Preparation and properties of vesicles made of nonpolar/polar/nonpolar fullerene amphiphiles

Article abstract of DOI:10.1021/ja200498g

Twenty potassium complexes of penta-[(4-substituted)phenyl][60]fullerene anions were synthesized and examined for their ability to form bilayer vesicles in water. The 4-substituents include alkyl groups ranging from methyl to icosanyl groups and perfluoromethyl, perfluorobutyl, and perfluorooctyl groups. The overall structure of the amphiphiles can be described as a nonpolar/polar/nonpolar (n-p-n′) motif as opposed to the usual polar/nonpolar motif of lipid amphiphiles. Despite the hydrophobicity of the fullerene moiety (n-part) and alkyl/perfluoroalkyl chains (n′-part), all compounds except for the one with perfluoromethyl groups were soluble in water because of the centrally located fullerene cyclopentadienide (p-part) and spontaneously formed a vesicle of 25- to 60-nm diameter with a narrow unimodal size distribution. The vesicles are stable upon heating to 90 ° or standing over one year in air, as well as on a solid substrate in air or in vacuum, maintaining their spherical form. The vesicle membrane consists of an interdigitated bilayer of the amphiphile molecules, in which the fullerene n-part is inside and the n′-side is exposed to water. These vesicles, in particular the one bearing icosanyl chains, exhibit the smallest water permeability coefficient ever found for a self-assembled membrane in water.

Full text of DOI:10.1021/ja200498g

ARTICLE
pubs.acs.org/JACS
Preparation and Properties of Vesicles Made of Nonpolar/Polar/
Nonpolar Fullerene Amphiphiles
Tatsuya Homma,† Koji Harano,† Hiroyuki Isobe,‡ and Eiichi Nakamura*,†
†Department of Chemistry, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
‡Department of Chemistry, Tohoku University, Aoba-ku, Sendai 980-8578, Japan
S Supporting Information
b
ABSTRACT: Twenty potassium complexes of penta-
[(4-substituted)phenyl][60]fullerene anions were synthesized and
examined for their ability to form bilayer vesicles in water. The
4-substituents include alkyl groups ranging from methyl to icosanyl
groups and perfluoromethyl, perfluorobutyl, and perfluorooctyl
groups. The overall structure of the amphiphiles can be described
as a nonpolar/polar/nonpolar (nꢀpꢀn0) motif as opposed to the
usual polar/nonpolar motif of lipid amphiphiles. Despite the hydro-
phobicity of the fullerene moiety (n-part) and alkyl/perfluoroalkyl chains (n0-part), all compounds except for the one with
perfluoromethyl groups were soluble in water because of the centrally located fullerene cyclopentadienide (p-part) and
spontaneously formed a vesicle of 25- to 60-nm diameter with a narrow unimodal size distribution. The vesicles are stable upon
heating to 90 °C or standing over one year in air, as well as on a solid substrate in air or in vacuum, maintaining their spherical form.
The vesicle membrane consists of an interdigitated bilayer of the amphiphile molecules, in which the fullerene n-part is inside and
the n0-side is exposed to water. These vesicles, in particular the one bearing icosanyl chains, exhibit the smallest water permeability
coefficient ever found for a self-assembled membrane in water.
