The role of the medium in electrochemical functionalization and dispersion of carbon nanotubes

Article abstract of DOI:10.1007/s11172-011-0169-9

An electrochemical method for dispersion of single-walled carbon nanotubes (SWNTs) is described. The technique is based on grafting of oxygen-containing functional groups to the nanotube surface during electrolysis in aqueous and nonaqueous potassium bromide solutions. A dependence of the degree of functionalization of nanotubes on the solvent was revealed experimentally. Nanotubes treated in DMSO have about 14 carbon atoms per oxygen atom from functional groups (cf. nearly four C atoms per oxygen atom in the nanotubes treated in aqueous solutions). The corresponding maximum specific capacities of the electrodes are nearly 10 and 60 F g^-1. The samples treated in solutions of KBr in DMSO have about 300 carbon atoms per bromine atom on the nanotube surface (cf. only 30 carbon atoms in the samples treated in aqueous solution). A mechanism of electrochemical modification of SWNTs is proposed. Its key step is production of atomic oxygen that oxidizes the nanotube surface with the formation of functional groups.

Full text of DOI:10.1007/s11172-011-0169-9

Russian Chemical Bulletin, International Edition, Vol. 60, No. 6, pp. 1071—1077, June, 2011
1071
The role of the medium in electrochemical functionalization
and dispersion of carbon nanotubes
A. G. Krivenko,a N. S. Komarova,a A. G. Ryabenko,a A. V. Naumkin,b E. V. Stenina,c
L. N. Sviridova,c and S. N. Sul´yanovd
aInstitite of Problems of Chemical Physics, Russian Academy of Sciences,
1 prosp. Akad. Semenova, 142432 Chernogolovka, Moscow Region, Russian Federation.
Fax: +7 (496) 522 3507. Eꢀmail: krivenko@icp.ac.ru
bA. N. Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences,
28 ul. Vavilova, 119991 Moscow, Russian Federation.
Fax: +7 (499) 135 5328
cDepartment of Chemistry, M. V. Lomonosov Moscow State University,
3 Bidg. 1, Leninskie Gory, 119992 Moscow, Russian Federation.
Fax: +7 (499) 939 0067
dA. V. Shubnikov Institute of Crystallography, Russian Academy of Sciences,
59 Leninsky prosp., 119333 Moscow, Russian Federation.
Fax: +7 (499) 135 1011
An electrochemical method for dispersion of singleꢀwalled carbon nanotubes (SWNTs) is
described. The technique is based on grafting of oxygenꢀcontaining functional groups to the
nanotube surface during electrolysis in aqueous and nonaqueous potassium bromide solutions.
A dependence of the degree of functionalization of nanotubes on the solvent was revealed
experimentally. Nanotubes treated in DMSO have about 14 carbon atoms per oxygen atom
from functional groups (cf. nearly four C atoms per oxygen atom in the nanotubes treated in
aqueous solutions). The corresponding maximum specific capacities of the electrodes are nearꢀ
ly 10 and 60 F g–1. The samples treated in solutions of KBr in DMSO have about 300 carbon
atoms per bromine atom on the nanotube surface (cf. only 30 carbon atoms in the samples
treated in aqueous solution). A mechanism of electrochemical modification of SWNTs is proꢀ
posed. Its key step is production of atomic oxygen that oxidizes the nanotube surface with the
formation of functional groups.
Key words: singleꢀwalled carbon nanotubes, dispersion, functionalization, electrolysis, diꢀ
methylsulfoxide.
At present, there are only a few fields of application of
nanosized carbon and, in particular, carbon nanotubes.
For instance, nanotubes can be used as components of
polymerꢀnanotube composites.1 Indeed, the introduction
of very small amounts (~10–3—10–2 wt.%) of singleꢀwalled
carbon nanotubes (SWNTs) into a polymer matrix
leads to a noticeable improvement of operational efficiency
of such composites.2 However, an increase in the SWNT
concentration to about 1—5 wt.% usually causes the
strength of a material to decrease.3 It is believed that such
a behavior is due to sliding motion of nanotubes relative to
one another in nanotube bundles formed at sufficiently
high concentration of SWNTs in the initial monomer.
