×105) was obtained from Sigma-
Another obstacle in dispersing the carbon nanotubes is the Aldrich. All other chemicals used were of high purity and
presence of various impurities including amorphous carbon used without further purification.
and other metal catalyst particles. These impurities are
responsible for the poor properties of carbon nanotube Purification of Carbon Nanotubes
reinforced composites [8]. In recent years, research has been The SWCNTs were purified using the following method.
focused on achieving the highest possible degree of dispersion The SWCNT (2 mg) was Soxhlet extracted with 50 ml of
without altering the properties of the CNTs themselves. toluene for 6 h. After Soxhlet extraction, the nanotubes were
Ideally, individual nanotubes must be separated from the rinsed with acetone and finally dried in the air oven at
bundles to exploit the intrinsic properties of SWCNTs. Among 100 oC. The Soxhlet extracted and dried nanotubes were
various approaches reported for dispersing CNTs in aqueous oxidized with 10 % hydrogen peroxide (H2O2) solution
medium, the use of surfactants is most popular and effective. under reflux for 1 h [8]. The sample was then filtered,
washed with deionized water, followed by acetone, and
**Corresponding author: dried in oven at 100 oC.
**Corresponding author:

Improved Dispersion of Carbon Nanotubes in Chitosan Fibers and Polymers 2008, Vol.9, No.4 411

Functionalization of Carbon Nanotubes Results and Discussion

The purified SWCNTs (30 mg) were treated with 70 %
nitric acid (10 ml) in an ultrasonic bath for about 30 min at Purification and Functionalization of Carbon Nanotubes
room temperature (30 oC) and the dispersion was allowed to Different CNT synthesis methods are known to yield
react for about 4 h [8]. This dispersion was then diluted with CNTs with different types and amounts of impurities, (e.g.
deionized water to approximately 20 times its initial volume. amorphous carbon and catalyst particles). Several strategies
The treated SWCNTs were vacuum filtered using an have already been developed over the past decade to purify
ultrafiltration membrane and washed repeatedly with deionized CNTs without significantly damaging the structure of nanotubes.
water until the washings showed no acidity. The filtered solid The optimum conditions reported in the literature [8] were
(SWCNT-COOH) was dried under vacuum for 24 h at 40 oC. used. After Soxhlet extraction of SWCNTs with toluene, a
weight loss of approximately 4 % was observed. The SWCNTs
Characterization of the SWCNT Samples were then treated with hydrogen peroxide solution for
The purified and functionalized SWCNTs were characterized removal of amorphous carbon. This treatment removes the
by Fourier transform infrared spectroscopy (FTIR) and carbonaceous particles that are shelled off from the SWCNTs.
thermogravimetric analysis (TGA). The FTIR measurements The reflux treatment with nitric acid has been reported for
were made on Perkin Elmer FTIR equipment (Spectrum oxidative functionalization of SWCNTs. However, such
BX). The dried samples were ground with potassium bromide severe treatment conditions are also reported to damage the
and then pressed into a pellet. A standard method (ASTM D- structure of nanotube walls and reduce their aspect ratio.
974) was followed to determine the number of carboxyl Therefore, the functionalization conditions were optimized
groups produced on the carbon nanotubes surface. Precisely by varying the temperature and time of treatment with nitric
weighed quantity of functionalized carbon nanotubes was acid solution. The treatment of purified carbon nanotubes
dispersed in water. The dispersion was titrated against standard with nitric acid at room temperature (~30 oC) for 4 h was
sodium hydroxide solution and the extent of functionalization found to result in optimum degree of carboxylation. The
was determined in terms of number of acid groups from the heavy metal catalyst particles present in the carbon
following relation. nanotubes are also removed during this oxidation process.
Further, the oxidation of the SWCNTs occurs at the defect
No of acid groups=
sites, known as the Stone-Wales (SW) defect on the carbon
[(titration reading−blank titration)×molarity of aq NaOH]/
nanotubes [11,22,23]. The SW defect is a typical topological
weight in g of SWCNT
defect in the nanotube and is comprised of two pairs of five-
TGA data was acquired on Perkin-Elmer (TGA7) thermal membered and seven-membered rings. It has been suggested
analyzer at a heating rate of 20 oC min−1 under nitrogen that SW defects could act as nucleation centers for the
atmosphere up to 800 ºC. formation of dislocations in the originally ideal graphite
network and constitute the onset of possible plastic
Dispersion Studies of Carbon Nanotubes deformation of a carbon nanotube. With a local deformation,
Transmission electron microscope (TEM) of Philips CM12 the nanotube containing SW defects might be more favorable
was used to study the stability of SWCNT dispersion in for subsequent reactions (shown in Figure 1).
ethanol and in chitosan solution. The purified and purified- Figure 2 shows FTIR spectra of pristine SWCNT and
functionalized SWCNTs were first dispersed in ethanol to functionalized SWCNT. The evidence for the purification
the extent of 1.2 mg in 20 ml of ethanol, which is a standard and the functionalization of the SWCNTs can be clearly
method for observing carbon nanotubes under TEM [8]. The observed from the FTIR spectra. In SWCNTs, the band at
dispersion was ultrasonicated for 15 min at 35 KHz and ~1600 cm−1 is assigned to the C=C stretching mode associated
10 % power. One drop of dispersion was placed on the 300 with SWCNT sidewall defects. The FTIR spectrum of the
mesh copper coated grid and imaged under various magnifi- functionalized sample shows three new peaks in comparison
cations of the electron beam. Similarly, the SWCNT chitosan to the as obtained nanotubes. The peak at 1737 cm−1 is clearly
dispersion was prepared by dispersing 1.2 mg SWCNT in
0.25 % chitosan solution and ultrasonication for 15 min and
was observed in electron beam.
The Raman measurements were carried out using RENISHAW
Raman system equipped with 785 nm excitation (Argon ion
LASER and power of 15 mW). For recording Raman spectra,
the dispersion was filled into 2 mm diameter quartz capillaries
and the capillaries were sealed with the help of paraffin wax
(to prevent evaporation of water). The samples were scanned
between 100 cm−1 to 3200 cm−1. Figure 1. Reaction scheme for oxidation of carbon nanotubes.
412 Fibers and Polymers 2008, Vol.9, No.4 Sukrut Ozarkar et al.