olar head/nonpolar tail is the archetypal structural motif of
lipids that plays crucial roles in the formation of compart-
We report in this article that the nꢀpꢀn0 motif of the
amphiphile is very general, as examined for 18 new fullerene
anions bearing methyl to icosanyl, perfluoromethyl, and per-
P
ment structures both in biology1 and in man-made systems.2 In
water, the polar heads of the lipid molecule align toward the
aqueous environment, while the tails cluster together to form a
bilayer membrane. The polymorphic behavior of lipid amphi-
philes is a typical characteristic of the lipid bilayer and endows the
bilayer with properties characteristic of cell membranes such as
mechanical stability and water permeability.3 The question arises
of whether the polar head/nonpolar tail motif is the only motif of
the molecules that form a bilayer membrane in water. We made a
discovery some time ago that a fullerene anion bearing five
phenyl groups on the top side of the molecule (PhK, Figure 1a)
in water spontaneously forms a bilayer vesicle that shows a
unimodal distribution with a diameter centering at ca. 30 nm4
and exhibits extremely low water permeability (Figure 1c and
1d).5 Although the structure is quite unusual, PhK still has a
polar/nonpolar motif (Figure 1b), and the cyclopentadienyl
anion and the phenyl groups are exposed to the vesicle surface
(Figure 1c and d). With this vesicle structure in mind, we asked
whether we could install a hydrophobic chain on each phenyl
group (cf. Figure 1e) to create a nonpolar/polar/nonpolar
(nꢀpꢀn0) motif, and whether such a molecule would still form
i
t
fluorobutyl chains (compounds C1KꢀC13K, C3K, C4K,
C20K, Rf1K, and Rf4K, respectively). All compounds except
Rf1K formed vesicles with a diameter of 20ꢀ60 nm simply by
injection of their THF solution into water, and the vesicles
exhibited a variety of properties that are unknown for lipid
vesicles. For instance, the size distribution of the vesicle is
unimodal without any purification, and water permeation
through the vesicle membrane is exceedingly slow. Spin coating
of the vesicle solution on a hydrophilic solid substrate of mica and
indium tin oxide (ITO) dispersed the spherical vesicle particles
uniformly on the surface with retention of the spherical shape, in
contrast to lipid vesicles.1 All available evidence indicates that the
interdigitated bilayer in the nꢀpꢀn0 fullerene vesicle is a hard
shell rather than a flexible membrane.
’ RESULTS AND DISCUSSION
Synthesis of Fullerene Derivatives and Vesicle Formation
in Water. The protio fullerene compounds (CnH and RfnH)
were synthesized by 5-fold addition of a 4-substituted phenyl-
copper reagent to [60]fullerene (58ꢀ99%) as shown in
a bilayer vesicle in water. We previously communicated6
a
remarkable observation that this type of anion, a fullerene anion
bearing five perfluorooctyl chains Rf8K (Figure 1a), indeed
forms a vesicle and that its surface is covered by fluorous chains.
Received: January 18, 2011
Published: April 01, 2011
r
2011 American Chemical Society
6364
dx.doi.org/10.1021/ja200498g J. Am. Chem. Soc. 2011, 133, 6364–6370
|

Journal of the American Chemical Society
ARTICLE
Figure 1. Schematic images of vesicles made of fullerene anions. (a) Chemical structures of potassium complexes of fullerene anions. (b) CPK drawing
of PhK. (c) Schematic model of bilayer vesicle of PhK. (d) A magnified image of the bilayer PhK vesicle. (e) CPK drawing of C20K. (f) Schematic model
of a vesicle of C20K. (g) A magnified image of the interdigitated bilayer of the C20K vesicle. The fullerene parts are shown in green, the hydrophilic parts
in blue, and substituents in yellow.
Scheme 1. Preparation of Fullerene Vesicles
Figure 2. A solution of fullerene pentaadduct and the corresponding
potassium salt in a triphase system of toluene (top phase), water (middle
phase), and C8F18 (bottom phase) and the size distribution determined
by DLS (carried out in pure solvent). PhH dissolves in toluene, Rf8K
dissolves in water, and Rf8H dissolves in C8F18 but only sparingly in
toluene.
protio compound Rf8H is insoluble in water, sparingly soluble in
toluene, and soluble in perfluorooctane (C8F18) as an aggregate
with a diameter of 8.1 nm as determined by dynamic light
Scheme 1.7 They were treated with t-BuOK in THF to form a
homogeneous solution of CnK and RfnK, which was injected
into water (followed by removal of THF under reduced pres-
sure) to form a homogeneous vesicle solution of the potassio
fullerene. The vesicle solution thus prepared is orange, homo-
geneous, clear without any purification, and stable for over a year
under air on a shelf.
