Synthesis and properties of new, spatially relaxed dendrons containing internal carboxyl groups

Article abstract of DOI:10.1246/bcsj.77.993

We synthesized a series of new dendrons with up to fourteen internal carboxyl groups. These dendrons are made from a branching unit and a spacer unit with a carboxyl group. The growth reactions (formation of the benzylic ether bonds) are completed within

Full text of DOI:10.1246/bcsj.77.993

ꢀ 2004 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 77, 993–1000 (2004)
993
Synthesis and Properties of New, Spatially Relaxed Dendrons
Containing Internal Carboxyl Groups
ꢀ
Yoshihiro Kikuzawa and Toshi Nagata
Department of Structural Molecular Science, The Graduate University for Advanced Studies,
Institute for Molecular Science, Myodaiji, Okazaki 444-8787
Received July 28, 2003; E-mail: toshi-n@ims.ac.jp
We synthesized a series of new dendrons with up to fourteen internal carboxyl groups. These dendrons are made
from a branching unit and a spacer unit with a carboxyl group. The growth reactions (formation of the benzylic ether
bonds) are completed within a few hours, which suggests high reactivity at the ‘‘focal’’ point of the dendritic framework.
The final deprotection of the internal ester groups also proceeded smoothly. These high reactivities were attributed to the
presence of spacer units, which caused spatially relaxed conformations of these molecules. The carboxylate salt form of
the dendron produced both normal and reverse micelles in a THF/water mixed solvent according to the fraction.
Dendrimers containing carboxyl functional groups are well-
studied as pH-controllable unimolecular micelles,1 shape-per-
sistent amphiphiles,2 polyionic materials,3 potential metal-
complexing agents,4 and multiply modifiable materials.5 They
are also utilized for solubilizing highly hydrophobic materials,
such as fullerenes6 and porphyrins.1e–g Most of these dendri-
mers have carboxyl groups in the periphery with internal frame-
works of polyimines,1c,4a polyamides,1a,b,d,3b,5b,6,7 benzyl
ethers,1e–g,3a,c,d,5a and phenylacetylenes.2 Some dendrimers
have carboxyl groups at the core that participate in self-assem-
bly through hydrogen bonds8 and metal-ligating interactions.4b
9
´
Scheme 1. Synthesis of the spacer unit 3.
On the other hand, Frechet reported on dendrimers with ‘‘inter-
nal’’ functional groups, and examined their catalytic activity as
unimolecular reverse micelles. Such dendrimers with internal
functional groups will also be useful for multiple functionaliza-
Compound 3 is crucial in the synthesis of this series of den-
drimers. Therefore, for these dendrimers to be useful, com-
pound 3 needs to be prepared in large quantities. The t-butyldi-
methylsilyl-protected compound 2 is useful because it is easily
handled in column chromatography. However, the usual depro-
tection with Bu4NF led to a mixture of 3 with tetrabutylammo-
nium salts, which could not be easily separated out. We used
KF instead, so that the by-products could be separated by utiliz-
ing differences in the solubility. A similar compound (an ethyl
ester) was recently prepared via a direct partial reduction of di-
ethyl 5-hydroxyisophthalate,12 but our three-steps procedure
gave higher overall yield (53% vs 30%).
The general synthetic procedures for the dendrons are shown
in Scheme 2. The terminal unit, G0, was synthesized from 4 and
1-bromooctane.11 The dendrons G1–3 were prepared by the
convergent method,13 by an alternating treatment with the
spacer unit 3 and the branching unit 4. Specifically, compound
Gn was converted to the bromide by CBr4/PPh3, and the bro-
mide was allowed to react with 3 in the presence of K2CO3
´
tion and as metal-complexing agents. However, Frechet’s den-
drimers are based on 3,5-dialkoxybenzyl ether frameworks that
lead to densely packed materials, which will make further deri-
vatization of the carboxyl groups rather difficult.
Herein we report on a new series of dendrons with internal
carboxyl groups. These dendrons are built from a spacer unit
containing a carboxyl group (methyl 3-hydroxy-5-(hydroxy-
methyl)benzoate, 3) and a branching unit (3,5-dihydroxybenzyl
alcohol, 4). The introduction of spacer units makes room
around the carboxyl groups, so that these groups maintain reac-
tivity even after the construction of a high generation of den-
drons. The growth reactions are also easier because of the high
reactivity at the focal point. Molecular dynamics (MD) simula-
tions have shown that these dendrons have void space inside,
where the carboxylate groups are accessible by the solvent
molecules. An interesting micelle-forming behavior of the car-
boxylate salt form of the dendron is also reported.
and 18-crown-6 to give the spacer-connected compound Gn.5
.