This sliding motion causes nanotubes to split off from
the polymeric material.4 Therefore, the design of highꢀ
strength, nanotubeꢀbased composites requires individualꢀ
ization of nanotubes followed by their dispersion in the
polymer matrix. Also, the number of functional groups
covalently bound to the nanotube surface should be optiꢀ
mal. A particular number of functional groups is necessary
to form strong bonds between the individual SWNT and
the polymer matrix, but if the number of such bonds is too
large, the properties of the nanotube deteriorate.
Both these problems, viz., dispersion and controlled
functionalization of nanotubes, are far from being solved.
There are neither commonly accepted method for the synꢀ
thesis of particular types of individual nanotubes nor releꢀ
vant procedures for their purification.5 Usually, the availꢀ
able nanotube preparations represent mixtures of carbon
nanotubes with different diameters, lengths, and chiraliꢀ
ties. Moreover, they often contain uncontrolled amounts
of metallic and graphitic impurities. In addition, nanoꢀ
tubes can form, through van der Waals forces, either hexaꢀ
gonally packed bundles typical of "short" SWNT with an
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 6, pp. 1046—1052, June, 2011.
1066ꢀ5285/11/6006ꢀ1071 © 2011 Springer Science+Business Media, Inc.

1072
Russ.Chem.Bull., Int.Ed., Vol. 60, No. 6, June, 2011
Krivenko et al.
as is. All potentials (E) are given relative to aqueous saturated
calomel electrode.
average length of 1 μm or coils of "long" nanotubes with
average length ranging from 10 to 1000 μm. The energy of
van der Waals interaction between individual SWNTs is
0.5—1.0 eV per nanometer.6 Such a strong interaction is
the major obstacle to dispersion of individual SWNTs in
aqueous and organic solutions and polymer matrices.
Unfortunately, considerable literature on methods for
functionalization and dispersion of SWNTs includes only
few studies concerning electrochemical approaches and
methods. In particular, electrochemical disruption of
nanotube bundles (without oxidation of the surface) by
incorporating Li ions solvated by DMSO molecules into
channels between tubes was reported.7 However, the elecꢀ
trolyte decomposes at the potentials used by the authors of
Ref. 7. This precludes complete individualization of nanoꢀ
tube from bundles. There is also an electrochemical methꢀ
od for partial disruption of nanotubes during electrolysis
of aqueous chloride solutions.8 Some chlorine atoms proꢀ
duced in the course of electrolysis at a specified potential
move into the voids between tubes and adsorb on the surꢀ
face of nanotubes as well as form covalent bonds with
nanotube carbon atoms. Other Cl atoms interact with waꢀ
ter with the formation of hypochlorite ions which decomꢀ
pose and thus serve as sources of atomic oxygen that causꢀ
es the oxidation of the nanotube surface and the formation
of various covalent bonds. In this work, we compare the
effect of electrolysis of KBr solutions on the dispersion
and functionalization of SWNTs in aqueous and nonaꢀ
queous (DMSO) solutions. Potassium bromide was chosen
as the background electrolyte because it is well soluble in
aqueous and nonaqueous electrolytes and bromine ions
have a low oxidation potential; DMSO is a readily availꢀ
able aprotic solvent of relatively low toxicity. Owing to
easy penetration through the skin of a living organism,9 it
is widely used in biology and medicine as drug delivery
agent. One can assume that the use of DMSO as organic
solvent for dispersion of SWNT bundles will make it posꢀ
sible to extend the field of application of individual nanoꢀ
tubes in medicine. In addition, the authors of some studies
reported the efficiency of DMSO as a solvent for funcꢀ
tionalized nanotubes10,11 and high wettability of carbon
nanotubes in DMSO.12
