Microwave-assisted covalent sidewall functionalization of multiwalled carbon nanotubes

Article abstract of DOI:10.1002/chem.200501314

Thermal cycloaddition of 1,3-dipolar azomethine ylides to the side-walls of multiwalled carbon nanotubes (MWNTs) has been used to prepare MWNTs that contain 2-methylenethiol-4-(4-octadecyloxyphenyl) (4), N-octyl-2-(4- octadecyloxyphenyl) (5) or 2-(4-octad

Full text of DOI:10.1002/chem.200501314

FULL PAPER
DOI: 10.1002/chem.200501314
Microwave-Assisted Covalent Sidewall Functionalization of Multiwalled
Carbon Nanotubes
[
a, b]
[a]
Junxin Li
and Helena Grennberg*
Abstract: Thermal cycloaddition of 1,3-
dipolar azomethine ylides to the side-
walls of multiwalled carbon nanotubes
the amount of added groups after only
2 h of MiW heating at 2008C, deter-
mined by using thermogravimetric
analysis, was found to be in the same
range as that obtained after 100–120 h
of conventional heating of soluble and
insoluble MWNTs. Solubility is a key
feature for a successful MiW-heated re-
action; MWNTs insoluble in the reac-
tion medium yielded considerably less
addends in the MiW-heated reactions
than in the conventionally heated reac-
tion. The location and even distribution
of the pyrrolidine units over the outer-
most layer of the MWNTs was verified
by transmission electron microscopy
analysis of 4 that had been treated with
gold nanoparticles and thoroughly
washed to remove gold particles adsor-
bed on nonfunctionalized parts of the
MWNTs.
(
MWNTs) has been used to prepare
MWNTs that contain 2-methylenethiol-
4
2
-(4-octadecyloxyphenyl) (4), N-octyl-
-(4-octadecyloxyphenyl) (5) or 2-(4-
octadecyloxyphenyl)pyrrolidine
units. All these contain the 4-octadecyl-
oxyphenyl substituent that acts as a
(6)
ACHTREUNG
solubilizing group. Microwave (MiW)-
assisted heating was found to be highly
efficient for soluble MWNTs, for which
Keywords: cycloaddition · gold ·
nanoparticles · nanotubes
Introduction
Both the ends and the sidewalls of the nanotubes can be
[7]
[8]
targeted; both noncovalent and covalent approaches
have been used. By proper choice of the structure added,
practically insoluble NTs can be solubilized without affect-
ing other properties of the NTs. The dominant method has
been the use of carboxylic acid groups introduced at the
ends and other defect sites during cleaning of the raw NT
The remarkable physical, electronic, and chemical proper-
ties of carbon nanotubes (NTs) make them interesting for
[1]
applications in many fields, but to progress from ideas to
applications reliable methods for synthesis, purification, and
further manipulations of the NTs are of great importance.
Chemical functionalization of NTs is a branch of this re-
search that has attracted considerable attention over the
[8–11]
[8b,9,11]
[10]
material,
creating amide
or ester
linkages, as
well as the formation of salts through acid–base reac-
[
2]
[9d,e,11]
past few years. By introducing, for example, solubilizing or
substrate-targeting units, the electronic and mechanical
tions.
Perhaps the most versatile group of side-wall-tar-
geting reactions is the 1,3-dipolar cycloaddition of azome-
thine ylides, originally developed by Prato et al. for
[1]
properties of the fascinating single-, double-, or multiwal-
led (SW, DW, or MW, respectively) structures may find real
[12]
[60]fullerene
SWNTs.
and more recently demonstrated for
[3]
[4]
[13,14]
use as, for example, field emitters, memory elements,
[5]
[6]
sensors, actuators, and so forth. Moreover, covalent or-
ganic functionalization of NTs will bring the area closer to
the fields of macro- and biomolecular chemistry.
Microwave (MiW) heating has become an important tool
[15]
for efficient execution of organic reactions;
excellent
yields are often obtained within reaction times considerably
shorter than those measured when conventional heating is
[16]
used. In the field of carbon nanotube functionalization, a
few MiW-assisted examples have been reported for the puri-
[
a] Dr. J. Li, Prof. Dr. H. Grennberg
Department of Organic Chemistry, Uppsala University
PO Box 599, 75124 Uppsala (Sweden)
Fax : (+46)18-471-3818
[17]
fication of raw NTs. Very recently, chemical functionaliza-
tion of SWNTs was achieved in which both end carboxyl
E-mail: helena.grennberg@kemi.uu.se
[11,18]
units and sidewalls were targeted.
[
b] Dr. J. Li
School of Sciences
We have now further confirmed that MiW heating is
indeed highly useful for sidewall-targeting addition reactions
to MWNTs by demonstrating that it is possible to obtain
Hebei University of Science and Technology
Shijiazhuang 050018 (China)
Chem. Eur. J. 2006, 12, 3869 – 3875
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3869

soluble sidewall-functionalized MWNTs that also have
useful properties. In particular, pyrrolidine units with gold-
targeting groups have been added in fast and convenient re-
actions, and with control of the amount of addends at the
outermost layer of the MWNTs.
Our findings are important as this sidewall-targeting ap-
proach has a drawback that needs to be considered careful-
[
19,20]
ly.
Since the reaction involves the p electrons of the
tube wall the electronic structure of a sidewall-functional-
ized NTs will be disturbed. Haddon et al. reported that, at
25% functionalization with dichlorocarbene, the SWNT
band structure was completely disrupted and that thermal
treatment above 3008C did not restore the original SWNT
[21]
electronic structure. It is thus surprising that reports de-
tailing successful sidewall modification of MWNTs are not
[13b,22]
more common,
in spite of the fact that a MW (or DW)
NT material would be more useful than a SWNT material in
applications that require both the new properties of the
functionalized outer tube and the electronic properties of
the intact inner layers.