range of 400-600 oC. This observation is in agreement with

the reported literature for the carbonaceous materials. After
purification and functionalization of SWCNT, a weight loss
of about 10 % was observed. This may be attributed to
incomplete removal of carbonaceous matter [24].

Dispersion Studies of Carbon Nanotubes

The pristine, purified, and purified-functionalized carbon
nanotubes were dispersed in ethanol to the extent of 1 mg/
10 ml ethanol and ultrasonicated at 35 KHz for 15 min. As
expected, the pristine SWCNT could not be dispersed even
for a few minutes. It can be observed from the digital
photographs shown in Figure 4, that both the purified and
purified-functionalized carbon nanotubes dispersed equally
well during ultrasonication. However, the stability of the
dispersion prepared using functionalized carbon nanotubes
Figure 2. FTIR spectrum of (a) purified and (b) purified - was observed to be better. The dispersion was stable up to
functionalized SWNT. five days. However, this partial functionalization may not be
sufficient to form stable dispersions for long time.
The dispersion of the SWCNT was also investigated using
assigned to the C=O stretching vibration in nitric acid treated TEM. The dispersability of SWCNT is also a function of the
SWCNT. The broad peak at ~3434 cm−1 is assigned to the dispersion medium. The non-functionalized carbon nanotubes
O-H stretching vibration while the peak in the frequency were seen to be agglomerated into a cluster where it was
range of 1380-1430 cm−1 may be due to the bending vibrations difficult to distinguish the nanotubes individually (Figure
of O-H groups or nitro groups [17]. The authors must change 5(a)). In the case of the functionalized carbon nanotubes
this statement. These observations are a direct evidence for (Figure 5(b) and (c)) individual carbon nanotubes are well
introduction of a carboxylic acid group on the nanotubes. distinguishable and appear as bundles.
However, the peak observed at about 2900 cm−1 is not Further, in order to study the dispersion of SWCNTs in
clearly understood. chitosan solution, the purified and functionalized carbon
The degree of functionalization (a measure of the number nanotubes were dispersed in chitosan solution to the extent
of carboxyl groups present on the surface of the carbon of 0.25 wt% based on the polymer. The chromaticity of the
nanotubes) depends on the oxidative treatment time and dispersions was exploited as the criteria of dispersion state.
temperature. Under the above treatment conditions, ~4.33×103 Figure 6 shows that the color of the dispersion of purified
carboxylic acid groups per gram of SWCNT were formed [9]. SWCNT dispersion in chitosan appears much darker than
As shown in the thermogravimetric analysis in Figure 3, that of the functionalized SWCNT. This is an indication of
the pristine SWCNT showed a weight loss at a temperature improved dispersion of functionalized carbon nanotubes.
of 380 oC corresponding to loss of amorphous carbon in the The purified and the functionalized SWCNTs were also
nanotubes. The other impurities degraded in the temperature dispersed in dilute chitosan solution (0.25 % concentration)
followed by ultrasonication for 15 min. From the TEM
images of the SWCNT dispersions (Figure 7), it is clearly

