Novel vesicular aggregates of crown-appended cholesterol derivatives which act as gelators of organic solvents and as templates for silica transcription

Article abstract of DOI:10.1021/ja001623f

New diazacrown-appended cholesterol gelators 1 and 2 were synthesized, and their gelation ability was evaluated in organic solvents. Very surprisingly, 1 + acetic acid gel results in spherical vesicles with two distinct sizes 200 and 2500 nm in diameter.

Full text of DOI:10.1021/ja001623f

8648
J. Am. Chem. Soc. 2000, 122, 8648-8653
Novel Vesicular Aggregates of Crown-Appended Cholesterol
Derivatives Which Act as Gelators of Organic Solvents and as
Templates for Silica Transcription
Jong Hwa Jung, Yoshiyuki Ono, Kazuo Sakurai, Masahito Sano, and Seiji Shinkai*
Contribution from the Chemotransfiguration Project, Japan Science and Technology Corporation (JST),
2432 Aikawa, Kurume, Fukuoka 839-0861, Japan
ReceiVed May 11, 2000
Abstract: New diazacrown-appended cholesterol gelators 1 and 2 were synthesized, and their gelation ability
was evaluated in organic solvents. Very surprisingly, 1 + acetic acid gel results in spherical vesicles with two
distinct sizes 200 and 2500 nm in diameter. In particular, the smaller vesicles are linked linearly, which is
considered to serve as a driving-force for gelation. In contrast, 2 has a multilayered tubular structure. To
characterize their aggregation modes in the gel phase, the organogels were observed by CD spectroscopy. The
CD spectrum of 1 + acetic acid gel exhibits a negative sign for the first Cotton effect, indicating that the
dipole moments in the gelator aggregate orient into an anticlockwise direction. On the other hand, 2 exhibits
a positive sign for the first Cotton effect, indicating that they orient into a clockwise direction. The results
indicate that the aggregate of 2 is stabilized by intermolecular cholesterol-cholesterol and azobenzene-
azobenzene interactions, whereas the CD sign from the aggregate of 1 is indicative of an intramolecular
azobenzene-azobenzene interaction. The spherical vesicle structures of organogel 1 were successfully transcribed
into silica structures by the sol-gel polymerization of tetraethoxysilane (TEOS) in the gel phase. The TEM
observation established that the wall of the spherical silica obtained in the acidic conditions consists of the
multilayered vesicle structure. On the other hand, addition of Pd(NO3)2 changed the silica structure into fluffy
globules with ∼6000 nm in diameter. The EPMA observation established that Pd(II) ions are densely deposited
on the surface of this globular silica. Hence, this process is useful as a new method to create metal catalytic
sites on the silica support. These results indicate that the spherical multilayered structure of the organogel can
be precisely transcribed into the silica structure. We thus believe that the sol-gel polymerization using molecular
assembly templates strongly built in the organogel phase is a new strategy to create superstructured silica
materials.
Introduction
weight compounds formed by noncovalent interactions are
responsible for such gelation phenomena. Hence, the xerogels
can exhibit various superstructures, reflecting the monomeric
structure of each gelator. This is why the organogel studies are
considered to be a new field of supramolecular chemistry. More
recently, it was found that certain cholesterol derivatives can
gelate even tetraethoxysilane (TEOS), which results in silica
by the sol-gel polymerization.11 Very interestingly, it was
shown that the sol-gel polymerization of gelated TEOS
solutions affords novel silica with hollow lamellar,11a,b helical,11e
or linear fiber11c,d structures because the superstructures act as
a template to create an inner tube during the TEOS polymeri-
Gelation of organic solvents by low-molecular weight com-
pounds is the subject of increasing attention, not only because
of numerous applications of the organogels to industrial purposes
(e.g., in foods, deodorants, cosmetics, athletic shoes, and
chromatography)1 but in particular because of their striking
properties to create various superstructures with respect to self-
assembly phenomena.1-10 Fibrous aggregates of low-molecular
* To whom correspondence should be addressed. E-mail: seijitcm@
mbox.nc.kyushu-u.ac.jp.
(1) (a) Esch, J. V.; Schoonbeek, F.; Loos, M. d.; Kooijman, H.; Spek,
A. L.; Kellogg, R. M.; Feringa, B. L. Chem. Eur. J. 1999, 5, 937. (b) Lu,
l.; Cocker, T. M.; Bachman, R. E.; Weiss, R. G. Langmuir 2000, 16, 20.
(c) Adballah, D. J.; Weiss, R. G. Langmuir 2000, 16, 352.