Singleꢀwalled carbon nanotubes for the preparation of nanoꢀ
paper were synthesized by the electric arc method using a Ni/Y
catalyst. Nanopaper represented a black thin layer of a paperꢀ
like material 0.03—0.04 mm thick with a specific surface weight
of 1—2 mg cm–2 containing 98–99% of nanotubes and less than
0.5% of metal. According to electron microscopy data, nanoꢀ
paper represents ribbons comprising 500—10000 SWNTs with
an average diameter of about 1.4 nm and an average length of
nearly 1 μm. In turn, these ribbon form larger carpetꢀlike aggreꢀ
gates up to a few micrometres in size. Carbon nanotubes used in
this study were extensively characterized by electron microscopy,
IR, NMR spectroscopy, Xꢀray photoelectron spectroscopy
(XPS), and Raman spectroscopy, as well as by mass spectroꢀ
metry and elemental analysis.13 The preparation of electrodes
based on carbon nanopaper was described earlier.14
The content of carbon, oxygen, and nitrogen was determined
with an Elementar Vario cube automated CHNS/O analyzer. To
determine bromine, the sample was combusted in oxygen in
a Schoeninger flask followed by titrimetric determination of Br–
ions. Products of electrolysis of potassium bromide solutions were
analyzed using a Shimadzu LCꢀ2020 liquid chromatoꢀmass
spectrometer. Thermogravimetric and differential thermal analꢀ
ysis were done with a NETZSCH STAꢀ409 Luxx thermal
analyzer with a mass spectrometric accessory for analysis of exꢀ
haust gases.
Structural features of SWNTs were studied with a Carl Zeiss
LEO 912 AB OMEGA transmission electron microscope using
samples prepared by dispersion of nanopaper in isopropyl alcoꢀ
hol under sonication. The UV, visible, and nearꢀIR absorption
spectra were recorded using a Lumex FTꢀ02 IR Fourier spectroꢀ
meter (Lumex Instruments Ltd., Russia). The Raman spectra
were acquired with a Thermo Scientific Nicolet 9810 Raman
Fourier spectrometer using a laser operating at λ = 976 nm as the
excitation source. The Auger spectra and Xꢀray photoelectron
spectra were recorded with a MK II VG Scientific spectrometer.
Results and Discussion
Passage of electric current through the interface beꢀ
tween the solution and the electrode made of nanotube
bundles is accompanied by the formation of reactive interꢀ
mediates (radicals or radical ions) on the surface of SWNTs
and near it. These species participate in subsequent homoꢀ
geneous and heterogeneous reactions resulting in stable
molecules or ions. Both intermediates and stable species
can penetrate into cavities and channels between SWNTs
that form nanotube bundles through van der Waals interꢀ
action. One can assume that the molecules and ions are
then physically sorbed and chemisorbed on the surface
of nanotubes. It is also probable that the most reactive
intermediates form covalent bonds with the surface of
nanotubes.8
Experimental
Electrochemical studies were carried out by cyclic voltamꢀ
metry (CV) using an IPCꢀPro L potentiostat. Aqueous solutions
were prepared using water triple distilled in quartz vessels. DMSO
was stored over alkali and distilled in vacuo. From 1H NMR
spectra recorded on a Brucker Nova 500 spectrometer operating
at 500 MHz it follows that the concentration of water in DMSO
solutions was at most 0.22 mol.% while the concentration of the
main substance was 99.7%. Prior to measurements, the dissolved
oxygen was removed by bubbling argon (high purity grade) over
a period of about 40 min. The salts of extra pure grade were used
In 1.0 M aqueous KBr solutions at potentials E ≈
≈ (1—1.5) V on the nanopaper electrode, one deals with
electrochemical oxidation of bromine ions. According to
chromatoꢀmass spectroscopic analysis in solution, the

Role of the medium in dispersion of nanotubes
Russ.Chem.Bull., Int.Ed., Vol. 60, No. 6, June, 2011
1073
I/mA
an increase in the surface area of the SWNTꢀelectrode