Figure 4. Photographs of vials containing dispersion of purified

(N) and functionalized (F) SWCNT (0.1 mg/ml) in ethanol. Left:
Figure 3. Thermogravimetric analysis of SWCNTs. Immediately after sonication; right: 24 h after sonication.
Improved Dispersion of Carbon Nanotubes in Chitosan Fibers and Polymers 2008, Vol.9, No.4 413

Figure 5. TEM micrographs of SWCNTs dispersed in ethanol; (a) non-functionalized SWCNT at 110,000×, (b) functionalized SWCNT at
110,000×, and (c) functionalized SWCNT at 300,000×.

molecules can act as polycationic surfactant and get

adsorbed on the SWCNT surface, thus improving the
dispersion of the SWCNTs. On the contrary, the purified but
non-functionalized carbon nanotubes show a strong tendency
to aggregate by virtue of very high van der Waal’s forces of
attraction. In Figure 7(b) the functionalized SWCNT observed
at lower magnification clearly shows uniform dispersion of
SWCNTs. This may be attributed to ionic interaction of
amino functional groups of chitosan with carboxylic acid
groups of SWCNT (introduced during functionalization).
The introduction of carboxyl groups during oxidative
functionalization is also revealed by FTIR characterization.
As shown in Figure 7(c), at higher magnification, the
individual carbon nanotubes can be clearly identified.
Generally, the SWCNTs are dispersed in very small quantities,
Figure 6. Photographs of (A) chitosan aqueous solution (3 wt.%), and hence the FTIR spectroscopy fails to collect the signals and
(B) purified SWCNT dispersion in chitosan (0.1 % wt of chitosan), UV visible spectroscopy suffers from strong photoluminescence
and (C) functionalized SWCNT dispersion in chitosan (0.1 % wt of effect of SWCNTs. On the contrary, Raman spectroscopy is
chitosan). a powerful tool to analyze the dispersion of SWCNT; it can
detect the signals from carbon nanotubes even when dispersed
observed that the functionalized SWCNTs disperse well in at very low concentration. The changes in the peaks observed at
the chitosan solution implying that chitosan is a better various Raman shifts differ in their intensities and position
dispersion medium for the SWCNTs. The chitosan macro- depending upon the quality of the dispersion [25-27]. Figure 8

Figure 7. TEM micrographs of SWCNT dispersed in chitosan (a) non-functionalized, (b) functionalized SWNT at 110,000×, and (c)
functionalized SWNT at 300,000×.
414 Fibers and Polymers 2008, Vol.9, No.4 Sukrut Ozarkar et al.

Technology (DST), Govt. of India for providing funding for

nanotechnology program at IIT Delhi. The authors also
acknowledge Reninshaw Instrument (UK) for providing the
support for carrying out Raman studies.