Results and Discussion
This compound Gn.5 was similarly converted to Gn+1 by a treat-
ment with CBr4/PPh3, followed by a reaction with 4. Finally,
the ester groups in the dendron G3 were hydrolyzed by excess
KOH in THF/MeOH, and G3(CO2H) was obtained.
Synthesis. The spacer unit 3 was synthesized according to
Scheme 1. The silyl-protected diester 110 was partially reduced
by 0.5 molar equivalents of LiAlH4, and the monoalcohol 2 was
isolated by column chromatography.
The formation of the ether linkage and hydrolysis of the ester
Published on the web May 6, 2004; DOI 10.1246/bcsj.77.993

994 Bull. Chem. Soc. Jpn., 77, No. 5 (2004)
Dendrons with Internal Carboxyl Groups
Fig. 1. 1H NMR spectrum of the dendron G3 in CDCl3.
Scheme 2. Synthesis of the dendrons G1–3
.
groups proceeded smoothly in short periods (2 to 3 h). This is in
contrast with other benzyl-ether-based dendrons that often re-
quire several days for the reaction to complete,3a,3d,5a and is a
useful feature for synthesizing higher generations of this series
of dendrons. The high reactivity of the ester groups will also be
useful for preparing functionalized derivatives via ester/amide
linkages.
Characterization. The 1H NMR spectrum of G3 is shown
in Fig. 1. All signals were assigned to the corresponding kinds
of protons with the expected integral values: aromatic (with and
without ester groups), benzylic, methyl ester, and octyloxy
groups. Each synthetic step was monitored by the 1H NMR sig-
nals of the focal benzylic protons (i.e., the ones at the core ben-
zene ring). The typical spectral change is shown in Fig. 2. The
CH2OH signal at 4.64 ppm (a) moved to 4.40 ppm after conver-
Fig. 2. 1H NMR spectra of the focal benzylic protons in
CDCl3. (a) G2.5, (b) the bromide, and (c) G3.
sion to CH2Br (b); after ether formation, a new CH2OH signal
appears at 4.54 ppm (c). The hydrolysis of the methyl ester was
1
also monitored by H NMR (Fig. 3).
The MALDI-TOF spectra of the dendrons are shown in

Y. Kikuzawa et al.
Bull. Chem. Soc. Jpn., 77, No. 5 (2004)
995
Fig. 3. 1H NMR spectra of methyl ester groups of G3 in
CDCl3 (a), andG3(CO2H)inDMSO-d6 afterhydrolysis(b).
Fig. 4. In all of the spectra, the peaks of the highest intensities
were assigned as M + Naþ. In addition, the peaks assigned as
M + Kþ were observed for G2 and G3. In all cases, no peaks
with a lower m/z than these peaks were observed, except for
those from the matrix, which indicate the purity of these com-
pounds.
The analytical gel-permeation chromatograms of these com-
pounds (Fig. 5) revealed a change of the retention times corre-
sponding to the molecular sizes. The symmetric peaks also in-
dicate the purity of these compounds.
Molecular Dynamics (MD) Simulations. In order to vis-
ualize the molecular shape of these dendrons, we performed
MD simulations on the G3 dendron using the AMBER7 pro-
gram package.14 Such simulations will also aid us to design
functional molecules based on these dendrons. Two types of
simulations were examined, one without a solvent (i.e., in va-
cuo simulation) and the other in an explicit solvent box of
chloroform.
Fig. 4. MALDI-TOF mass spectra of the dendrons G1–3
.
100-ps run of molecular dynamics at 300 K without solvent,
and (c) after a 100-ps run of molecular dynamics at 300 K, 1
atm with an explicit solvent. In the absence of a solvent the den-
dron collapsed completely within 100 ps, whereas in the pres-
Figure 6 shows the resultant three conformers of the G3 den-
dron: (a) the initial, energy-minimized conformation, (b) after a

996 Bull. Chem. Soc. Jpn., 77, No. 5 (2004)
Dendrons with Internal Carboxyl Groups
Fig. 5. Analytical gel-permeation chromatograms of the
dendrons of G0.5–3
.
ence of a solvent the dendron retained its extended conforma-
tion during the same time period. These results reveal the pres-
ence of spatial ‘‘voids’’ in this molecule, which can only be
maintained by solvation. They also emphasize the importance
of including explicit solvent molecules in a simulation of this
type of dendrons.