(2) Yoza, K.; Amanokura, N.; Ono, Y.; Akao, T.; Shinmori, H.; Takeuchi,
M.; Shinkai, S.; Reinhoudt, D. N. Chem. Eur. J. 1999, 5, 2722.
(3) (a) Vries, E. J. de.; Kellogg, R. M. J. Chem. Soc., Chem. Commun.
1993, 238. (b) Loos, M. de; Esch, J. van; Stokroos, I.; Kellogg, R. M.;
Feringa, B. L. J. Am. Chem. Soc. 1997, 119, 12675.
(4) (a) Wang, R.; Geiger, C.; Chen, L.; Swanson, B.; Whittern, D. G. J
Am. Chem. Soc. 2000, 122, 2399. (b) Geiger, C.; Stanesch, M.; Chen, L.
H.; Whittern, D. G. Langmuir 1999, 15, 5. 2241.
(5) (a) Hanabusa, K.; Okui, K.; Karaki, K.; Koyama, T.; Shirai, H. J.
Chem. Soc. Chem. Commun. 1992, 1371. (b) Hanabusa, K.; Kawakami,
K.; Kimura, M.; Shirai, H. Chem. Lett. 1997, 191 and references therein.
(6) Sohna, J.-E. S.; Fages, F. Chem. Commun. 1997, 327.
(7) (a) Otsuni, E:; Kamaras, P.; Weiss, R. G. Angew. Chem., Int. Ed.
Engl. 1996, 35, 1324 and references therein. (b) Terech, P.; Furman, I.;
Weiss, R. G. J. Phys. Chem. 1995, 99, 9558 and references therein.
(8) (a) Murata, K.; Aoki, M.; Suzuki, T.; Harada, T.; Kawabata, H.;
Komori, T.; Ohseto, F.; Ueda, K.; Shinkai, S. J. Am. Chem. Soc. 1994,
116, 6664 and references therein. (b) James, T. D.; Murata, K.; Harada, T.;
Ueda, K.; Shinkai, S. Chem. Lett. 1994, 273. (c) Jeong, S. W.; Murata, K.;
Shinkai, S. Supramol. Sci. 1996, 3, 83.
(9) Brotin, T.; Utermo¨hlen, R.; Fages, F.; Bouas-Laurent, H.; Desvergne,
J.-P. J. Chem. Soc., Chem. Commun. 1991, 416.
(10) (a) For recent comprehensive reviews, see Terech, P.; Weiss, R.
G. Chem. ReV. 1997, 97, 3133. (b) Shinkai, S.; Murata, K. J. Mater. Chem.
(Feature Article) 1998, 8, 485.
(11) (a) Jung, J. H.; Ono, Y.; Shinkai, S. Langmuir 2000, 16, 1643. (b)
Jung, J. H.; Ono, Y.; Shinkai, S. J. Chem. Soc., Perkin Trans 2 1999, 1289.
(c) Ono, Y.; Nakashima, K.; Sano, M.; Kanekiyo, Y.; Inoue, K.; Hojo, J.;
Shinkai, S. Chem. Commun. 1998, 1477. (d) Ono, Y.; Kanekiyo, Y.; Inoue,
K.; Hojo, J.; Shinkai, S. Chem. Lett. 1999, 23. (e) Ono Y.; Nakashima, K.;
Sano, M.; Hojo, J.; Shinkai, S. Chem. Lett. 1999, 1119.
10.1021/ja001623f CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/26/2000

Vesicular Aggregates of Cholesterol DeriVatiVes
J. Am. Chem. Soc., Vol. 122, No. 36, 2000 8649
Table 1. Gelation Abilitya of 1 and 2 in Organic Solvents
zation process. One may thus consider that the unique super-
structures in the organogels can be transcribed into the various
silica structures: this means that the superstructures of organic
aggregates, temporarily formed by noncovalent interactions, can
be permanently fixed in inorganic materials.
entry
solvent
1
2
entry
solvent
1-decanol
(S)-phenylethylamine
(R)-phenylethylamine
(S)-diaminocyclohexane
(R)-diaminocyclohexane
pyridine
1
2
1
2
3
4
5
6
7
8
chloroform
benzene
acetonitrile
DMSO
1-butanol
1-hexanol
1-octanol
1-nonanol
S
S
I
I
I
G
G
G
S
S
I
G
G
G
G
G
9
G
G
G
G
G
G
PG
G
G
S
S
S
S
S
S
S
10
11
12
13
14
15
16
Gokel et al. and others found that certain azacrown-appended
cholesterol derivatives can form unique vesicular or lamellar
structures in the absence and the presence of metal salts in
aqueous solution.12 They possess a polar azacrown headgroup
and a suitable hydrophobic group to form stable aggregates,
and the aggregate morphology can be adjusted by the ring size
of the azacrown headgroup or the length of the hydrophobic
group. It thus occurred to us that these superstructures created
from the azacrown-appended cholesterol derivatives might be
useful as a template for the transcription into the silica structure.