1. P. J. F. Harris, “Carbon Nanotubes and Related Structures”,

pp.1-12, Cambridge University Press, 1999.
2. H. G. Chae and S. Kumar, Indian J. Fibre Text., 31, 29
3. R. Ramasubramaniam, J. Chen, and H. Liu, Appl. Phys.
Lett., 83, 2928 (2003).
4. Q. Chen, C. Saltiel, and S. Manickavasagam, J. Colloid
Figure 8. Raman spectra of SWCNT (0.1 % wt of chitosan) Interf. Sci., 280, 91 (2004).
dispersed in 3 % aq chitosan solution.
5. J. Tkac and T. Ruzgas, Electrochem. Commun., 8, 899
shows the Raman spectroscopic data for purified and 6. Y.-C. Tsai and J.-D. Huang, Electrochem. Commun., 8,
functionalized SWCNTs dispersed in chitosan solution. The 956 (2006).
strong G band observed in the radial breathing mode at 1586 7. N. Grossiord, S. J. Loos, S. O. Regev, and C. E. Koning,
cm−1 for both the samples is a characteristic feature of the Chem. Mater., 18, 1089 (2006).
SWCNT. Additionally the SWCNTs show peaks between 8. L. A. Montoro and M. Rosolen, Carbon, 44, 3293 (2006).
100 cm−1 to 400 cm−1. The remaining Raman scattering 9. X. Yangchuan, L. Liang, C. Charles, and R. V. Hull,
intensities are due to the presence of polymer solution. The Langmuir, 21, 4185 (2005).
very sharp peak at 1586 cm−1 for the functionalized SWCNTs 10. S. Kumar, A. Govindaraj, J. Mofokeng, G. N. Subbanna,
suggests the better dispersion of the SWCNTs. The peak at and C. N. R. Rao, J. Phys. B-At. Mol. Opt., 29, 4925
1100 cm−1 indicates the presence of amine group in the (1996).
chitosan polymer. The peak intensity at 1100 cm−1 for the 11. A. Curulli, S. Nonziante, and A. Coppe, Microchim. Acta,
purified-functionalized SWCNT dispersed in chitosan is 152, 225 (2006).
lower than that observed for the purified SWCNT dispersed 12. A. Felten, C. Bittencourt, J. J. Pireaux, G. Van Lier, and J.
in chitosan. This drop in the intensity of the amine peak at C. Charlier, J. Appl. Phys., 98, 074308 (2005).
1100 cm−1 in the case of functionalized SWCNT indicates 13. B. N. Khare, P. Wilhite, B. Tran, E. Teixeira, K. Fresquez,
that some interaction between the carboxyl group of the D. N. Mvondo, C. Bauschlicher, and M. Meyyappan, J.
oxidized SWCNT and the amine group is possible. Hence, Phys. Chem. B, 109, 23466 (2005).
the intensity of the peak at 1100 cm−1 can be used to 14. H. Jia, Y. Lian, O. Midori, T. Nakahodo, Y. Maedab, T.
understand the nature of dispersion of SWCNT in chitosan. Tsuchiya, T. Wakahara, and T. Akasaka, Sci. Technol. Adv.
Mater., 6, 571 (2005).
Conclusion 15. D. Hill, Y. Lin, L. Qu, A. Kitaygorodskiy, J. W. Connell, L.
F. Allard, and S. Y. Ping, Macromolecules, 38, 7670
The SWCNTs were successfully purified and functionalized (2005).
to form carboxyl functional groups on the surface by 16. S. M. Lefrant, M. Baibaraca, I. Baltogb, I. Y. Mevellec, C.
covalent bonds. The optimal conditions to prepare stable Godona, and O. Chauvet, Diam. Relat. Mater., 14, 867
dispersions of SWCNTs in chitosan in acidic conditions were (2005).
demonstrated. Compared to non-functionalized SWCNT, the 17. Y. Wang, I. Zafar, and S. Mitra, Carbon, 43, 1015 (2005).
functionalized SWCNT showed a very stable dispersion in 18. K. Mylvaganam and L. C. Zhang, J. Phys. Chem. B, 108,
chitosan solution, which can be very useful to investigate the 15009 (2004).
potential application of carbon nanotubes in bio-actuators. 19. F. Liang, J. M. Beach, P. K. Rai, R. E. Smalley, and W. E.
Further, preparation of chitosan/SWCNT composite fibers Billups, Chem. Mater., 18, 1520 (2006).
exhibiting activity to electric field is under study in our 20. D. S. Bag, R. Dubey, N. Zhang, V. K. Varadan, D. Lal, and
laboratory. G. N. Mathur, Smart Mater. Struct., 13, 1263 (2004).
21. N. Grossiord, J. Loos, O. Regev, and C. E. Koning, Chem.
Acknowledgement Mater., 18, 1089 (2006).
22. C. Y. Hong, Y. Z. You, and C. Y. Pan, Polymer, 47, 4300
The authors acknowledge Department of Science & (2006).
Improved Dispersion of Carbon Nanotubes in Chitosan Fibers and Polymers 2008, Vol.9, No.4 415

23. C. Wang, G. Zhou, H. Liu, J. Wu, Y. Qiu, B. L. Gu, and W. 26. L. M. Ericson and P. E. Pehrsson, J. Phys. Chem. B, 109,
Duan, J. Phys. Chem. B, 110, 10266 (2006). 20276 (2005).
24. S. Porro, S. Musso, M. Vinante, L.Vanzetti, M. Anderle, F. 27. S. Lefrant, J. P. Buisson, O. Chauvet, J. M. Benoit, P.
Trotta, and A. Tagliaferro, Physica E, 37, 58 (2007). Bernier, and M. Baltog, Proc. Materials Research Society
25. C. G. Salzmann, T. T. C. Bryan, T. Gerard, A. L. Simonl, Symposium, 706, 231 (2002).
and L. H. G. Malcol, Carbon, 45, 907 (2007).