Results and Discussion
Pretreatment: All work was carried out using commercially
available MWNTs. The as-received material was treated
with aqueous HNO3 or an HNO –H SO mixture
Figure 1. FTIR spectra of MWNTs 1–3, pyrrolidinated MWNTs 4–6, and
-octadecyloxybenzaldehyde.
4
3
2
4
[9a,10c]
A
C
H
T
R
E
U
N
G
(Scheme 1),
to remove various non-NT carbon materi-
als and catalyst particles. The more reactive ends of the NTs
and defect sites at the sidewalls were oxidized by the addi-
tion of oxygen-containing functional groups to the rims of
the added hydrocarbon chain. Although trace amounts of
paramagnetic metal leads to broad signals, the presence of
the hydrocarbon chain is clear from H NMR spectra, not
[8b,9,11]
1
the opened NTs.
For SWNTs, carboxyl IR features
[11]
appear after treatment with aqueous HNO3, whereas simi-
lar treatment of MWNTs resulted in a clean MWNT materi-
al 1, essentially free of carboxyl groups (Figure 1). To obtain
carboxylated MWNTs 2 that can be solubilized by esterifica-
tion, sonication in concentrated HNO –H SO was required.
shown here.
Dipolar addition to insoluble and soluble MWNT materials:
The azomethine ylides are among the most versatile 1,3-di-
poles owing to the large number of methods available for
the generation of the active reagent, the possibilities to fur-
3
2
4
À1
The IR spectrum of 2 exhibits a carbonyl peak at 1710 cm ,
C=C stretching peaks at 1635 and 1560 cm , a broad peak
at 1100 cm , and hydroxyl absorption in the 3500 cm re-
À1
ther modify added groups, and their reactivity towards elec-
À1
À1
[12b,23]
tron-poor olefins.
A very simple and powerful approach
[10a–c]
gion.
The soluble MWNT–ester 3, formed in the reac-
to fabricate azomethine ylides is the decarboxylation of imi-
nium salts derived from the condensation of amino acids
and aldehydes/ketones. It is possible to carry out this proce-
dure as a one-pot reaction. As the aldehyde component we
have used 4-octadecyloxybenzaldehyde, which, owing to its
long flexible hydrocarbon chain, could induce solubility to
the addition product. The aldehyde and either cysteine, N-
octyl glycine or glycine (Scheme 2) were allowed to react
with insoluble MWNT 1 and with the soluble MWNT–ester
[
10c]
tion between the anion of 2 and octadecyl bromide,
ex-
À1
hibits carbonyl stretching at 1720 cm and peaks centered
À1
at 2917 and 2848 cm , corresponding to the CÀH stretch of
3
by using either conventional heating or MiW heating.
The conventionally heated reactions yielded the expected
soluble products both from 1 and 3, whereas in the MiW-
heated reactions, only the soluble MWNT–ester 3 reacted to
any considerable extent. The low reactivity observed for the
MWNTs 1 is probably due to the insolubility of the materi-
al. When using conventional heating, the extended reaction
times with constant stirring lead to sufficient exposure of
Scheme 1. Transformation of as-received MWNTs into the clean insolu-
ble MWNT 1 and the soluble MWNT–ester 3. Reagents and conditions:
a) 2.6m aqueous HNO
sonication, 5 h; c) aqueous NaOH, hexadecyltrimethylammonium bro-
mide, C18 37Br, 1008C, 2 h.
3 2 4 3
, 1008C, 48 h; b) conc. H SO /HNO (3:1 v/v),
H
3870
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3869 – 3875

Functionalization of Carbon Nanotubes
FULL PAPER
MWNTs 1 and, as expected,
higher for the MWNT–COOH
2
. The latter ratio is very simi-
lar to that of the MWNT–ester
, thus the hydrocarbon chains
3
introduced do not significantly
affect the Raman properties of
the MWNTs. On the other
Scheme 2. Pyrrolidination of the MWNTs. Conditions for pyrrolidination of 1: DMF, conventional heating at hand, the D:G ratio of the pyr-
1
1
308C for 5 d. Conditions for pyrrolidination of 3: toluene, MiW heating at 2008C for 2 h (4a), 4 h (4b), at
308C for 16 h (4c and 6a), and conventional heating at 1108C for 19 h (4d) or 4 days (6b).
rolidinated MWNTs 4 and 5 are
significantly higher than that of
also the inner parts of the MWNT bundles to the reagents,
whereas only the outermost tubes of insoluble NT bundles
can react when mechanical treatment is lacking, as is the
case during MiW heating. In contrast, the sidewalls of the
soluble MWNT–ester 3 were efficiently functionalized with
pyrrolidine within highly convenient reaction times, in line
with our findings for MiW-assisted SWNT end functionaliza-
[11]
tion.
Characterization of sidewall-pyrrolidinated MWNT materi-
als: The IR spectra of the addition products (Figure 1) show
similarities with those of 1 or 3, although the features are
more pronounced, and signals from the added pyrrolidine
units are also present. As expected, the features of the
added group depend on the components of the azomethine
ylide. Neither the heating method nor the reaction tempera-
ture influence the structure of the added group; this result is
exemplified in Figure 1, which displays close to identical IR
features for 4a–d. This was also the case for 6, in which 6b,
obtained from a conventionally heated reaction, has an
almost identical spectrum to 6a, which was obtained from a
MiW-heated reaction. The intensities reflect the degree of
sidewall functionalization. The not-so-prominent peaks in 5
Figure 2. Raman spectra of MWNTs 1–3 and pyrrolidinated MWNTs 4
and 5 with intensities normalized to the D band at ~1315 cm . Inset
shows the D:G intensity ratio.