The microscopic structure around the innermost (i.e., closest
to the core) carboxylate functional groups are further picturized
by calculating the radial distribution functions of both the sol-
vent and dendron atoms (Fig. 7). With explicit solvents (part
(a)), no dendron atoms, but the solvent atoms, are found at dis-
ꢀ
tances less than 5 A. On the other hand, the results without ex-
plicit solvents (part (b)) show the presence of dendron atoms at
ꢀ
distances as short as 3 A. These results suggest high accessibil-
ity of the ester functional groups from the solution phase.
Formation of Micelles by the Carboxylate Salt Form of
the G3 Dendron. In order to examine the micelle-forming
properties of these dendrons, the carboxylate salt form of the
G3 dendron, G3(CO2Na), was prepared by neutralization of
G3(CO2H) with an equivalent amount of aqueous sodium hy-
droxide. After 5 min of sonication, the following solvents gave
uniform emulsions: THF, CH2Cl2, CHCl3, and CH3CN. On the
other hand, MeOH, EtOH, acetone, hexane, EtOAc, and water
did not give emulsions, but suspensions of amorphous solid ma-
terials. Among the examined solvents, only a THF/water mix-
ture gave a colorless transparent solution with the G3(CO2Na)
when the water fraction was 2–45% by volume. When the water
fraction was less than 2%, part of G3(CO2Na) precipitated. As
the water fraction increased from 45%, the solution turned
white and turbid, and became opaque when the water fraction
was more than 60%, suggesting the formation of micelles.
The average diameters of micellar particles were estimated
by dynamic light scattering (DLS) measurements. Figure 8
shows the results, together with the intensities of the scattered
light. From 90% to 60% of water fraction, the intensities were
slightly diminished, and the estimated diameters increased
from 100 to 400 nm. From 60% to 45%, the diameters changed
irregularly between 200–700 nm, and intensities were drastical-
ly diminished. From 45% to 0%, the intensities were constantly
low, suggesting that very small amount of large micellar parti-
cles were present; the estimated diameters were almost constant
at around 400 nm.
Fig. 6. Ball-and-stick representations of the G3 dendron, (a)
before the MD run, (b) after 100-ps MD without solvent,
and (c) after 100-ps MD with explicit solvent.
Figure 9 shows a model to account for these results. From
90–60% of water, the dendron molecules form normal micelles

Y. Kikuzawa et al.
Bull. Chem. Soc. Jpn., 77, No. 5 (2004)
997
Fig. 7. The radial distribution functions from the innermost
ester carbonyl atoms. (a) Calculated from the MD results
with explicit solvents; the dendron atoms (solid line) and
the solvent atoms (dotted line). (b) Calculated for the den-
dron atoms from the MD results without solvent. For the
dendron atoms, those atoms are omitted that are included
in or directly connected to the aromatic ring to which the
ester carbonyl atom in question is attached.
Fig. 9. A proposed model of switching between normal and
reverse micelles in a mixed solvent of THF/water.
with hydrophobic octyloxy groups and THF inside (phase A).
As the fraction of THF in solution increases, the micellar par-
ticles continue to swell by uptaking more THF molecules
(phase B). However, between 60–45% of water, these micellar
particles become unstable, and break down to form smaller par-
ticles of reverse micelle (phase C). Note that these dendrons,
which have hydrophilic groups inside and hydrophobic groups
outside, can form reverse-micelles structurally more easily than
normal micelles. Consequently, the reverse-micelle particles in
this phase may consist of a smaller number of dendrons, or may
even be unimolecular. From 45% to 2% of water, these small
reverse-micelle particles are mainly present together with a
small number of large (ca. 400 nm) particles (phase D). When
the water fraction is less than 2%, the dendrons are no longer
capable to form reverse micelles and begin to precipitate as
solid (phase E).
In summary, we developed new dendrons with internal car-
boxyl groups. The introduction of a spacer group made these
dendrons spatially relaxed, which caused high reactivity both
in the growth process of the dendritic framework and in the fi-
nal manipulation of the internal carboxyl groups. The structural
feature led to interesting reverse/normal switching of micelle
Fig. 8. Estimated average diameters of micelles (solid line,
circle) and intensities of the scattered light (broken line,
square) at various volume fractions of water.