accessible to the electrolyte solution, while the decrease in
the capacity on longꢀterm electrochemical treatment
seems to be due to violation of its continuity. This concluꢀ
sion was also confirmed by transmission electron microꢀ
scopy data, viz., nanotube bundles in the electrochemicalꢀ
ly treated samples are disrupted to smaller systems whose
width is 5—10 times smaller than that of the initial nanoꢀ
paper bundles (see Fig. 2, a, b). Figure 2, c also shows no
distortion of the SWNT structure upon electrochemical
treatment. Figure 3 presents the smallꢀangle Xꢀray scatꢀ
tering (SAXS) patterns for the initial nanopaper samples
and for the nanopaper samples electrochemically treated
in aqueous KBr solutions. The Xꢀray diffraction patterns
were recorded at different times of electrochemical treatꢀ
ment. The intensity of the first reflection corresponding to
the average distance between tubes in nanotube bundles
significantly decreases during electrolysis. Nanotube bunꢀ
dles form a twoꢀdimensional hexagonal lattice whose secꢀ
tion is shown in the inset in Fig. 3. Electrochemical treatꢀ
ment of nanopaper causes the distances between tubes to
increase randomly; as a consequence, the interplanar spacꢀ
ings are no longer constant and the reflection amplitude
decreases. This suggests that electrochemical treatment
leads to a greater disordering in the arrangement of nanoꢀ
tubes in the bundles. Figure 4 presents the UV, visible, and
nearꢀIR absorption spectra of the initial and electrochemꢀ
ically treated nanotubes. Initial SWNTs have a pronounced
van Hove spectrum; however, the band intensities decrease
as the time of electrochemical treatment of the samples
increases. In particular, the band intensities are halved
during the treatment time necessary to achieve a maxiꢀ
mum electrical capacity C. This indicates partial violation
of the electronic structure of nanotubes, which may probꢀ
ably be due to the appearance of functional groups on their
surfaces. Thus, analysis of the absorption spectra confirms
the conclusion drawn after comparison of the micrographs
3
2
0.4
0
1
–0.4
–0.8
–0.4
–0.2
0
E/V
Fig. 1. Cyclic voltammograms of nanotube electrode in 1 M aqueous
KBr solutions at a potential sweep rate of 0.05 V s–1: initial
sample (1) and samples electrolyzed for 25 (2) and 50 min (3).
For clarity, the currents of the initial sample are magnified (×4).
major products of the reaction include molecular bromine
–
and Br2 ions. The process is accompanied by considerꢀ
able changes in the CV currents for this electrode in the
potential range from 0.1 to –0.5 V (Fig. 1). The absolute
values of the forward and reverse potential sweep currents
significantly increase, whereas the shapes of the CV curves
change only slightly. This corresponds to domination of
the charging currents of the double electrical layer over
the electrode reaction currents and makes it possible to
evaluate the electrical capacity of the samples (C) from
a simple ratio C = Ic/v, where Ic is the charging current
and v is the sweep rate of the potential E. In the course
of electrolysis, the initial capacity of electrodes increases
12—50 times, reaches a maximum value after nearly 60 min,
and then decreases. The increase in C is apparently due to
a
b
c
100 nm
100 nm
20 nm
Fig. 2. Micrographs of nanotube bundles in the SWNT sample before (a) and after electrochemical treatment in 1 M aqueous KBr
solution during 50 min (b, c).

1074
Russ.Chem.Bull., Int.Ed., Vol. 60, No. 6, June, 2011
Krivenko et al.
dw—v = 0.32 nm
trolyte at E ≥ 1.0 V. This decomposition is accompanied
by the appearance of specific smell and darkening of the
solution. Analysis of the 13C NMR spectra recorded beꢀ
fore and after electrolysis revealed no signals correspondꢀ
ing to the DMSO decomposition products at the potenꢀ
tials applied.