À1
À1
and the broad CÀO stretch peak at ~1100 cm in 5, 4a, and
the starting materials 3 and 1, respectively. Moreover, the
ratio increases with increasing reaction times as seen in the
MiW-assisted formation of 4 from 3.
4
d indicate a lower degree of sidewall functionalization
[24]
than in 4b, 4c, and 6.
All of the first-order Raman spectra (Figure 2) exhibit a
strong G band at ~1595 cm , which is the Raman allowed
The indications from IR and Raman spectroscopic meas-
urements of increasing functionalization with increasing re-
action times were verified by thermogravimetric analysis
(TGA), for which the weight loss during heating can be
used to estimate the relative mass of various groups attach-
ed to the material. The insoluble MWNTs 1 and 2 only ex-
hibited an inflection at ~4708C that indicates tube decom-
À1
phonon high-frequency E2g first-order mode, and a large
À1
disordered-induced (D) band at 1315 cm . The bands are
À1
shifted with respect to the 1580 cm (G), and 1346 or
À1
1
350 cm (D) bands reported for other carbon nanotube
[10c,25]
samples.
Since the positions of the bands are not
[10c,26]
changed during the functionalization process, the shift is
either due to inherent properties of the NT material or an
effect of the lasers used in the measurements. The similari-
ties between our spectra imply that neither esterification
nor sidewall functionalization seriously affect the bulk
graphite structure of the MWNTs, which already have a
high degree of defect sites before any cleaning or defect-tar-
geting oxidative cleaning. The slight differences in the rela-
tive intensities of the D and G bands (inset in Figure 2) give
indications of the extent of disorder introduced in the func-
tionalization steps. The D:G ratio is quite low in the cleaned
position (Figure 3),
whereas the weight loss from 3 be-
tween 200 and 3008C indicates an ester content of about
13%, which apparently is sufficient to solubilize the materi-
al well enough for efficient MiW-heated reactions to occur.
A reaction time of 2 h at 2008C yielded pyrrolidinated
MWNTs 4a with a total of 33 mass% added groups and a
sidewall addend/ester/MWNT ratio of 0.35:0.15:1 (Table 1).
Doubling the time at this temperature gave a product 4b
with 68% of non-NT groups, a value that is approaching sat-
uration. The 83% of non-NT groups present in 4c (16 h at
1308C) is likely to be a maximum under these conditions, as
Chem. Eur. J. 2006, 12, 3869 – 3875
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3871

H. Grennberg and J. Li
that of 4a, with a MiW reaction time of only 2 h. This last
observation is in good agreement with the similarities of the
D:G intensity ratio of the Raman spectra for 4a and 5.
The position and distribution profile of the units added to
the sidewalls were determined by transmission electron mi-
croscopy (TEM) by using the interaction between the thiol
[22a,27]
of the pyrrolidine in 4 and gold.
Treatment of 4 with
2
~6 nm-sized gold particles and thorough washing of the
sample on a filter membrane prior to TEM analysis verified
the success of the sidewall-functionalization reactions
(
Figure 4). For the first time, direct evidence for a rather
Figure 4. TEM images of a) 3–Au, b) 4b–Au, c) 4c–Au, and d) 4a–Au.
even distribution of the added groups over the MWNTs was
obtained. As a control, 3–Au, that is, 3 treated using the
same procedure as for the preparation of 4–Au, was ana-
lyzed. Remarkably, TEM results show a nearly complete ab-
sence of gold nanoparticles at the NTs, thus our washing
procedure is sufficiently effi-
Figure 3. TGA curves of 1, 2, 3, 4a, 4b, 4c, and 5 (top), and 6a and 6b
bottom).
(
cient to remove nanoparticles
Table 1. Relative amount and distribution of non-NT groups for MiW and conventionally heated (c.h.) pyrroli-
dinated MWNTs, as determined by TGA.
not in contact with a thiol unit.
Furthermore, the amount of
nanoparticles is lower in 4a–Au
than in 4b–Au and 4c–Au, an
observation that is in excellent
agreement with the results ob-
tained from IR and Raman
spectroscopy and TGA.
Sample
Wt%
Sidewall addends
Relative distribution
Sidewall/ester/MWNT
[a]
Organic total
Ester
MWNT
3
13
33
68
83
18
25
81
48
–
13
10
4.6
2.2
12.4
86
65
30
14
80
70
19
51
0/0.15/1
0.35/0.15/1
2.1/0.15/1
5.8/0.15/1
0.07/0.15/1
0.36/0/1
[
b]
4
4
4
4
5
6
6
a (MiW 2 h
)
)
23
63
81
5.6
25
78
40
[
b]
b (MiW 4 h
[
c]
c (MiW 16 h )
[
d]
d (c.h. 19 h
)
[
e]
(c.h. 5 d
)
[
c]
a (MiW 16 h )
2.9
7.9
4.1/0.15/1
0.79/0.15/1
[
d]
b (c.h. 4 d
)
[
a] Weight losses in the 200–4608C range are used to estimate the mass of organic groups attached to the
MWNTs. [b] 2008C. [c] 1308C. [d] 1108C (toluene). [e] 1308C (DMF).
this number is close to the yield of 81% of non-NT groups
obtained for 6a, prepared using a smaller amino acid glycine
and the same long-chain aldehyde. For comparison a sample
of 6b prepared using conventional heating for four days,
was found to contain 56 mass% of sidewall addends, which
is similar to the value obtained for 4b.