Mesoporous inorganic materials with vesicular structures have
received special attention because of their potential applications
as sorption media, molecular sieves, and catalysts.13,14 However,
all of vesicular (multilayered) mesostructures reported to date
had shells of undesirable thickness and shape. Recently,
Pinnavaia and co-workers reported that arrangements of me-
soporous inorganic materials with vesicular structures which
showed 3-70 nm of thin shells have been accomplished by a
“gemini”-type surfactants as template in aqueous solution.14
However, these particles have shells that consist of uni- and
multi-lamellae.
acetic acid
aniline
a Gelator ) 5.0 wt %; G: stable gel formed at room temperature;
S: soluble; I: insoluble; PG: partially gelatinized.
Figure 1. CD spectra of (A) 1 + acetic acid gel, (B) 1‚Pd(NO3)2 +
aniline gel, and (C) 2 + cyclohexane gel.
16 different organic solvents containing 1 or 2. As summarized
in Table 1, 1 can gelate 11 out of 16 solvents, indicating that it
acts as a versatile gelator of organic solvents. In particular, it is
seen from Table 1 that the gelation ability for long-chain
alcohols and amines is excellent. On the other hand, 2 can gelate
only 6 out of 16 solvents. The difference suggests that the
“gemini”-type cholesterol gelator is superior to conventional
“mono”-type cholesterol gelator.
To characterize the aggregation mode in the organogel
systems, CD (circular dichroism) spectra of the organogels were
measured for 1 + acetic acid gel, 1‚Pd(NO3)2 + aniline gel,
and 2 + butanol gel (Figure 1). In their CD spectra, the λθ ) 0
values appear at 345 nm (absorption spectrum: λmax) 345 nm),
368 nm (absorption spectrum: λmax) 368 nm), and 390 nm
(absorption spectrum: λmax) 375 nm), respectively. One can
thus assign these CD spectra to the exciton-coupling bands. It
is known that azobenzene-appended cholesterol gelators with
natural (S) C-3 configuration tend to give a positive sign for
the first Cotton effect, indicating that the dipole moments of
the azobenzene chromophores tend to orient into the clockwise
direction.8a Gelator 2 bearing only one cholesterol moiety also
exhibits a positive sign for the first Cotton effect. In contrast,
the CD spectra of 1 + acetic acid gel and 1‚Pd(NO3)2 + aniline
gel exhibit a negative sign for the first Cotton effect, indicating
that the dipole moments of the azobenzene chromophores orient
into the anticlockwise direction. The CD maximum of 1 + acetic
acid gel appears at shorter wavelength than that of 1‚Pd(NO3)2
+ aniline gel, indicating that 1 in acetic acid can form the more
stable gel than that of 1 + aniline gel in the presence of Pd-
(NO3)2. This shift in CD and UV can be explained by a better
stacking of the azobenzene moieties.16c In all cases, it was
We therefore designed gelator 1 which has two cholesterol
and azobenzene moieties as the aggregate-forming sites and one
azacrown moiety as a polar headgroup. We also prepared the
gelator 2 as a reference for 1. We have found that 1 not only
gelates various organic solvents but also acts as a template in
the presence of metal salts or acidic medium for the sol-gel
polymerization of TEOS to produce novel spherical structures
of the silica. Here, we report a direct and highly efficient method
toward the creation of novel spherical mesoporous silica with
very thin shells ( ∼5-20 nm) and ∼200-2500 nm diameter
by the sol-gel polymerization in the organogel phase. To the
best of our knowledge, this is the first example not only for the
multilayered spherical structure in the organogel system but also
for the direct transcription of the molecular assemblies into the
sphere of the silica in the organogel phase.
Results and Discussion
Characterization of Organogel Superstructures by Cir-
cular Dichroism (CD). The gelation ability was estimated for
(12) (a) Echegoyen, L. E.; Hernandez, J.; Kaifer, A. E.; Gokel, G. W.;
Echegoyen, L. J. Chem. Soc. Chem. Commun. 1988, 836. (b) Nakano, A.;
Hernadez, J. C.; Dewall, S. L.; Wang, K.; Berger. D. R.; Gokel, G. W.