I (rel.units)
8000
6000
4000
2000
0
1
Dn
As in aqueous solutions, electrolysis in solutions of
KBr in DMSO is accompanied by changes in the shape of
the CV curves of the SWNT electrode. However, the iniꢀ
tial capacity of the electrode immersed into the DMSO
solution is three—four times higher than that of the elecꢀ
trode immersed into aqueous solution (Table 1). Apparꢀ
ently, the working surface area of nanotubes in the elecꢀ
trolyte solution in the nonaqueous medium is larger than
in the aqueous solution. It seems that nanotubes are charꢀ
acterized by better wettability in DMSO than in water and
therefore the electrolyte (DMSO) can penetrate faster into
channels between individual nanotubes in the bundle. In
this case, the initial electrode capacity is nearly doubled
during electrolysis in the solution of KBr in DMSO, which
is much smaller than in aqueous KBr solutions. Longꢀ
term electrolysis in the solution of KBr in DMSO causes
the capacity C to decrease, but the decrease is observed at
longer electrolysis times (~80 min) compared to aqueous
solutions. Electrochemical treatment in the solutions of
KBr in DMSO over a period of 80 min leads to a small
(~5%) decrease in the intensity of the van Hove band (see
Fig. 4). As in the case of electrolysis in aqueous KBr soluꢀ
tions, this decrease can be associated with violation of the
electronic structure of nanotubes, which may be due to the
formation of functional groups on their surface. An analyꢀ
sis of the 1H NMR spectra of the electrolyte suggests
that the concentration of water in the DMSO solutions
(≤0.22 mol.%) remains unchanged during electrolysis. At
the same time, assuming that transformation of one H2O
molecule requires one electron, it appears that the charge
consumed during electrolysis is more than ten times highꢀ
er than the number of electrons required for total transforꢀ
mation of all water molecules present in the solution. The
fact that water molecules are not involved in electrolysis is
dw—v
d10
4
2
3
Dn + dw—v
5
10
15
20
2θ/deg
Fig. 3. Smallꢀangle Xꢀray scattering patterns of the samples of
initial nanopaper (1) and nanopaper electrochemically treated in
1 M aqueous KBr solutions during 15 (2), 30 (3), and 60 min (4).
For clarity, the intensities in the SAXS patterns 2—4 are magꢀ
nified (×2, ×3, and ×5, respectively). Inset: a schematic view of
a SWNT bundle; dw—v is the van der Waals interꢀtube distance,
Dn is the nanotube diameter, and d10 is the lattice constant corꢀ
responding to the first SAXS reflection (2θ ≈ 6°).
of the SWNT samples (see Fig. 2) that the nanotubes
retain their structure until the electrode capacity increases.
The results obtained cast some doubt on the statement
that weakening of the first reflection in the Xꢀray diffracꢀ
tion patterns of SWNT is due to breakdown of electroꢀ
chemically treated nanotubes and to formation of amorꢀ
phous phase.
Electrolysis in 0.2 M solutions of KBr in DMSO at
anode potentials E ≈ 0.8—0.9 V on the nanotube electrode
is also accompanied by release of bromine. However, unꢀ
like aqueous solutions, the use of higher positive potenꢀ
tials is impossible because of decomposition of the elecꢀ
A (rel.units)
1.6
1.2
Table 1. Changes in the capacity of nanopaper electrodes during
electrochemical treatment in different solvents
1
2
0.8
0.4
3
4
Treatment time
/min
Capacity/F g–1
H2O, 1 M KBr
DMSO, 0.2 M KBr
0
1—2
10—12
22—26
58—68
60—68
45—50
—
4—5
6—7
7—8
10—12
12—15
14—16
11—12
250
750
1250
1750
2250 λ/nm
15
30
45
60
75
90
Fig. 4. Absorption spectra of initial SWNTs (1) and SWNTs
electrochemically treated in 0.2 M solution of KBr in DMSO
until the maximum electrical capacity (2), in 1 M aqueous KBr
solution for 5 min (3), and until the maximum capacity (4). All
absorption spectra are matched at a wavelength of 1319 nm.