998 Bull. Chem. Soc. Jpn., 77, No. 5 (2004)
Dendrons with Internal Carboxyl Groups
by column chromatography (silica gel hexane/EtOAc, and/or
preparative GPC) gave the product.
formation. The utilization of these dendrons for the develop-
ment of new materials is currently underway.
Compound G1. 3,5-Bis(octyloxy)benzyl bromide11 (15.1 g,
35.4 mmol), 3 (10.7 g, 58.9 mmol), K2CO3 (6.1 g, 44 mmol),
and 18-crown-6 ether (0.703 g, 2.66 mmol) in acetone 350 mL
were allowed to react according to Procedure B to give G0.5
(12.8 g, 69%). G0.5 (5.29 g, 10.0 mmol) was converted to the bro-
mide (5.70 g, 9.60 mmol, 96%) according to Procedure A (CBr4,
4.97 g, 15.0 mmol; PPh3, 3.94 g, 15.0 mmol; THF, 40 mL). This
bromide (4.97 g, 8.40 mmol) was allowed to react with 4 (0.561
g, 4.00 mmol) in the presence of K2CO3 (2.76 g, 20.0 mmol)
and 18-crown-6 (0.317 g, 1.20 mmol) in acetone (50 mL) for 30
min according to Procedure B, to give G1 (4.34 g, 93%). 1H NMR
(CDCl3) ꢀ 7.68 (s, 2H, Ar), 7.57 (s, 2H, Ar), 7.23 (s, 2H, Ar), 6.59
(d, 2H, J ¼ 2 Hz, Ar), 6.54 (d, 4H, J ¼ 2 Hz, Ar), 6.50 (t, 1H, J ¼
2 Hz, Ar), 6.39 (t, 2H, J ¼ 2 Hz, Ar), 5.02, 5.01 (2s, 4H + 4H,
ArOCH2Ar0), 4.62 (s, 2H, CH2OH), 3.92 (t, 8H, J ¼ 7 Hz,
OCH2CH2), 3.89 (s, 6H, CO2Me), 1.74 (m, 8H, OCH2CH2),
1.41 (m, 8H, OCH2CH2CH2), 1.22–1.30 (m, 32H, CH2), and
0.86 (m, 12H, CH3). HRMS (MALDI-TOF). Found: m/z
1183.710. Calcd for [M + Na]þ 1183.706.
Compound G2. G1 (3.72 g, 3.20 mmol) was converted to the
bromide (3.85 g, 3.14 mmol, 98%) according to Procedure A
(CBr4, 2.65 g, 8.00 mmol; PPh3, 2.10 g, 8.00 mmol; THF 24
mL). This bromide (3.67 g, 3.00 mmol) was allowed to react with
3 (0.574 g, 3.15 mmol) in the presence of K2CO3 (0.518 g, 3.75
mmol) and 18-crown-6 ether (0.0596 g, 0.225 mmol) in acetone
(64 mL) for 2 h according to Procedure B, to give G1.5 (4.34 g,
93%). G1.5 (3.98 g, 3.00 mmol) was converted to the bromide
(3.75 g, 90%) according to Procedure A (CBr4, 5.97 g, 18.0 mmol;
PPh3, 4.72 g, 18.0 mmol; THF, 40 mL). This bromide (2.63 g, 1.89
mmol) was allowed to react with 4 (0.126 g, 0.900 mmol) in the
presence of K2CO3 (0.622 g, 4.50 mmol) and 18-crown-6
(0.0713 g, 0.270 mmol) in acetone (36 mL) for 2 h according to
Procedure B to give G2 (2.28 g, 91%). 1H NMR (500 MHz, CDCl3)
ꢀ 7.69 (s, 4H, Ar), 7.68 (s, 2H, Ar), 7.57 (m, 4H, Ar), 7.56 (m, 2H,
Ar), 7.24 (s, 4H, Ar), 7.21 (s, 2H, Ar), 6.67 (d, 4H, J ¼ 2 Hz, Ar),
6.59 (d, 2H, J ¼ 2 Hz, Ar), 6.55 (m, 10H, Ar), 6.49 (t, 1H, J ¼ 2
Hz, Ar), 6.39 (t, 4H, J ¼ 2 Hz, Ar), 4.99–5.02 (4s, 24H,
ArOCH2Ar0), 4.59 (d, 2H, J ¼ 6 Hz, CH2OH), 3.91 (t, 16H, J ¼
7 Hz, OCH2CH2), 3.88 (s, 18H, CO2Me), 1.99 (t, 1H, J ¼ 7 Hz,
OH), 1.74 (m, 16H, OCH2CH2), 1.41 (m, 16H, OCH2CH2CH2),
1.26–1.31 (m, 64H, CH2), and 0.86 (t, 24H, J ¼ 7 Hz, CH3).