Supramol. Sci. 1997, 8, 21. (c) Echegoyen. L. E.; Portugal, L.; Miller, S.
R.; Hernandez, J. C.; Echegoyen, L.; Gokel, G. W. Tetrahedron Lett. 1988,
29, 4065. (d) Okahara, M.; Kuo, P. L.; Yamamura, S.; Ikeda, I. J. Chem.
Soc. Chem. Commun. 1980, 586.
(13) (a) Schacht, S.; Huo, Q.; Voigt-Martin, I. G.; Stucky, G. D.; Schu¨th,
F. Science 1996, 273, 768. (b) Huo, Q.; Feng, J.; Schu¨th, F.; Stucky, G. D.
Chem. Mater. 1997, 14, 119. (c) Caruso, F. Chem. Eur. J. 2000, 6. 413. (d)
Caruso, F.; Caruso, R. A.; Mo¨hwald, H. Science 1998, 282, 1111.
(14) (a) Tanev, P. T.; Ling, Y.; Pinnavaia, J. Am. Chem. Soc. 1997, 119,
8616. (b) Kim, S. S.; Zhang, W.; Pinnavaia, T. J. Science 1998, 282, 1302.
(15) The xerogel more suitable to the SEM observation was obtained
from cyclohexane solution (mp: 6.47 °C) than from a 1-butanol solution
(mp: -90 °C), because this was prepared by a freezing-and-pumping
method. The structure of xerogel does not depended on organic solvent
according to ref 11.
(16) (a) Kunitake, T. Angew. Chem., Int. Ed. Engl. 1992, 21, 709. (b)
Kunitake, T.; Ishikawa, Y.; Shimoumura, M. J. Am. Chem. Soc.1986, 108,
327. (c) Fuhrhop, J.-H.; Ko¨ning, J. Membranes and Molecular As-
semblies: The Synkinetic Approach; The Royal Society of Chemistry: 1994.

8650 J. Am. Chem. Soc., Vol. 122, No. 36, 2000
Jung et al.
Figure 3. Schematic representation of sol-gel transcription of the
multilayered spherical structure of the organolgel 1.
2 + cyclohexane gel shows the rolled film-like structure (Figure
2E), which is occasionally observed for less stable, inferior
organogels.11,15 Very surprisingly, 1 + acetic acid gel results
in two different spherical structures, one with ∼200 nm diameter
and the other with ∼2500 nm diameter (Figure 2A and B: the
overview picture supporting this explanation is shown in
Supporting Information). The spherical morphology makes a
sharp contrast from more common forms of organogels, fibrous
or platelike. Further interesting is the finding that the smaller
vesicles are connected to one another like a pearl necklace. Thus
far, it has been believed that the fibrous aggregates are
indispensable to gelation. Presumably, this new class of cross-
link in the present system can be also the origin of the gelation
phenomena. When 1 + acetic acid gel was dissolved by heating
and sonicated by a 35 W tip sonicator for 15 min, the resultant
SEM picture shows the increase in the concentration of the larger
vesicle. One particle was cut by ultramicrotome (Figure 2B). It
is seen from Figure 2B that the particle has the shell wall of
∼2000 nm thickness and inner sphere of ∼1000 nm. The clearer
image of the larger vesicle was obtained from the TEM pictures.
The 1 + acetic acid gel was translucent even in the presence of
uranyl acetate, which was subjected to the TEM observation.
The TEM picture (Figure 2C) clearly established that the shell
wall consists of the multilayered structure with ∼5 nm layer
thickness and 200∼350 nm diameter. In 1, the solvophilic
azacrown moiety and the solvophobic cholesterol moieties
should occupy the surface and the inner part of the layer,
respectively. Hence, the layer thickness of ∼5 nm is com-
mensurate with the vesicle wall structure as illustrated in Figure
3: that is, 1 adopts a folded conformation in which the
cholestrol-cholesterol interaction occurs in the inner part in an
interdigitated manner and the azobenzene-azobenzene interac-
tion occurs primarily in an intramolecular manner and second
in an intermolecular manner toward the lateral direction. This
aggregation mode is compatible with the ∼5 nm layer thickness
and well explains the reason the CD spectra are different from
those of 2 which features the one-dimensional molecular
stacking8a (Figure 1). To the best of our knowledge, this is the
first example for the creation of multilayered spherical vesicles
in organic media (acetic acid for 1). This system is quite
Figure 2. (A and B) SEM micrographs of the xerogel 1 prepared from
acetic acid; (C) TEM micrograph of the xerogel 1 obtained from acetic
acid (negatively stained by UO22+), (D) SEM micrograph of the xerogel
1‚Pd(NO3)2 prepared from the aniline gel, and (E) SEM micrograph of
the xerogel 2 prepared from the cyclohexane gel.