The influence of the solubility of the starting material
proved to be as expected: the conventionally heated pyrroli-
dination of insoluble 1 at 1308C for five days (120 h) yielded
Conclusion
Sidewall pyrrolidination of MWNTs by using 1,3-dipolar ad-
ditions of in situ formed azomethine ylides was achieved by
using both conventional and MiW-assisted heating. The
latter heating method was found to be highly efficient for
MWNTs 3 with solubilizing ester units, but was less useful
for the insoluble unit 1; this result is in contrast to previous-
[
18]
ly reported literature. The amount of sidewall functionali-
zation after MiW-assisted heating of 3 for 2–4 h at 2008C is
in the same range as that obtained from 3 or 1 after 4–5
days (100–120 h) of conventional heating. As the amount of
added groups in our conventionally heated reactions is in
5
with a pyrrolidine content lower than that in 6b. The prod-
uct 6b was prepared under similar reaction conditions, but
from a soluble starting material 3. Moreover, the mass%
sidewall addends/MWNT ratio of 5 is in the same range as
3872
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3869 – 3875

Functionalization of Carbon Nanotubes
FULL PAPER
the same range as that reported for similar additions, the
MiW-heating method is 50–60 times faster. The structure of
the added groups was not affected by the heating method,
thus MiW heating is a generally useful method, provided
the starting material is soluble. In addition, we have demon-
strated that the amount of addends can be controlled by al-
tering the reaction time and temperature, and that the
added groups are evenly distributed over the MWNTs even
at lower degrees of functionalization. Since the sidewall re-
actions only take place on the outermost layer of the
MWNTs, we anticipate that the electronic properties of the
inner layers of the functionalized MWNTs will be unaffect-
ed; this result is of interest for certain applications. In a
more general sense, the solubility and substrate-targeting
properties of the resulting MWNTs can be tailored by a
proper choice of added groups at the ends and sides of the
nanotubes to fit any desired application.
ethanol. The dark product was washed off the filter with ethanol and
dried for 12 h at a reduced pressure (vacuum oven at 1208C). The IR
spectrum of 1 exhibited no features indicating the presence of a carbonyl
group.
Formation of MWNT–COOH 2: The as-received MWNTs (100 mg) were
oxidatively cleaned and converted into the acid 2 by treatment with an
HNO :H SO (20 mL, 1:3 v/v) mixture at 508C for 5 h with continuous
3
2
4
ultrasonication (20 W). The resultant solid, MWNT–COOH 2, was col-
lected by filtration as described above. The IR spectrum of 2 shows
À1
À1
peaks at 1710 cm (carbonyl of acid) and a broad peak at 1100 cm (CÀ
[10,24]
O stretch).
Preparation of soluble MWNT–octadecyl ester 3: This ester was pre-
pared by the reaction of 2 with octadecyl bromide, inspired by a litera-
[
10c]
ture method.
MWNT–COOH 2 (200 mg) was sonicated in aqueous
NaOH (1 mm, 200 mL) for 2 min. Hexadecyltrimethylammonium bro-
mide (100 mg) and octadecylbromide (500 mg) were added to the result-
ing black homogeneous suspension of MWNT–COONa and the mixture
was refluxed with vigorous stirring. After 2 h, a black precipitate was
formed and the aqueous phase became clear and colorless. After the ad-
dition of CHCl
layer was collected, washed with aqueous NaCl (15%), and dried over
anhydrous CaSO . After a reduction of the volume, slow addition of the
3
and filtration to remove unreacted MWNTs, the organic
4
solution into ethanol yielded the MWNT–octadecyl ester 3 as a black
solid that was collected by filtration, washed several times with ethanol,
and dried in a vacuum oven at 1008C for 12–15 h. The IR spectrum
Experimental Section
À1
À1
shows peaks centered at 2917 and 2848 cm (alkyl CÀH), 1720 cm (car-
General: MWNTs (diameter range 10–30 nm) prepared by a chemical
vapor deposition (CVD) method by using a Ni–La complex oxide cata-
lyst were purchased from Shenzhen Nanotech Port (China). Cysteine hy-
drochloride monohydrate (Fluka) was neutralized and dried before use.
Other reagents were obtained from commercial sources and were used as
received. The solvents used were of analytical grade, and were dried and
À1
bonyl of ester), and broad peaks at 1560 and 1100 cm (corresponding to
[
10, 24] 1
C=C and CÀO, respectively).
olution due to residual paramagnetic impurity): d=0.89 (t, 3H), 1.0–1.7
m, 32H), 3.45 ppm (m, 2H).
3
H NMR (399.9 MHz, CDCl , poor res-
(
Synthesis of 4-octadecyloxybenzaldehyde: Under nitrogen protection, a
mixture of 4-hydroxybenzaldehyde (534 mg, 5.04 mmol), potassium car-
bonate (880 mg, 6.37 mmol), and octadecyl bromide (1681 mg,
5.05 mmol) in dry DMF (50 mL) was refluxed for 20 h. After cooling, the
solution was poured into NaOH (50 mL of a 5% aqueous solution); the
precipitate was collected by filtration and recrystallized from ethanol.
After drying under a reduced pressure for 15 h, 1.24 g (3.3 mmol, 66%)
of 4-octadecyloxybenzaldhyde was obtained as a slightly pinkish powder.
With a reaction time of 68 h, 1.60 g (4.3 mmol, 85%) of product was ob-
tained. M.p. 51–548C; IR (KBr): n=2918, 2849 (CÀH), 1693 (C=O), 1603
[
28]
distilled before use. Column chromatography was performed by using
Matrex silica gel 60A, 35–70 mm as the solid phase. Thin-layer chroma-
tography (TLC) was performed on Merck precoated silica gel 60-F254
plates.