scattering (DLS; Figure 2), perhaps micelles that expose the
fluorous chain on the outer surface. In contrast, the potassio
compound Rf8K dissolved in THF as a monomer (ca. 1 nm in
diameter by DLS), suggesting that the solubility in THF is gained
by formation of a solvent-separated ion pair, as was found
previously in the crystal structure of PhK.8 The potassium
complex Rf8K dissolves quite well in water (up to 10 g/L at
25 °C) to form an orange solution of vesicles with a diameter of
36.2 nm (Figure 2). Notably, the lithium and sodium complexes
The protio and the potassio fullerenes behave very differently
from each other. Rf8H and Rf8K are illustrative of this fact. The
6365
dx.doi.org/10.1021/ja200498g |J. Am. Chem. Soc. 2011, 133, 6364–6370

Journal of the American Chemical Society
ARTICLE
Figure 3. Diameter of fullerene vesicles in water determined by DLS measurement at 25 °C. The data with error bars were determined by at least three
vesicle preparations. The data for PhK and Rf8K agree with the values reported in refs 4 and 6.
Figure 4. Schematic illustrations of two membrane structures. (a)
Archetypal bilayer where the area per one fullerene molecule is about
0.6 nm2 and (b) an interdigitated bilayer where the area is approximately
twice as large.
Figure 5. Schematic model of the bilayer showing a solvent-separated
ion-pair structure. Alkyl- or perfluoroalkyl-substituted fullerenes in gray,
potassium cation in pink separated from the cyclopentadienide anion on
fullerene in blue, and water in light blue.
precipitated in water, indicating that the potassium cation is
responsible for the high solubility of Rf8K in water. Rf4K
behaves similarly to Rf8K, but the vesicle of Rf1K is unstable
and precipitates in water and hence was not examined further.
The 16 alkylphenyl compounds CnK dissolve well in water,
like Rf4K and Rf8K, to form a vesicle solution of up to 2ꢀ4 mM
concentration. However, the solubility of the protio compounds
CnH depended markedly on the alkyl group. For instance, C20H
dissolves well both in toluene and in hexane, whereas PhH
dissolves in toluene but not in hexane.
Laser Light Scattering (LLS) Studies of Fullerene Vesicles
in Water. The 19 vesicle-forming potassium complexes (PhK,
CnK, Rf4K, and Rf8K) showed a unimodal diameter distribution
ranging between 20 and 60 nm as judged by DLS (Figure 3). The
vesicle size was found not to change much after one year on
a shelf or over 1 h at 90 °C. Zeta potentials are negative (ꢀ30
to ꢀ50 mV), which suggests that the vesicle exists as a solvent-
separated ion pair in water.
1.22 nm2/molecule area found in the crystal of a structurally similar
penta-4[(trimethylsilyl)ethynylphenyl]fullerene compound.10,11
We consider that the potassium cation located in the aqueous
phase endows water solubility to the vesicles as illustrated in
Figure 5 on the basis of the following experimental results. The
first is the negative zeta potentials of the vesicle surface, suggest-
ing that the amphiphile exists as a solvent-separated ion pair.
Second, the reported crystal structures of PhK indicated that the
bulky five phenyl groups force the bulky potassium ion away from
the fullerene’s anion center, prevents the formation of a contact
ion pair, and consequently creates a solvent-separated ion pair.8
The fact that lithium and sodium analogs of PhK do not dissolve
in water suggests that these metals form a contact ion pair and do
not contribute to the increase of the water solubility.
NMR Studies of Fullerene Vesicles in Water. Neither the
C8K nor the Rf8K vesicle in D2O exhibited any signals in the
1H, 13C, or 19F NMR spectra at 25 and 80 °C. On the other hand,
the 1H NMR spectra of the C20K vesicle showed no signals below
40 °C but suddenly showed, at 50 °C or higher, the terminal CH3
and CH2 signals (Figure 6), a behavior suggesting the melting of
the alkyl chains. This behavior is entirely different from that of
lipid molecules in a bilayer membrane, where the alkyl chains
To determine the structure of the vesicle further, we analyzed
the Rf8K vesicle by static light scattering (SLS). For a solution
that gave us a hydrodynamic radius (Rh; by DLS) of 26.7 nm, the
SLS analysis gave us a radius of gyration (Rg) of 26.7 nm. The Rg/
Rh ratio being 1.00, we conclude that the vesicle is hollow.9 The
Zimm square plot gave a molecular weight of 1.94 ꢁ 107 Da,
indicating that a vesicle contains 5.98 ꢁ 103 molecules of Rf8K.