confirmed that the contribution of linear dichroism (LD) to the
true CD spectra is negligible. These CD data support the view
that the aggregation mode of 1 is somewhat different from that
of 2. It is undoubted that the organogels of 2 are formed by
both intermolecular cholesterol-cholesterol and azobenzene-
azobenzene interactions using the former interaction in the
central columnar aggregate and the latter interaction in the side
chain aggregate around the central column.8a On the other hand,
1 in the organogels may adopt a folded conformation to enjoy
efficient intramolecular cholesterol-cholesterol and azoben-
zene-azobenzene interaction, which may be the origin of the
negative sign for the first Cotton effect. This problem will be
discussed later in more detail.
Vesicular Structure of Organogels as Observed by SEM
and TEM. To visually observe the aggregation modes in the
organogel systems, the xerogels of 1 + acetic acid gel, 1‚Pd-
(NO3)2 + aniline gel, and 2 + cyclohexane gel were prepared
by a freezing-and-pumping method from their gel phases below
the sol-gel phase transition temperature and their SEM and
TEM pictures were taken (Figure 2). The xerogel obtained from

Vesicular Aggregates of Cholesterol DeriVatiVes
J. Am. Chem. Soc., Vol. 122, No. 36, 2000 8651
Figure 4. SEM micrographs of the silica obtained from (A) 1 + acetic acid gel, (B) 1‚Pd(NO3)2 + aniline gel, and (C) 2 + acetic acid gel.
Figure 5. TEM micrographs of the silica obtained from 1 + acetic acid gel (A) before and (B) after sectioning.
different from azacrown-appended cholesterol surfactants previ-
ously reported by Gokel et al. which have long methylene groups
between the azacrown headgroup and the cholesterol group.12a-c
This surfactant can only form vesicular and lamellar structures
in aqueous solution, whereas 1 can form the vesicular structure
even in organic solvents. In addition, Kunitake and co-workers
have reported on a bilayer assembly of azobenzene amphiphiles
containing a cyclic tetramine group as a hydrophilic head. Since
they only form rodlike aggregates with 3-4 nm diameters and
several hundred to 200 nm lengths in aqueous solution,16 the
present system is also different from Kunitake,s system.
Sol-Gel Polymerization of TEOS and SEM/TEM Obser-
vation of Silica Structures. To transcribe these novel super-
structures constructed in the organogels to silica, sol-gel
polymerization of TEOS was carried out in the gel phase of 1
and 2 at room temperature for 7 days.11 One may expect that
oligomeric silica species with the anionic charge interact with
protonated azacrown moieties with the cationic charge and the
multilayered vesicular structure is successfully transcribed into
the silica structure (as illustrated in Figure 3). Figure 4 shows
the SEM micrographs of the silica taken after calcination. From
1 + acetic acid gel, the spherical silica structures are yielded
(Figure 4A). As expected, one can observe both the cross-linked
smaller particles with ∼200 nm diameter and the isolated larger
particles with ∼2500 nm in diameter. Figure 4A accidentally
catches one broken particle (the enlarged particle is shown in
the right). It is seen from these images that the inside of these
particles have cavity inside. Presumably, this breakage is induced
by vacuum necessary to the SEM sample preparation.
To further corroborate that the spherical structure of the
organogel really acted as a template for the creation of the
spherical silica, we took TEM pictures after removal of 1 by
calcination. Figure 5A reveals that both the smaller and the
larger spherical particles have a hollow structure. After sectioned
with a ultramicrotome, the edge of the small particle was
observed with TEM (Figure 5B). It is seen from this TEM image
that the shell wall consists of the multilayered lamella having
5 nm spacing.17 These results clearly support the view that the
multilayered structure of the organogels is precisely transcribed
into the silica structure. We believe that such a precise
transcription becomes possible for the first time by using the

8652 J. Am. Chem. Soc., Vol. 122, No. 36, 2000
Jung et al.
structure in acetic acid, whereas 1 results in the fluffy globular
aggregates in the presence of Pd(NO3)2. In contrast, 2 shows
the tubular lamellar structure. Sol-gel polymerization of TEOS
in the presence of the multilayered spherical structure of 1 and
the lamellar aggregates of 2 is useful to create the multilayered
spherical and the tubular structure of the silica, respectively.