MiW heating and sonication were performed by using Biotage Initiator
Sixty and Vibro Cell Sonics equipment, respectively. Centrifugation-as-
sisted filtration of MWNTs was carried out by using 500 mL polypropy-
lene centrifuge tubes fitted with a removable chamber with a polypro-
pene filter membrane, pore size 0.45 mm (Pall Life Sciences or Whatman
VectaSpin), in an Eppendorf MiniSpin centrifuge at 12 000 rpm for 10–
À1
1
(
C=C), ~1100 cm (CÀO); H NMR (399.9 MHz, CDCl
3
): d=0.88 (t,
H), 1.26 (m, 28H), 1.48 (m, 2H), 1.81 (m, 2H), 4.04 (t, 2H), 6.99 (d,
3
2
1
1
5 min. Liquid that was added to the chamber passed the membrane into
1
3
H), 7.83 (d, 2H), 9.87 ppm (s, 1H); C NMR (100.6 MHz, CDCl
4.33, 22.92, 26.19, 29.29, 29.58, 29.77, 29.80, 29.93 (from 29.29 to 29.93,
3
): d=
the lower compartment of the tube, whereas the insoluble material re-
mained at the filter. Cycles of solvent removal from the lower compart-
ment and addition of fresh solvent to the filter compartment could easily
wide peaks), 31.12, 32.16, 68.68, 114.99, 130.01, 132.20, 164.52,
190.99 ppm.
1
be repeated until the solids were considered sufficiently clean. H and
1
3
[29]
C NMR spectra were recorded on a Varian Unity 400 spectrometer
Preparation of N-octylglycine ethyl ester: N-octylglycine was prepared
1
13
(
399.9 and 100.6 MHzfor H and C, respectively). The chemical shifts
in a two-step process. A solution of ethyl bromoacetate (0.45 mL,
4.06 mmol) in dry dichloromethane (10 mL) was added to a solution of
octylamine (1.2 mL, 7.3 mmol) and potassium carbonate (355 mg,
2.57 mmol) in dry dichloromethane (10 mL) at 08C over a period of
30 min. The mixture was stirred at room temperature for 12–15 h and
was then washed with brine (3”10 mL), the organic phase was dried
were reported by using the residual solvent signal as an indirect reference
1
13
1
3 2
to TMS: CDCl 7.26 ppm ( H) and 77.0 ppm ( C), D O 4.79 ppm ( H).
FT-IR spectra were recorded for analytes in KBr tablets by using KBr as
the background on a Perkin–Elmer 1760 spectrometer. TGA was ob-
tained on a TA Instrument TGA Q500. Samples of about 5 mg were
À1
heated at a rate of 108Cmin under a flow of nitrogen. Raman spectra
over CaSO , and the solvents evaporated. The crude product was purified
4
were recorded on a Renishaw System 2000 micro-Raman spectrometer
equipped with a 783 nm diode laser (25 mW) with 2 cm resolution.
by flash chromatography with ethyl acetate as eluent yielding 308 mg
(35%) of N-octylglycine ethyl ester as a clear yellowish oil. H NMR
À1
1
Solid samples were placed on a stainless-steel plate during the measure-
ments. TEM was obtained on a Jeol2000FXII instrument operating at
(399.9 MHz, CDCl ): d= 0.80 (t, 3H), 1.20 (m, 13H), 1.43 (q, 2H), 1.56
3
1
3
(s, 1H), 2.49 (t, 2H), 3.29 (s, 2H), 4.14 ppm (q, 2H); C NMR d=13.98,
4.14, 22.56, 27.16, 29.16, 29,40, 30.00, 31.74, 49.59, 50.97, 60.55,
2
00 kV by using a 300 mesh holey carbon grid.
1
Pretreatment of as-received carbon nanotubes to remove residual non-
172.47 ppm.
nanotube material and form 1 and 2 was carried out by using modified
Preparation of N-octylglycine: A solution of N-octylglycine ethyl ester
[
9–11]
literature procedures:
Formation of cleaned MWNTs 1: The as-received MWNTs (typically
0 mg) were oxidatively cleaned by treatment with aqueous HNO (typi-
cally 25 mL of a 2.6m solution) at 1008C in an oilbath, with stirring for
(
526 mg, 2.45 mmol) and sodium hydroxide (214 mg, 5.35 mmol) in etha-
nol/water (1:1 v/v, 15 mL) was stirred at room temperature for 24 h.
Then, aqueous HCl (2m) was added while stirring until a neutral pH was
attained; the solvent was evaporated and the product dried under re-
duced pressure until a constant weight was measured (24 h), yielding
5
3
[
9f]
2
d. The mixture was then diluted with water (three times the reaction
volume) and the solids were collected by filtration and washed with
water, until the filtrate became pH neutral, and then repeatedly with
0.69 g of a solid material containing N-octylglycine and NaCl in an ap-
1
proximately 1:2 molar ratio. H NMR (399.9 MHz, D
2
O): d=0.93 (t,
Chem. Eur. J. 2006, 12, 3869 – 3875
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3873

H. Grennberg and J. Li
1
3
3
H), 1.35 (m, 10H), 1.63 (m, 2H), 3.10 (t, 2H), 3.66 (s, 2H); C NMR
alized MWNTs (4a, 4b, and 4c) and almost no Au at the sidewalls of the
control (3).
d=13.64, 14.43, 22.23, 25.76, 25.93, 28.42, 31.26, 47.85, 49.44, 171.83 ppm.