From this number and the surface area of a vesicle, we obtain an
area of 1.39 nm2/molecule, which is twice as large as the
0.59 nm2/molecule area of the PhK vesicle (an archetypal bilayer
shown in Figures 4a and 1d).4b We consider that Rf8K forms an
interdigitated bilayer (Figures 4b and 1g) on the basis of the
1
give H NMR signals at room temperature because they are
mobile.12 The CH2/CH3 signal ratio changed from 24:3
(50 °C), 30:3 (60 °C), to 32:3 (70 °C) (Figure S3). Thus,
the terminal groups are more mobile than the internal CH2
6366
dx.doi.org/10.1021/ja200498g |J. Am. Chem. Soc. 2011, 133, 6364–6370

Journal of the American Chemical Society
ARTICLE
Figure 6. 1H NMR spectra of C20K vesicle in D2O at (a) 20 and (b) 70 °C.
Figure 8. Illustration of decomposition of the PhK vesicle on HOPG.
Green balls represent the fullerene core, and the blue region represents
the hydrophilic cyclopentadienide part.
When drop-cast on mica, the vesicles of PhK, CnK (n = 4, 8,
12, 20; 2 mM as for the fullerene molecule), and Rf8K retained
their spherical shape and became uniformly dispersed without
the formation of domains or islands. The density of the vesicles
was found to be rather uniform: 63, 80, 75, 100, 85, and 60
vesicles/μm2, respectively (see Supporting Information).
High-resolution SEM analysis of the Rf8K vesicle on a
semiconducting ITO surface afforded an interesting piece of
information on the manner in which the vesicles interact with the
solid substrate. The low-magnification top view in Figure 9a
shows that the vesicle particles uniformly cover a wide area of the
ITO substrate. A closer look (Figure 9b) reveals that the vesicles
sit on the grain boundaries of the ITO surface to maximize
contact with the surface. The 30°-tilted side view (Figure 9c) not
only confirms this observation but also indicates that the vesicles
largely retain their spherical structures under the rather harsh
SEM conditions of 10ꢀ5 Pa, which attests to the unusual
robustness of the fullerene vesicle. The average diameter of the
Rf8K vesicle particles was found to be 35.6 ( 0.4 nm as measured
for 190 vesicles (Figure 9a), which matched very well with the
DLS diameter (36.2 ( 0.2 nm, Figure 2).6
It is interesting to note that we could not fill the ITO grain
boundary completely with the vesicle particles (94% of the 545
particles are separated from each other) and the surface coverage
remained at the rather constant value of 228 ( 14 particles/μm2
as examined under two independently prepared samples by spin
coating of a 2 mM Rf8K solution. Possibly, the vesicle particles
recognize some very special structures of the grain boundary.
Contact Angle of Water on Vesicle-Coated Mica and Soda-
Lime Glass Surfaces. The coverage of the hydrophilic surface of
mica and soda-lime glass with the hydrophobic vesicles was
found to make the surface water-repellent, as examined for the
vesicles of PhK, CnK (n = 4, 8, 12, 20), and Rf8K.
Figure 7. AFM images of the HOPG surface after drop-casting of a
2 mM solution of C20K and PhK. The cross-section profiles along the
light-blue line in each figure are shown below the AFM images. (a) The
HOPG surface covered by C20K vesicles. Note that the diameter of the
vesicle in the lateral direction is greatly overestimated because of the
intrinsic limitations of AFM. (b) The HOPG surface after drop-casting
of PhK, where we could not identify any vesicular objects and instead
found ill-defined objects with 1- to 2-nm heights.
groups, which agrees with our proposal that alkyl groups are
exposed to the aqueous phase.
Microscopic Studies of Fullerene Vesicles on Highly Or-
iented Pyrolytic Graphite (HOPG), Mica, and ITO. Microscopic
studies of the fullerene vesicles on a solid surface revealed that
these vesicles are nearly perfectly spherical and are robust enough
to retain their shape in air or in vacuum. This behavior is entirely
unknown for lipid vesicles, which are flexible and lose their
structural integrity upon drying or in vacuum.