Role of the medium in dispersion of nanotubes
Russ.Chem.Bull., Int.Ed., Vol. 60, No. 6, June, 2011
1075
I (rel.units)
probably due to the bulk properties of DMSO. The bulk
properties of pure DMSO and its aqueous solutions have
been well documented.15—17 These systems are characterꢀ
ized by selfꢀassociation and association which are favored
by high polarity of the DMSO molecule (μ = 3.96 D)
whose positive effective charge (+0.5) is localized on the
sulfur atom. Owing to the dipoleꢀdipole interaction and
hydrogen bonds, chain and cyclic aggregates exist in liquid
DMSO. These aggregates are responsible for the ordered
structure of liquid DMSO at T < 40 °C. Owing to protoꢀ
philic properties of DMSO, associates readily break down
upon the addition of proton donors. Mixing DMSO with
water is a highly exothermic process. The structure characꢀ
teristic of pure DMSO breaks down and a new structure is
formed, which includes hydrogenꢀbonded water moleꢀ
cules. The highly hydrophilic S=O group can even form
bonds with two water molecules. It seems not improbable
that water molecules involved in such bonds are much less
reactive than in water and therefore do not participate in
functionalization of nanotubes in the DMSO solutions.
In an attempt to establish the mechanism of electroꢀ
chemical action on the SWNT surface during electrolysis
of KBr solutions it is of interest to reveal how the solvent
nature affects the elemental composition and the state of
the nanotube surface. Once the maximum capacity during
electrolysis in aqueous solution was reached, the electrode
material was analyzed for Br, C, O, and N and the results
obtained were compared with the composition of the iniꢀ
tial nanopaper (Table 2). From the data of Table 2 it
follows that electrochemical treatment leads to an increase
in the content of bromine and oxygen in the samples. To
establish the nature of the interaction of bromine with the
nanotube surface, a treated sample was washed with water
and heated to 50—60 °C. Then, thermogravimetric analyꢀ
sis was carried out and the mass spectra of the gases liberꢀ
ated were recorded. It was found that in these samples, the
content of bromine chemically bound to nanotube carbon
atoms is low. This conclusion is confirmed by the strucꢀ
ture of the Xꢀray photoelectron spectrum of the treated
sample which shows a peak at 70.2 eV characteristic of the
C—Br bond. From analysis of the C1s peak in the XPS
spectra it follows that electrochemical treatment in DMSO
does not bring about significant changes in the spectrum,
whereas a similar process in aqueous solutions leads to
pronounced spectral changes in the highꢀenergy region.
This can probably be due to the formation of C—O—C,
C=O, and C(O)O bonds. In addition, XPS spectra in the
O1s region for all samples are almost the same (Fig. 5, a).
This suggests that oxygen atoms in these samples are in
the same chemical state, although are present in different
concentration. The results of quantitative analysis for carꢀ
bon, oxygen, nitrogen, and bromine on the surface of nanoꢀ
paper (see Table 2) correlate with the changes in the inꢀ
tensity ratio of the D/G modes in the Raman spectra of
nanotubes during electrolysis. This ratio increases to nearꢀ
a
1
2
3
1
0
530
535
Binding
energy/eV
I (rel.units)
b
1
2
40
3
20
240
270
Kinetic
energy/eV
Fig. 5. Xꢀray photoelectron spectra in the O1s region (a) and the
lineshape of the C KVV Auger spectra (b) of initial SWNTs (1)
and SWNT samples after electrochemical treatment in 0.2 M
solution of KBr in DMSO (2) and in 1 M aqueous KBr solution (3).