HRMS (MALDI-TOF). Found: m/z 2776.541, 2792.536. Calcd
for [M + Na]þ 2776.543, [M + K]þ 2792.517.
Experimental
THF and CH2Cl2 were distilled from sodium diphenylketyl and
CaH2, respectively. Other reagents were obtained commercially
and used as received. 1H NMR spectra (500 MHz) were measured
on a JEOL LAMBDA-500 spectrometer and chemical shifts were
reported in the ꢀ scale relative to CHCl3 (7.24) and DMSO-d6
(2.49) in ppm. MALDI-TOF-MS was performed on an Applied
Biosystems Voyager System DE-DTR using 2-(4-hydroxyphenyl-
azo)benzoic acid as a matrix. GPC experiments were performed
on a Japan Analytical Industry Co. model LC-908 with JAIGEL
2H/2.5H (preparative) and TOSOH model GPC-8020 (analytical).
The analytical chromatograms were obtained under the following
conditions; columns, TSKgel super HM-H/HM-M/H3000; sol-
ꢁ
vent, CHCl3; flow rate, 0.6 mL/min; column temperature, 30 C.
Medium-pressure preparative column chromatography was per-
formed with Yamazen PUMP540-2CSC and Wakogel C-200 silica
gel. DLS was performed on an Otsuka Electronics model
DLS-6000. The intensity of scattering light was analyzed on a
Shimadzu RF-5300PC spectrofluorometer.
Methyl 3-(t-Butyldimethylsilyloxy)-5-(hydroxymethyl)ben-
zoate (2). LiAlH4 (1.90 g, 50.0 mmol) was added in portions to
a solution of 1 (32.5 g, 100 mmol) in THF (100 mL) at 0 ꢁC.
The mixture was stirred vigorously for 30 min, then quenched
slowly with water (3.60 mL). The solid was filtered and washed
with CH2Cl2 (500 mL). The combined filtrate was concentrated
and purified by column chromatography (silica gel, hexane/
1
AcOEt). Colorless oil. Yield 16.3 g (55%). H NMR (500 MHz,
CDCl3) ꢀ 7.59 (s, 1H, 6-C6H3), 7.37 (s, 1H, 2-C6H3), 7.03 (s,
1H, 4-C6H3), 4.66 (s, 2H, CH2), 3.87 (s, 3H, –CO2Me), 1.94 (s,
1H, –OH), 0.97 (s, 9H, t-Bu), and 0.19 (s, 6H, Me).
Methyl 3-Hydroxy-5-(hydroxymethyl)benzoate (3). A solu-
tion of KF (5.81 g, 100 mmol) in MeOH (150 mL) was added to 2
(14.7 g, 50.0 mmol) at room temperature, and then the solution was
stirred for 10 min, followed by the addition of 4 M HCl aq. (12.5
mL). After removal of the solvent, MeOH (20 mL), CHCl3 (400
mL), and anhydrous Na2SO4 were added. The solid was filtered
and washed with CHCl3 (200 mL), and the combined filtrate was
evaporated. The residue was recrystallized with hot CHCl3 (50
1
mL). Yield 8.70 g (96%). H NMR (500 MHz, DMSO-d6) ꢀ 9.74
(s, 1H, OH), 7.36 (s, 1H, 6-C6H3), 7.19 (s, 1H, 2-C6H3), 6.97 (s,
1H, 4-C6H3), 5.25 (t, 1H, J ¼ 6 Hz, CH2OH), 4.46 (d, 2H, J ¼ 6
Hz, CH2), and 3.81 (s, 3H, CO2Me).