When we drop-cast the C20K vesicle on HOPG and examined
it by atomic force microscopy (AFM) under air, we saw
uniformly dispersed spherical particles having a height of 20 to
30 nm (Figure 7a). The height agrees well with the diameter
determined by DLS. Interestingly, the C20K vesicles on
HOPG formed a uniform monolayer without forming do-
mains or multilayers while the PhK vesicle showed ill-defined
objects with 1- to 2-nm heights (Figure 7b). We ascribe the
latter observation to the decomposition of the vesicle struc-
ture because of the more favorable interaction of the fullerene
part of the PhK molecule with HOPG than the hydrophilic
cyclopentadienide part that was exposed to the aqueous phase
before HOPG is coated (Figure 8).
As summarized in Figure 10 for mica (blue) and glass (red),
the contact angle properties may be classified into three classes,
6367
dx.doi.org/10.1021/ja200498g |J. Am. Chem. Soc. 2011, 133, 6364–6370

Journal of the American Chemical Society
ARTICLE
Figure 10. (a) Contact angles of water on a vesicle-coated mica surface
(blue) and a vesicle-coated glass surface (red). (b) Photographs of water
droplet (3 μL) on vesicle-coated mica surface.
magnitude smaller than known values for lipid vesicles. The very
slow water permeation is due to a very large negative entropy
barrier that has been ascribed to the capture of gaseous water
molecules by fullerenes,14 accompanied by a large negative
activation enthalpy (Figure 11c, leftmost column). Note that
the barrier for water permeation through a lipid membrane is due
to a positive activation enthalpy (a barrier for water
gasification).1,3,15 In addition, the permeability peaks at ca.
40 °C and decreases toward 80 °C where the vesicle still
maintains its structural integrity (see Figure 11a, blue line
bending at ca. 40 °C). This may be ascribed to a phase transition
characteristic of fullerene derivatives in the solid state, as found
for some fullerene derivatives.16
Comparison of the permeability properties of the PhK vesicle
with those of the C4K, C8K, C12K, C20K, and Rf8K vesicles
first indicated that the C4K vesicle is several times more leaky
than the PhK vesicle and the permeability decreases as the alkyl
chain becomes longer or perfluorooctyl groups are introduced
(Figure 11b). To our knowledge, the permeation constant of P =
(1.69 ( 0.10) ꢁ 10ꢀ7 m/s is the smallest value reported for a
self-assembled membrane.2b It is reasonable to assume that the
membrane becomes leaky because of the steric bulk of the p-
alkylphenyl groups (i.e., n0-group), which loosens the molecular
packing (see above for the increase in the area occupancy per
molecule as it goes from PhK to Rf8K, and the likely change of
the molecular packing as illustrated in Figure 4), but becomes less
leaky as the hydrophobicity of the n0-group becomes larger as the
chain becomes longer. We consider that the low water perme-
ability of C20K is due to the hydrophobic effects of the long alkyl
chains and the enhanced stability of the membrane caused by the
intermolcular van der Waals interactions. Interestingly, the
permeability constant P in Figure 11b is inversely related to
the magnitude of the contact angle shown in Figure 10. It is
possible that the contact angle reflects the properties of the
fullerene bilayer, because the contact angle has been reported to
reflect the properties of a self-assembled monolayer up to a depth
of about 2 nm.17
Figure 9. SEM images of Rf8K vesicles on ITO under ca. 10ꢀ5 Pa. (a)
SEM image of Rf8K vesicles on the ITO surface. (b) Magnified view of
Figure 9a. (c) A different area of the same sample as viewed with a 30° tilt
of the sample stage.
PhK, CnK, and RfnK. For the CnK vesicles, the contact angle on
mica for the alkyl substituted fullerenes increased as the
hydrophobicity (alkyl chain length) increased: C4K (14.2 (
2.0)° < C8K (27.3 ( 4.1)° < C12K (52.0 ( 4.2)° < C20K
(85.9 ( 3.6)°. Once the surface was dried, the surface
remained water-repellent even after repeated rinsing with
water and after storage in air at room temperature for several
months. The behavior of the PhK vesicle is similar to that of
C8K, while that of Rf8K is similar to that of C20K. We surmise
that the drying changed the surface structure of the vesicle.