The Pd particles are deposited on the surface of the silica by
host-guest interaction.
In general, organic materials are capable of construction of a
variety of supramolecular structures reflecting their own mo-
lecular shape, whereas such “shape design” is very difficult from
inorganic materials. The present findings suggest, as also
suggested by a few other research groups, that various novel
assembly structures created by weak intermolecular forces can
be imprinted as permanent structures in inorganic materials. The
present study clearly demonstrates that the organogel system is
one of the most suitable molecular assemblies for this transcrip-
tion. We believe, however, that the present silica with the unique
higher-order morphology is still very useful for drug delivery,
nanosized microcapsulation, compartmentalization, etc.
Figure 6. EPMA pictures of the silica obtained from 1‚Pd(NO3)2 +
aniline gel: (A) silicon map, (B) palladium map, and (C) oxygen map.
organogel superstructures as a template which have the crystal-
like characters.8,10,11
In contrast, the silica obtained from the acetic acid gel of 2
features the roll-paperlike tubular structure with 450-500 nm
outer diameter and 150-170 nm inner diameter (Figure 4C). It
is known that this structure results from the sol-gel polymer-
ization of TEOS along the curved films of 2 (Figure 2E).11a
Deposition of Palladium Particle on Silica Surface by
Host-Guest Interaction. It is known that azacrown ethers have
a high affinity with metal cations. When sol-gel polymerization
of TEOS is carried out around the metal-binding organogel
superstructures, the metals would be deposited on the surface
of the silica after calcination. With this intriguing idea in mind,
we added Pd(NO3)2 to the organogel system. Although many
organic solutions containing 1 and Pd(NO3)2 could not be
gelated, we eventually found that the aniline solution is
sufficiently gelated. A SEM picture of the xerogel obtained from
the aniline solution is shown in Figure 2D. The vesicular
structure is no longer observable: instead, the fluffy globular
aggregates with ∼6000 nm diameter appear. Careful examina-
tion of the aggregates reveals that the surface consists of the
canna-flower-like coating with ∼50 nm thickness. The results
indicate that Pd(II) ions bound to the surface azacrown drasti-
cally change the morphology.
As the preliminary step toward the design of the metal-
deposited silica for catalytic application,18 we carried out the
sol-gel polymerization in 1‚Pd(NO3)2 + aniline gel. It is seen
from Figure 4B that the silica structure which is very similar to
that of the xerogel (Figure 2D) results. The metal deposition
on the silica is usually detectable by the TEM observation,11b
but in the present sample the silica particles were too thick to
obtain the clear TEM picture. We thus estimated the Pd content
by X-ray electron probe microanalyzer (EPMA). Figure 6 clearly
shows that Pd(0) and Pd(II) are deposited on the silica surface,
the content being 1.8 wt %.19 The results indicate that the
azacrown-containing gelator is useful as a new metal-deposition
method on the silica matrix.
Experimental Section
Apparatus for Spectroscopy Measurement. 1H and 13C NMR
spectra were measured on a Bruker ARX 300 apparatus. IR spectra
were obtained in KBr pellets using a Shimadzu FT-IR 8100 spectrom-
eter, and MS spectra were obtained by a Hitachi M-250 mass
spectrometer.
TEM Observation. Samples for transmission electron microscope
(TEM) studies were embedded in White acrylic resin (hard) and
sectioned on an ultramicrotome. The thin sections ( ∼20 nm) were
supported on the 400 mesh copper grids. Transmission electron
microscopy was done with a Hitachi H-7100, using accelerating voltage
of 100 kV and a 16 mm working distance. Multilayered structure of
the organogel was attached to Butvar/carbon-coated, 400 mesh copper
grids by applying 10-15 µL drops of the organogel solution containing
to 15 µL of uranyl acetate onto the grids and allowing them to remain
for 1-5 min. Excess fluid was wiped off the rids by touching their
edges by filter paper.
SEM Observation. Scanning electron microscopy (SEM) was done
with a Hitachi S-4500. The thin gel was prepared in a 1-2 mL bottle
and frozen in liquid nitrogen or dry ice-acetone. The frozen specimen
was evaporated by a vacuum pump for 24 h. The dry sample was coated
by palladium-platinum. The accelerating voltage of SEM was 5-15
kV, and the emission current was 10 µA.