The spectra were normalized to the most intense peak.
ly 30.0% after electrochemical treatment in aqueous KBr
solution for 60 min (cf. about 8.0% for DMSO solution).
Taking into account the fact that in the initial sample this
ratio was at most 5.1%, one can suggest that the number
of defects formed on the SWNT surface during electrolysis
in aqueous KBr solutions is much larger than in the DMSO
solutions. According to XPS data, there are about 14 carꢀ
bon atoms per oxygen atom from functional groups after
treatment in DMSO (cf. nearly 4 carbon atoms after treatꢀ
ment of nanopaper in aqueous solutions). The correspondꢀ
ing maximum specific electrode capacitances are ~15 and
~70 F g–1. The samples treated in the solutions of KBr in
DMSO are characterized by a value of about 300 carbon
atoms per bromine atom on the nanotube surface (cf. only

1076
Russ.Chem.Bull., Int.Ed., Vol. 60, No. 6, June, 2011
Krivenko et al.
Table 2. Results of quantitative XPS analysis of initial nanopaper
Scheme 1
and nanopaper electrochemically treated in different media*
Br–
Br
(1)
(2)
(3)
(4)
(5)
SWNTs
C1s
O1s
N1s
at.%
Br3d
2 Br + H2O
Br + SWNT
Br + Br
BrO–
BrO– + Br– + 2 H+
SWNTꢀBr
Initial sample
95.0 5.0 [1.0] 0.0 [0.1] 0.0 [0.8]
91.6 6.6 1.0 0.3
After electrolysis
in 0.2 M solution
of KBr in DMSO
After electrolysis
in 1 M aqueous
KBr solution
Br2
77.7 18.7 [8.0] 1.0 [0.7] 2.6 [1.3]
BrO3– + 2 Br–
*Results of elemental analysis are given in square brackets.
(6)
30 carbon atoms for the samples treated in aqueous KBr
solution).
SWNTox is oxidized SWNT.
Figure 5, b presents the Auger spectra of the initial and
treated samples of the nanopaper electrodes. It follows
that, as for the Xꢀray photoelectron spectra, the most signiꢀ
ficant changes are observed in the spectra of the SWNT
samples obtained by electrochemical treatment in aqueꢀ
ous KBr solutions. Earlier,18 it was shown that the shift of
the spectra toward lower energies of Auger electrons indiꢀ
cates some changes in the πꢀband of the density of states
in region near the Fermi level. This can be due to both
weakening of the interꢀtube interaction and oxidation of
the sample. A possible oxidation is confirmed by the fact
that electrolysis in aqueous KBr solution is accompanied by
considerable dispersion of nanotubes and formation of
oxygenꢀcontaining functional groups on the nanotube
surface.
formation of atomic oxygen which oxidizes the nanotube
surface with the formation of functional groups (6) that
impart hydrophilic properties to nanotubes. The affinity
of water molecules to nanotube surface becomes higher
than the force of the interaction between tubes; this finally
leads to disruption of SWNT bundles. Summing up, the
main route of functionalization under the electrolysis of
bromine ions is the oxidation of SWNT surface rather
than halogenation. By choosing the electrolysis medium
one can vary the ratio of the degree of functionalization to
the degree of dispersion of SWNTs.
References
Analysis of the results obtained during electrolysis of
aqueous and nonaqueous potassium bromide solutions
using a nanostructured carbon electrode as anode suggests
a key role of water in the functionalization and dispersion
of nanotubes. Relevant mechanisms of these processes are
shown in Scheme 1. In the first step, the oxidation of Br–
ion at the electrode results in bromine atom as a reactive
intermediate (1) which can react in different manner. The
main route includes a reaction with water resulting in hypoꢀ
bromite ion (2). Yet another route implies a direct interacꢀ
tion of bromine atom with the graphene surface of nanoꢀ
tube (3); in addition, bimolecular deactivation (4) is posꢀ
sible. Ignoring the amount of oxygen in the initial samꢀ
ples, the data listed in Table 2 can be used to estimate the
lower bound of the ratio of efficiencies of the main and
side reaction routes of Br atom. It appears that, in the
presence of both solvents, the main reaction route is five
to seven times more probable than the side routes.