One possibility would be that dehydration of the potassium
cation pushed the metal atom from the surface of the vesicle
into the membrane interior.
The permeability curve (Figure 11a) peaks at ca. 40 °C for the
PhK and C4K vesicles, at 30 °C for Rf8K, and at 20 °C or below
for C8K, and no peak appears for C20K. This trend agrees with
the profile of activation enthalpy and entropy, where both
parameters become less sensitive as the chain becomes longer
from C12 to C20 (Figure 11c). We surmise that the fullerene part
determines the profile of water permeability in the PhK vesicle
while the hydrophobicity of the alkyl chains contributes to the
Water Permeability of Fullerene Vesicles. Water permea-
tion is one of the fundamental properties of lipid and surfactant
membranes,1,3,12 and hence those of the C4K, C8K, C12K,
C20K, and Rf8K vesicles were studied by the 17O NMR method
reported for the PhK vesicle.5,13
The PhK vesicle exhibits unusual kinetics and very low
water permeability.5 The permeation constant P is 3 orders of
6368
dx.doi.org/10.1021/ja200498g |J. Am. Chem. Soc. 2011, 133, 6364–6370

Journal of the American Chemical Society
ARTICLE
Figure 11. Water permeation through fullerene membranes. (a) Temperature dependence of water permeability of fullerene vesicles PhK, C4K, C8K,
C12K, C20K, and Rf8K. Error bars show the standard errors of the mean of three measurements. (b) Permeability coefficients of water through fullerene
membranes measured at 20 °C. Permeability coefficients at 20 °C: PhK: P = (5.45 ( 0.32) ꢁ 10ꢀ7 m/s, C4K: P = (3.36 ( 0.05) ꢁ 10ꢀ6 m/s, C8K: P =
(1.28 ( 0.01) ꢁ 10ꢀ6 m/s, C12K: P = (4.82 ( 0.12) ꢁ 10ꢀ7 m/s, C20K: P = (1.69 ( 0.10) ꢁ 10ꢀ7 m/s, Rf8K: P = (3.01 ( 0.09) ꢁ 10ꢀ7 m/s. (c)
Activation enthalpy (ΔH‡) and activation entropy (ꢀTΔS‡) values for water permeation at 80 °C. Data for the PhK vesicle are cited from ref 5.
decrease of the permeability in the C20K vesicle (and the Rf8K
vesicle).
vesicle surface. The solubility of the molecule appears to be
produced by the potassium cation, which is solvated by water
molecules and located in the aqueous phase. The nꢀpꢀn0
amphiphile provides a hydrophobic surface to a water-soluble
hollow object, and this can be achieved by self-assembly. This
property suggests applications of the vesicles to the targeted
delivery of organic and inorganic materials.19 In addition, we can
envisage the utility of the surface-coating property of the vesicles
in the nanolevel surface modification of a solid substrate.20 These
nꢀpꢀn0-type amphiphiles will open a new avenue in the
research on nanometer-sized compartment structures in solution
and on solid surfaces.
’ CONCLUSION
We have studied the behavior of 20 fullerene anions, PhK,
C1KꢀC20K, iC3K, tC4K, Rf4K, and Rf8K, and found that the
fullerene anions having a nonpolar/polar/nonpolar (nꢀpꢀn0)
motif spontaneously form an interdigitated bilayer vesicle in
water. All vesicles obtained in this study show essentially the
same physical properties despite the diversity of the substituents.
The SEM and AFM images indicate that the vesicles are spherical
and very robust, indicating that the vesicle is more like an eggshell
than a biological cell. This robustness is one reason for the low
water permeability.