Addition of N-octyl-2-(4-octadecyloxyphenyl)pyrrolidine units to the in-
soluble MWNTs 1: this resulted in the formation of 5 by using adapted
[
13,14]
conventional heating methods, previously reported in the literature.
Acknowledgements
A mixture of insoluble MWNTs 1 (10 mg), an N-octylglycine and NaCl
mixture (194 mg, containing ~0.64 mmol N-octylglycine), and 4-
octadecyloxybenzaldhyde (561 mg, 1.5 mmol) in dry DMF (12 mL) was
heated at 1308C, while stirring, for 5 d. The resulting solid materials were
collected by centrifuge-assisted filtration on a polytetrafluoroethylene
This work was supported by the Swedish Foundation for Strategic re-
search within the CARAMEL consortium, the Swedish Research Coun-
cil, and the Gçran Gustafsson Foundation. We thank Prof. Jan Lindgren,
Dr. Jun Lu, and Mr. Basse Asplund, Uppsala University, for assistance
with Raman, TEM, and TGA measurements, respectively.
A
C
H
T
R
E
U
N
G
(
PTFE) filter membrane and washed with acetone and water (several
portions each, until the filtrate solution displayed only the solvent UV/
Vis absorption spectrum), and then dried in a vacuum oven at 1008C for
approximately 12 h, yielding 11–12 mg of a dark gray solid that was very
soluble in toluene, chloroform, and dichloromethane. TGA analysis indi-
cated 25 wt% of added groups; this value is in the same range as that
[
[
1] a) S. Iijima, Nature 1991, 354, 56–58; b) T. W. Ebbesen, H. J. Lezec,
H. Hiura, J. W. Bennett, H. F. Ghaemi, T. Thio, Nature 1996, 382,
5
4–56; c) H. Dai, J. H. Hafner, A. G. Rinzler, D. T. Colbert, R.
[
13c] 1
previously reported for SWNTs with similar added units.
CDCl , poor resolution due to residual paramagnetic impurity): d=0.89,
.0–1.7, 1.80, 3.95, 6.91, 8.01 ppm. TGA analysis: 25 mass% of pyrroli-
H NMR
Smalley, Nature 1996, 384, 147–150.
(
3
2] a) A. Hirsch, Angew. Chem. 2002, 114, 1933–1939; Angew. Chem.
Int. Ed. 2002, 41, 1853–1859; b) J. L. Bahr, J. M. Tour, J. Mater.
Chem. 2002, 12, 1952–1958; c) Special Issue On Carbon Nanotubes ,
Acc. Chem. Res. 2002, 35, 997–1113.
1
dine units.
Addition of 2-methylenethiol-4-(4-octadecyloxyphenyl)pyrrolidine units
to 3, forming 4: 4-Octadecyloxybenzaldhyde (200 mg, 0.535 mmol) and
cysteine (170 mg, 1.405 mmol) were added to a solution of the MWNT–
ester 3 (10 mg) in dry toluene (10 mL). The mixture was sonicated for
[3] W. A. Deheer, A. Chatelain, D. Ugarte, Science 1995, 270, 1179–
1180.
[4] T. Rueckes, K. Kim, E. Joselevich, G. Y. Tseng, C. Cheung, C. M.
Lieber, Science 2000, 289, 94–97.
5
min, then nitrogen gas was bubbled through the solution for 10 min
prior to MiW heating at 2008C for 2 or 4 h or at 1308C for 16 h (4a, b, c,
respectively), or conventional heating at the reflux temperature for 19 h
[5] J. Kong, N. R. Franklin, C. W. Zhou, M. G. Chapline, S. Peng, K. J.
Cho, H. J. Dai, Science 2000, 287, 622–625.
(
4d), after which time acetone was added. The solid material was collect-
[6] R. H. Baughman, C. X. Cui, A. A. Zakhidov, Z. Lqbal, J. N. Barisci,
G. M. Spinks, G. G. Wallace, A. Mazzoldi, D. De Rossi, A. G. Rin-
zler, O. Jaschinski, S. Roth, M. Kertesz, Science 1999, 284, 1340–
1344.
ed by centrifuge-assisted filtration on a PTFE filter membrane and
washed with acetone and water (several portions each, until the filtrate
solution displayed only the solvent UV/Vis absorption spectrum).
[
7] a) O. Matarredona, H. Rhoads, Z. Li, J. H. Harwell, L. Balzano,
D. E. Resasco, J. Phys. Chem. B 2003, 107, 13357–13367; b) J. E.
Riggs, D. B. Walker, D. L. Carroll, Y.-P. Sun, J. Phys. Chem. B 2000,
104, 7071–7076.
8] a) E. T. Mickelson, I. W. Chiang, J. L. Zimmerman, P. J. Boul, J.
Lozano, J. Liu, R. E. Smalley, R. H. Hauge, J. L. Margrave, J. Phys.
Chem. B 1999, 103, 4318–4322; b) J. Chen, M. A. Hamon, H. Hu, Y.
Chen, A. M. Rao, P. C. Eklund, R. C. Haddon, Science 1998, 282,
The products were dried in a vacuum oven at 1008C for 12–15 h yielding
1
1 mg of 4a, 22 mg of 4b, 42 mg of 4c, and 10 mg of 4d as dark gray
solids, which were very soluble in toluene, chloroform, and dichlorome-
À1
1
thane. IR (KBr): n=2917, 2848, 2590, 1720, 1635, 1254 cm
CDCl , poor resolution due to residual paramagnetic impurity): d=0.89,
.0–1.7, 1.77, 3.45, 3.97, 6.88, 7.40 ppm; TGA analysis: mass% of pyrroli-
; H NMR
[
(
1
3
dine units: 4a 23%, 4b 63%, 4c 81%, and 4d 18%.