Gelation Test of Organic Fluids. The gelator and the solvent were
put in a septum-capped test tube and heated in an oil bath until the
solid was dissolved. The solution was cooled at room temperature. If
the stable gel was observed at this stage, it was classified as G in Table
1.
Sol-Gel Polymerization of TEOS. In a typical preparation a 5.8
× 10-6 M quantity of gelator in the absence and the presence of metal
salt (5.8 × 10-6 M) were dissolved in 1.0 g of tetrahydrofurane. The
solution was evaporated to dryness. The residual solid was added to
acetic acid or aniline (95 mg)/TEOS (15.0 mg)/water (5.7 mg)/
benzylamine (5.6 mg) and warmed until a transparent solution was
obtained. The reaction mixture was placed at room temperature under
the static conditions for 7 days. The product was dried by a vacuum
pump at room temperature. Finally, the gelator was removed by
calcination at 200 °C for 2 h and 500 °C for 2 h under a nitrogen
atmosphere and 500 °C for 4 h under aerobic conditions.
4-N-Monobromobutoxyl-4′-((cholesteryloxy)carbonyl)azoben-
zene (4). 4-((Bromo- butyloxypheyl)azo)benzoic acid 3 (0.7 g, 1.86
mmol) and cholesterol (0.718 g, 2.23 mmol) were dissolved in 20 mL
of dichloromethane under a nitrogen atmosphere. The solution was
maintained at 0 °C with an ice bath. The dicyclohexylcarbodiimide
(DCC) (0.383 g, 1.86 mmol) and (dimethylamino)pyridine (DMAP)
(0.022 g, 0.186 mmol) were then added, the reaction mixture being
stirred for 4 h at 0 °C. The reaction mixture was filtered, and the filtrate
Conclusions
The present study has demonstrated that azacrown-appended
cholesterol gelator 1 creates the novel multilayered spherical
(17) In fact, we have already measured XRD. The peak which deserves
publication was not observed, however. Judging from the TEM pictures,
the space between silica layers is about 5 nm. One should note that this is
the “averaged” value and there is a significant distribution in the spacing.
We thus consider that this distribution is the reason we could not observed
the peak in XRD. The many reproducible TEM pictures support, however,
that there exist a clear space between the silica.
(18) (a) Mehnert, C. P.; Weaver, D. W.; Ying, J. Y. J. Am. Chem. Soc.
1998, 120, 12289. (b) Mehnert, C. P.; Ying, J. Y. Chem. Comm. 1997,
2215.
(19) Chemical component of Pd was confirmed by Electron Spectroscopy
for Chemical Analysis (ESCA): it consists of Pd(0) and Pd(II) derivative.

Vesicular Aggregates of Cholesterol DeriVatiVes
J. Am. Chem. Soc., Vol. 122, No. 36, 2000 8653
was washed with acidic and basic aqueous solutions (50 mL each).
The organic layer was evaporated to dryness. The residue was purified
by a silica gel column eluting with THF/n-hexane (1:6 v/v) to give
compound 4 in yield 26% as yellow solid (mp ) 141.5 °C). 1H NMR
(300 MHz, CDCl3) δ (ppm) ) 8.17 (2H, d, J ) 9.0 Hz), 7.72 (2H, d,
J ) 9.0 Hz), 7.90 (2H, d, J ) 9.0 Hz), 7.10 (2H, d, J ) 9.0 Hz), 5.45
(1H, d, J ) 6.3 Hz), 5.02-4.88 (1H, m), 41 (2H, t, J ) 6.3 Hz), 3.52
(2H, t, J ) 6.2 Hz), 2.49 (2H, d, J ) 6.2 Hz), 2.28-0.94 (35H, m),
0.88 (3H, s). 13C NMR (75 MHz, CDCl3) δ (ppm) ) 165.1, 161.88,
155.20, 146.98, 139.9, 130.2, 125.18, 122.84, 122.28, 114.72, 67.22,
66.67, 56.67, 56.11, 50.01, 42.30, 39.71, 39.50, 38.20, 37.01, 36.64,
36.17, 35.79, 33.32, 31.92, 31.86, 29.3, 28.32, 28.01, 27.88, 27.78,
24.28, 23.82, 22.83, 22.56, 21.04, 19.38, 19.38, 18.71, 11.86. MS(SIMS)
) 745 [M + H]+. IR (KBr) 3005, 1722, 1603, 1579, 1500, 1468, 1284,
27.96, 27.84, 26.96, 24.24, 23.78, 22.79, 22.53, 21.00, 19.34, 18.67,
11.81. MS(SIMS) ) 1019 [M + 2H]+. IR (KBr) 3010, 2943, 2868,
1711, 1595, 1585, 1500, 1463, 1417, 1404, 1275, 1140, 1109 cm-1
.