The exposure of the hydrophobic groups to the aqueous
environment that we propose is supported by several experi-
ments. First, VT NMR experiments indicated the higher mobility
of terminal CH3 groups compared to the internal CH2 groups in
the C20K vesicle (Figure 6). Second is the observed systematic
dependency of the water contact angle on the alkyl chains
(Figure 10). The third is the fact that both PhK and C20K show
similar temperature dependence profiles of water permeation,
which is different from those reported for lipid vesicles
(Figure 11a). Finally, it is known that the fullerene/fullerene
cohesive power is extremely high (much higher than the hydro-
carbon/hydrocarbon interaction which is known to be very
weak), and it has been demonstrated in many two- and three-
dimensional solid states that fullerene molecules stick together
very tightly.18
One attractive feature of the present vesicle preparation is that
we can obtain a morphologically stable, monodispersed, nan-
ometer-size, spherical hollow object by a simple method and
without any purification. These nꢀpꢀn0-type fullerene anions
have little resemblance to lipid molecules, as both ends of the
molecule are hydrophobic and the polar anionic part is buried in
the molecular structure. The strong cohesive power of fullerene
must be instrumental in forming the vesicle assembly. The
hydrophobic chains exposed on the surface of the vesicle did
not hamper the vesicle formation in water and may have even
increased the stability of the vesicles by clustering together on the
’ ASSOCIATED CONTENT
S
Supporting Information. Procedures for the synthesis of
b
new compounds; preparation of vesicles; DLS, SLS, AFM, SEM,
and water permeability measurements of vesicles; water contact
angles of vesicle covered surface; and temperature dependence of
1H NMR of C20K vesicle. This material is available free of charge
via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
nakamura@chem.s.u-tokyo.ac.jp
’ ACKNOWLEDGMENT
We thank Mr. K. Murata (FEI Company) for SEM measure-
ments. We thank MEXT (KAKENHI for E.N., No. 22000008,
the Global COE Program for Chemistry Innovation, and JSPS
predoctoral fellowship for T.H.) for financial support. The
generous supply of [60]fullerene from Frontier Carbon Corpora-
tion is acknowledged.
’ REFERENCES
(1) (a) Vesicles; Rosoff, M., Ed.; Marcel Dekker: New York, 1996.
(b) Biomembranes; Gennis, R. B., Ed.; Springer: New York, 1989.
(2) (a) Kunitake, T.; Okahata, Y. J. Am. Chem. Soc. 1977, 99,
3860–3861. (b) Discher, B. M.; Won, Y. Y.; Ege, D. S.; Lee, J. C. M.;
6369
dx.doi.org/10.1021/ja200498g |J. Am. Chem. Soc. 2011, 133, 6364–6370

Journal of the American Chemical Society
ARTICLE
Bates, F. S.; Discher, D. E.; Hammer, D. A. Science 1999,
284, 1143–1146. (c) Fuhrhop, J. H.; Bartsch, H. Liebigs Ann. Chem.
1983, 5, 802–815. (d) Saji, T.; Hoshino, K.; Ishii, Y.; Goto, M. J. Am.
Chem. Soc. 1991, 113, 450–456. (e) Lee, H.-K.; Park, K. M.; Jeon, Y. J.;
Kim, D.; Oh, D. H.; Kim, H. S.; Park, C. K.; Kim, K. J. Am. Chem. Soc.
2005, 127, 5006–5007. (f) Hao, J.; Li, H.; Liu, W.; Hirsch, A. Chem.
Commun. 2004, 602–603. (g) Greenfield, M. A.; Palmer, L. C.; Vernizzi,
G.; de la Cruz, M. O.; Stupp, S. I. J. Am. Chem. Soc. 2009,
131, 12030–12031. (h) Babu, S. S.; M€ohwald, H.; Nakanishi, T. Chem.
Soc. Rev. 2010, 39, 4021–4035. (i) Versluis, F.; Tomatsu, I.; Kehr, S.;
Fregonese, C.; Tepper, A. W. J. W.; Stuart, M. C. A.; Ravoo, B. J.;
Koning, R. I.; Kros, A. J. Am. Chem. Soc. 2009, 131, 13186–13187.