Addition of 2-(4-octadecyloxyphenyl)pyrrolidine units to 3, forming 6:
The MWNT–octadecyl ester 3 (10 mg) was treated with 4-octadecyloxy-
benzaldhyde (224 mg, 0.6 mmol) and glycine (287 mg, 3,83 mmol) in a
solvent (10 mL) following, for 6a, the same procedure and workup as de-
scribed for the preparation of 4c (toluene, MiW heating, 1308C, 16 h),
or, for 6b, a similar preparation procedure as that used for 4d (1108C,
conventional heating, toluene, 4 d). The product was very soluble in tol-
uene, chloroform, and dichloromethane. IR (KBr): n=2917, 2849, 1609,
9
5–98.
[
9] a) J. Liu, A. G. Rinzler, H. Dai, J. H. Hafner, R. K. Bradley, P. J.
Boul, A. Lu, T. Iverson, K. Shelimov, C. B. Huffman, F. Rodriguez-
Macias, Y. Shon, T. R. Lee, D. T. Colbert, R. E. Smalley, Science
1
998, 280, 1253–1256; b) S. S. Wong, E. Joselevich, A. T. Woolley,
C. L. Cheung, C. M. Lieber, Nature 1998, 394, 52–55; c) S. S. Wong,
A. T. Woolley, E. Joselevich, C. L. Cheung, C. M. Lieber, J. Am.
Chem. Soc. 1998, 120, 8557–8558; d) M. A. Hamon, J. Chen, H. Hu,
Y. Chen, M. E. Itkis, A. M. Rao, P. C. Eklund, R. C. Haddon, Adv.
Mater. 1999, 11, 834–840; e) J. Chen, A. M. Rao, S. Lyuksyutov,
M. E. Itkis, M. A. Hamon, H. Hu, R. W. Cohn, P. C. Eklund, D. T.
Colbert, R. E. Smalley, R. C. Haddon, J. Phys. Chem. B 2001, 105,
À1
1
1
261 cm (both 6a and 6b); H NMR (CDCl
sidual paramagnetic impurity): d=0.89, 1.27, 1.47, 1.79, 3.46, 3.96, 6.92,
.35 ppm (both 6a and 6b); TGA: mass% of pyrrolidine units: 6a 81%
3
, poor resolution due to re-
7
and 6b 62%.
Solubilized Au nanoparticles: Nanoparticles of diameter 2–6 nm, as seen
2
525–2528; f) J. Zhang, H. Zou, Q. Qing, Y. Yang, Q. Li, Z. Liu, X.
[
27]
by TEM, were prepared according to the method of Brust et al. Brief-
ly, an aqueous solution of HAuCl (50 mm, 3 mL) was added to a solution
Guo, Z. Du, J. Phys. Chem. B 2003, 107, 3712–3718, and references
therein; g) H. Kajiura, S. Tsutsui, H. Huang, Y. Murakami, Chem.
Phys. Lett. 2002, 364, 586–592; h) E. Farkas, M. E. Anderson, Z.
Chen, A. G. Rinzler, Chem. Phys. Lett. 2002, 363, 111–116.
4
of hexadecyltrimethylammonium bromide in toluene(50 mm, 8 mL). The
mixture was stirred at room temperature for 30 min, then a freshly pre-
pared aqueous solution of NaBH (0.5m, 2 mL) was added dropwise with
4
stirring over 15 min; stirring was continued for a further 30 min. After
standing at 48C for at least 12 h, the aqueous phase was removed. The
Au-containing toluene phase was stored in a sealed container at 48C.
[
10] a) Y.-P. Sun, W. Huang, Y. Lin, K. Fu, A. Kitaygorodskiy, L. A.
Riddle, Y. Yu, D. L. Carroll, Chem. Mater. 2001, 13, 2864–2869;
b) M. A. Hamon, H. Hui, P. Bhowmik, H. M. E. Itkis, R. C. Haddon,
Appl. Phys. A 2002, 74, 333–338; c) Y. Qin, J. Shi, W. Wu, X. Li, Z.-
MWNT–Au assemblies: These assemblies and controls were prepared by
dropwise addition of the solution of solublilized Au nanoparticles in tol-
uene (typically 0.5 mL) over 10–15 min to a solution of 3, 4a, 4b, or 4c
X. Guo, D. Zhu, J. Phys. Chem. B 2003, 107, 12899–12901;
d) M. G. C. Kahn, S. Banerjee, S. S. Wong, Nano Lett. 2002, 2, 1215–
1218.
(
1.4 mg in 10 mL). The resulting mixture was then stirred at room tem-
[11] K. Chajara, J. Lindgren, H. Grennberg, unpublished results.
[12] a) M. Maggini, G. Scorrano, M. Prato, J. Am. Chem. Soc. 1993, 115,
9798–9799; b) M. Prato, M. Maggini, G. Scorrano, Synth. Met. 1996,
77, 89–91.
[13] a) V. Georgakilas, K. Kordatos, M. Prato, D. M. Guldi, M. Holzing-
er, A. Hirsch, J . Am. Chem. Soc. 2002, 124, 760–761; b) V. Georga-
perature for 30 min, during which time a precipitate formed. The solid
material was collected by centrifuge-assisted filtration on a PTFE filter
membrane and washed repeatedly with toluene until the washing solution
displayed only the solvent UV/Vis absorption spectrum. TEM analysis
showed substantial amounts of Au at the sidewalls for the thiol-function-
3874
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3869 – 3875

Functionalization of Carbon Nanotubes
FULL PAPER
kilas, N. Tagmatarchis, D. Pantarotto, A. Bianco, J.-P. Briand, M.