Anal. Calcd for C63H92N4O7: C, 74.41; H, 9.12; N, 5.91. Found: C,
73.70; H, 9.121; N, 5.51.
4-(N-1,10-Diaza-18-crown-6-butoxy)-4′-(bis(cholesteyloxy)carbo-
nyl)azobenzene (1). A mixture of compound 4 (0.25 g, 0.335 mmol),
diaza-18-crown-6 (0.17 g, 0.67 mmol), and sodium carbonate (0.70 g,
6.7 mmol) in dry butyronitrile (35 mL) was refluxed for 48 h. The
solution was filtered after cooling, the filtrate being concentrated to
dryness by a vacuum evaporator. The residue was purified by an
aluminum oxide column with ethanol/dichloromethane (1:50 v/v) to
give the desired product in yield 50.5% as yellow solid (mp ) 203.5-
1
205.5 °C). H NMR (300 MHz, CDCl3) δ (ppm) ) 8.17 (4H, d, J )
1116, 1047 cm-1
.
9.0 Hz), 7.93 (4H, d, J ) 9.0 Hz), 7.88 (4H, d, J ) 9.0 Hz), 6.97 (4H,
d, J ) 9.0 Hz), 5.45 (2H, d, J ) 6.3 Hz), 4.87 (2H, m), 3.98 (4H, t, J
) 6.3 Hz), 3.73-3.61 (26H, m), 2.83-2.80 (10H, m), 2.60 (4H, t, J )
6.2 Hz) 2.01-0.69 (82H, m). 13C NMR (75 MHz, CDCl3) δ (ppm) )
165.67, 161.52, 155.35, 145.02, 137.54, 130.48, 123.13, 122.24, 114.72,
86.94, 70.84, 70.73, 68.53, 56.62, 55.35, 53.94, 49.22, 39.23, 38.16,
37.01, 36.97, 36.35, 36.25, 35.12, 31.35, 27.96, 27.35, 24.24, 23.78,
22.79, 22.53, 21.00, 19.34, 18.67, 11.86. MS (SIMS) ) 1592 [M +
H]+. IR (KBr) 3005, 2940, 1722, 1603, 1579, 1500, 1468, 1284, 1116,
1047 cm-1. Anal. Calcd for C100H146N6O10: C, 75.43; H, 9.24; N, 5.28.
Found: C, 75.26; H, 9.23; N, 5.17.
4-(N-Monobenzylaza-18-crown-6-butoxy)-4′-((cholesteyloxy)car-
bonyl)azo- benzene (2). A mixture of compound 4 (0.13 g, 0.174
mmol), monobenzyldiaza-18-crown-6 (0.061 g, 0.1740 mmol), and
sodium carbonate (0.784 g, 1.74 mmol) in dry butyronitrile (15 mL)
was refluxed for 24 h. The solution was filtered after cooling, the filtrate
being concentrated to dryness by a vacuum evaporator. The residue
was purified by an aluminum oxide column with ethanol/dichlo-
romethane (1:30 v/v) to give the desired product in yield 75.8% as a
yellow solid (mp ) 113.3-115.2 °C). 1H NMR (300 MHz, CDCl3) δ
(ppm) ) 8.17 (2H, d, J ) 9.1 Hz), 7.93 (2H, d, J ) 9.1 Hz), 7.31-
7.26 (7H, m), 6.95 (2H, d, J ) 9.0 Hz), 5.45 (1H, d, J ) 6.5 Hz), 4.90
(1H, m), 4.08 (2H, t, J ) 6.5 Hz), 3.73-3.61 (32H, m), 2.78 (4H, t, J
) 12.1 Hz), 2.75 (2H, t, J ) 6.5 Hz), 2.49 (2H, d, J ) 6.3 Hz), 2.01-
0.69 (35H, m). 13C NMR (75 MHz, CDCl3) δ (ppm) ) 165.47, 162.13,
155.19, 146,80, 139.54, 134.77, 130.48, 125.13, 122.79, 122.22, 114.72,
74.79, 70.84, 70.73, 70.70, 70.35, 68.14, 56.62, 56.06, 53.94, 49.87,
42.25, 39.66. 39.45, 38.16, 36.97, 36.60, 36.12, 35.75, 31.88, 28.19,
Supporting Information Available: An overview pictures
of SEM for the xerogel 1 prepared from acetic acid (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
JA001623F