(3) (a) Haines, T. H.; Liebovitch, L. S. Permeability and Stability of
Lipid Bilayers; CRC: Boca Raton, FL, 1995. (b) Fettiplace, R.; Haydon,
D. A. Physiol. Rev. 1980, 60, 510–550. (c) Reeves, J. P.; Dowben, R. M.
J. Membr. Biol. 1970, 3, 123–141. (d) Sharma, R. R.; Chellam, S. J. Colloid
Interface Sci. 2006, 298, 327–340.
(4) (a) Sawamura, M.; Nagahama, N.; Toganoh, M.; Hackler, U. E.;
Isobe, H.; Nakamura, E.; Zhou, S. Q.; Chu, B. Chem. Lett. 2000,
29, 1098–1099. (b) Zhou, S. Q.; Burger, C.; Chu, B.; Sawamura, M.;
Nagahama, N.; Toganoh, M.; Hackler, U. E.; Isobe, H.; Nakamura, E.
Science 2001, 291, 1944–1947.
(5) Isobe, H.; Homma, T.; Nakamura, E. Proc. Natl. Acad. Sci. U.S.A.
2007, 104, 14895–14898.
(6) Homma, T.; Harano, K.; Isobe, H.; Nakamura, E. Angew. Chem.,
Int. Ed. 2010, 49, 1665–1668.
(7) (a) Sawamura, M.; Iikura, H.; Nakamura, E. J. Am. Chem. Soc.
1996, 118, 12850–12851. (b) Matsuo, Y.; Nakamura, E. Chem. Rev.
2008, 108, 3016–3028.
(8) Matsuo, Y.; Tahara, H.; Nakamura, E. Chem. Lett. 2005,
34, 1078–1079.
(9) Burger, C.; Hao, J.; Ying, Q.; Isobe, H.; Sawamura, M.; Nakamura,
E.; Chu, B. J. Colloid Interface Sci. 2004, 275, 632–641.
(10) Zhong, Y. W.; Matsuo, Y.; Nakamura, E. J. Am. Chem. Soc. 2007,
129, 3052–3053.
(11) The bilayer scheme in Figure 1e in our original Communication
(ref 6) needs to be revised.
(12) Shibata, A. Liposomes in Life Science Experimental Manual;
Terada, H., Yoshimura, T., Eds.; Springer: Tokyo, 1992.
(13) Haran, N.; Shporer, M. Biochim. Biophys. Acta 1976, 426,
638–646.
(14) Ludwig, R.; Appelhagen, A. Angew. Chem., Int. Ed. 2005,
44, 811–815.
(15) Zwolinski, B. J.; Eyring, H.; Reese, C. E. J. Phys. Chem. 1949,
53, 1426–1453.
(16) Matsuo, Y.; Iwashita, A.; Abe, Y.; Li, C.-Z.; Matsuo, K.;
Hashiguchi, M.; Nakamura, E. J. Am. Chem. Soc. 2008, 130, 15429–
15436.
(17) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides,
G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321–335.
(18) Li, C.-Z.; Matsuo, Y.; Nakamura, E. J. Am. Chem. Soc. 2010,
132, 15514–15515.
(19) (a) Maeda, R.; Noiri, E.; Isobe, H.; Homma, T.; Tanaka, T.;
Negishi, K.; Doi, K.; Fujita, T.; Nakamura, E. Hypertens. Res. 2008,
31, 141–151. (b) Isobe, H.; Nakanishi, W.; Tomita, N.; Jinno, S.;
Okayama, H.; Nakamura, E. Chem.ꢀAsian J. 2006, 1, 167–175. (c)
Maeda-Mamiya, R.; Noiri, E.; Isobe, H.; Nakanishi, W.; Okamoto, K.;
Doi, K.; Sugaya, T.; Izumi, T.; Homma, T.; Nakamura, E. Proc. Natl.
Acad. Sci. U.S.A. 2010, 107, 5339–5344. (d) Nakamura, E; Isobe, H.
Chem. Rec. 2010, 10, 260–270.
(20) Katagiri, K.; Hamasaki, R.; Ariga, K.; Kikuchi, J. J. Am. Chem.
Soc. 2002, 124, 7892–7893.
6370
dx.doi.org/10.1021/ja200498g |J. Am. Chem. Soc. 2011, 133, 6364–6370