Prato, Chem. Commun. 2002, 3050–3051; c) V. Georgakilas, D.
Voulgaris, E. Vazquez, M. Prato, D. M. Guldi, A. Kukovecz, H. Kuz-
many, J. Am. Chem. Soc. 2002, 124, 14318–14319; d) D. M. Guldi,
M. Marcaccio, D. Paolucci, F. Paolucci, N. Tagmatarchis, D. Tasis, E.
Vazquez, M. Prato, Angew. Chem. 2003, 115, 4338–4341; Angew.
Chem. Int. Ed. 2003, 42, 4206–4209.
[21] a) K. Kamaras, M. E. Itkis, H. Hu, B. Zhao, R. C. Haddon, Science
2003, 301, 1501–1501; b) H. Hu, B. Zhao, M. A. Hamon, K. Kama-
ras, M. E. Itkis, R. C. Haddon, J. Am. Chem. Soc. 2003, 125, 14893–
14900.
[22] a) T. Sainsbury, J. Stolarczyk, D. Fitzmaurice, J. Phys. Chem. B 2005,
109, 16310–16325; b) W. Wu, S. Wieckowski, G. Pastorin, M. Benin-
casa, C. Klumpp, J-P. Briand, R. Gennaro, M. Prato, A. Bianco,
Angew. Chem. 2005, 117, 6516–6520; Angew. Chem. Int. Ed. 2005,
44, 6358–6362.
[
[
14] a) Z. Yao, N. Braidy, G. A. Botton, A. Adronov, J. Am. Chem. Soc.
003, 125, 16015–16024; b) L. S. Cahill, Z. Yao, A. Adronov, J.
2
Penner, K. R. Moonoosawmy, P. Kruse, G. R. Goward, J. Phys.
Chem. B 2004, 108, 11412–11418.
15] a) F. Langa, P. De la Cruz, E. Espildora, A. De la Hoz, Proc. Elec-
trochem. Soc. 2000, 9, 168–178; b) U. M. Fernandez-Paniagua, B. Ill-
escas, N. Martin, C. Seoane, P. de la Cruz, A. de la Hoz, F. Langa, J.
Org. Chem. 1997, 62, 3705–3710; c) C. O. Kappe, Angew. Chem.
[23] J. W. Lown, 1,3-Dipolar Cycloaddition Chemistry, Wiley, New York,
1984.
[24] Spectroscopic Methods In Organic Chemistry (Eds. M. Hesse, H.
Meier, B. Zeeh), Thieme, Stuttgart, New York, 1997.
[25] H. Hiura, T. W. Ebbesen, K. Tanigaki, H. Takahashi, Chem. Phys.
Lett. 1993, 202, 509–512.
2
004, 116, 6408–6443; Angew. Chem. Int. Ed. 2004, 43, 6250–6284.
[26] a) M. Alvaro, C. Aprile, P. Atienzar, H. Garcia, J. Phys. Chem. B
2005, 109, 7692–7697; b) D. Bom, R. Andrews, D. Jacques, J. Antho-
ny, B. L. Chen, M. S. Meier, J. P. Selegue, Nano Lett. 2002, 2, 615–
619.
[
16] a) S. Caddick, Tetrahedron 1995, 51, 10403–10432; b) A. Loupy, A.
Petit, J. Hamelin, F. Texier-Boullet, P. Jacquault, D. Mathꢁ, Synthesis
1
998, 1213–1234; c) W. Zhang, C. H.-T. Chen, Y. Lu, T. Nagashima,
Org. Lett. 2004, 6, 1473–1476.
[27] M. Brust, M. Walker, D. Bethell, D. J. Schiffrin, R. Whyman, J.
Chem. Soc., Chem. Commun. 1994, 801–802.
[28] D. D. Perrin, W. L. F. Armarego, Purification Of Laboratory Chemi-
cal, 3rd ed., Pergamon, Oxford 1988.
[29] a) P. Sofou, Y. Elemes, E. Panou-Pomonis, A. Stavrakoudis, V. Tsi-
karis, C. Sakarellos, M. Sakarellos-Daitsiotis, M. Maggini, F. For-
maggio, C. Toniolo, Tetrahedron 2004, 60, 2823–2828; b) P. Wang, B.
Chen, R. M. Metzger, T. Da Tos, M. Prato, J. Mater. Chem. 1997, 7,
2397–2400.
[
17] a) C. J. Ko, C. Y. Lee, F. H. Ko, H. L. Chen, T. C. Chu, Microelec-
tron. Eng. 2004, 73–74, 570–577; b) E. Vazquez, V. Georgakilas, M.
Prato, Chem. Commun. 2002, 2308–2309; c) M. T. Martinez, M. A.
Callejas, A. M. Benito, W. K. Maser, M. Cochet, J. M. Andres, J.
Schreiber, O. Chauvet, J. L. G. Fierro, Chem. Commun. 2002, 1000–
1
001.
[
18] a) Y. Wang, Z. Iqbal, S. Mitra, Carbon 2005, 43, 1015–1020; b) J. L.
Delgado, P. de la Cruz, F. Langa, A. Urbina, J. Casado, J. T. Lopez
Navarrete, Chem. Commun. 2004, 1734–1735.
[
[
19] K. Nordlund, P. Hakonen, Nat. Mater. 2005, 4, 514–515.
20] J. Zhao, Z. Chen, Z. Zhou, H. Park, P. von R. Schleyer, J. Lu, Chem-
PhysChem 2005, 6, 598–601.
Received: October 21, 2005
Published online: February 27, 2006
Chem. Eur. J. 2006, 12, 3869 – 3875
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3875