Reactions included in the main route result in disproꢀ
portionation of some hypobromite ions to bromate ions
(5) detected in solution by chromatoꢀmass spectrometry.
Other hypobromite and bromate ions decompose with the
1. B. S. Banerjee, T. HemrajꢀBenny, Adv. Mater., 2005, 17, 17.
2. E. R. Badamshina, M. P. Gafurova, Ya. I. Estrin, Usp. Khim.,
2010, 79, 1027 [Russ. Chem. Rev. (Engl.Trans.), 2010,
79, 945].
3. M. Moniruzzaman, K. I. Winey, Macromolecules, 2006,
39, 5194.
4. P. M. Ajayan, J. M. Tour, Nature (London), 2007, 447, 1066.
5. V. G. Udovitsky, Fiz. Inzh. Poverkhn., 2009, 7, 351 [Physical
Surface Engineering (Engl. Transl.), 2009, 7, No. 4].
6. L. Vaisman, H. D. Wagner, G. Marom, Adv. Colloid Interꢀ
face Sci., 2006, 128—130, 37.
7. O. Tanaike, O. Kimizuka, N. Yoshizawa, K. Yamada,
X. Wang, H. Hatori, M. Toyoda, Electrochem. Commun.,
2009, 4, 1441.
8. P. M. Rafailov, C. Thomsen, U. DettlaffꢀWeglikowska,
S. Roth, J. Phys. Chem. B, 2008, 11, 25368.
9. C. NietoꢀDraghi, J. B. Avalosa, J. Chem. Phys., 2003,
119, 4782.
10. V. V. Chaban, O. N. Kalugin, J. Mol. Liq., 2010, 151, 113.
11. S. Banerjee, S. S. Wong, J. Am. Chem. Soc., 2002,
124, 8940.aa
12. M. Pavese, S. Musso, S. Bianco, M. Giorcelli, N. Pugno,
J. Phys. Condens. Matter, 2008, 20, 474206.

Role of the medium in dispersion of nanotubes
Russ.Chem.Bull., Int.Ed., Vol. 60, No. 6, June, 2011
1077
13. A. V. Krestinin, A. P. Haritonov, Yu. M. Shulga, O. M.
Zhigalina, E. I. Knerelman, M. Dubois, M. M. Brzhezinꢀ
skaya, A. S. Vinogradov, A. B. Preobrazhencky, G. I. Zveꢀ
reva, M. B. Kislov, V. M. Martynenko, I. I. Korobov, G. I.
Davydova, V. G. Zhigalina, N. A. Kiselev, Ros. nanoꢀ
tekhnologii, 2009, 4, 67 [Nanotechnologies in Russia
(Engl. Transl.), 2009, 4, 60].
14. A. G. Krivenko, V. I. Matyushenko, E. V. Stenina, L. N.
Sviridova, A. V. Krestinin, G. I. Zvereva, V. A. Kurmaz,
A. G. Ryabenko, S. N. Dmitriev, V. A. Skuratov, Electroꢀ
chem. Commun., 2005, 7, 199.
15. K. Mizur, S. Imafuji, T. Ochi, J. Phys. Chem. B, 2000,
104, 11001.
16. M. C. R. Symons, G. Eaton, J. Chem. Soc., Faraday Trans. 1,
1985, 81, 1963.
17. Yu. A. Kukk, Zh. Klavil´ye, Elektrokhimiya, 1977, 13, 841
[Sov. Electrochemistry (Engl.Trans.), 1977, 13, 712].
18. A. P. Dementjeva, K. I. Maslakova, A. V. Naumkin, Appl.
Surf. Sci., 2005, 245, 128.
Received March 4, 2011