General Procedure for the Synthesis of the Benzylic Bro-
mide from the Alcohol (Procedure A). CBr4 was dissolved in
Compound G3. G2 (1.76 g, 0.640 mmol) was converted to the
bromide (1.73 g, 96%) according to Procedure A (CBr4, 1.27 g,
3.84 mmol; PPh3, 1.01 g, 3.84 mmol; THF 20 mL). This bromide
(1.58 g, 0.560 mmol) was allowed to react with 3 (0.107 g, 0.588
mmol) in the presence of K2CO3 (97 mg, 0.70 mmol) and 18-
crown-6 (11 mg, 0.042 mmol) in acetone (30 mL) for 2 h according
to Procedure A, to give G2.5 (1.22 g, 75%). G2.5 was converted to
the bromide (0.897 g, 91%) according to Procedure A (CBr4, 0.657
g, 1.98 mmol; PPh3, 0.519 g, 1.98 mmol; THF, 40 mL). This bro-
mide (0.815 g, 0.273 mmol) was allowed to react with 4 (18.2 mg,
0.130 mmol) in the presence of K2CO3 (90 mg, 0.65 mmol) and 18-
crown-6 (10 mg, 0.039 mmol) in acetone (16 mL) for 2 h according
ꢁ
a solution of benzylic alcohol in THF at 0 C for 1–2 min. PPh3
was added, and when all solids were dissolved (1–2 min), the solu-
tion was stirred under N2 at room temperature. A white precipitate
appeared after 5–8 min. Before the color turned yellowish, the re-
sulting mixture was cooled to 0 ꢁC and monitored by 1H NMR and
TLC. CBr4 and PPh3 were added if necessary, until NMR and TLC
showed no starting benzylic alcohol. After the usual workup, the
product was purified by two consecutive column chromatography
(silica gel, CH2Cl2, then hexane/EtOAc).
General Procedure for Formation of the Ether Linkage
(Procedure B). The bromide, the phenolic compound (3 or 4),
18-crown-6, and K2CO3 were suspended in acetone, and the sus-
pension was purged with Ar. The resulting suspension was stirred
vigorously and heated at 80 ꢁC, The usual workup and purification
1
to Procedure B, to give G3 (0.422 g, 55%). H NMR (500 MHz,
CDCl3) ꢀ 7.64–7.68 (m, 14H, Ar), 7.51–7.57 (m, 14H, Ar),
7.17–7.24 (m, 14H, Ar), 6.65–6.68 (m, 12H, Ar), 6.47–6.55 (m,
25H, Ar), 6.37 (m, 8H, Ar), 4.94–4.99 (m, 56H, ArOCH2Ar0),
4.54 (s, 2H, CH2OH), 3.84–3.90 (m, 74H, OCH2CH2, CO2Me),

Y. Kikuzawa et al.
Bull. Chem. Soc. Jpn., 77, No. 5 (2004)
999
1.70 (m, 32H, OCH2CH2), 1.38 (m, 32H, OCH2CH2CH2), 1.23–
1.26 (m, 128H, CH2), and 0.84 (t, 48H, J ¼ 7 Hz, CH3). HRMS
(MALDI-TOF). Found: m/z 5962.223, 5978.103. Calcd for [M
+ Na]þ 5962.218, [M + K]þ 5978.191.
the viscosities and refractive indices of aqueous mixtures of THF
were taken from the literature.21 The intensities of the scattered
light were measured with light of 530 nm wavelength as incident
light, and the value of the peaktop intensity was adopted.
G3(CO2H). To a solution of G3 (96 mg, 0.016 mmol) in THF
(5 mL) was added KOH (0.39 g, 6.95 mmol) in water (0.6 mL).
MeOH (2 mL) was added to the two-phase mixture to afford a
We thank Mr. Seiji Makita (Institute for Molecular Science)
for MALDI-TOF-MS measurements, and the Research Center
for Computational Science of Okazaki National Research Insti-
tutes for the use of the NEC SX-7 computer. This work was
ꢁ
homogeneous solution. The reaction mixture was heated at 80 C
for 3 h and monitored by 1H NMR. The solution was quenched
by HCl aq. and worked up usually, to give the product (86 mg,
94%). 1H NMR (500 MHz, DMSO-d6) ꢀ 7.57–7.60 (m, 14H,
Ar), 7.41–7.45 (m, 14H, Ar), 7.26–7.27 (m, 14H, Ar), 6.69–6.71
(m, 12H, Ar), 6.60–6.63 (m, 4H, Ar), 6.50–6.55 (m, 21H, Ar),
6.32 (m, 8H, Ar), 5.00–5.08 (m, 56H, ArOCH2Ar0), 4.38 (s, 2H,
CH2OH), 3.82 (t, 32H, J ¼ 6 Hz, OCH2CH2), 1.58 (m, 32H,
OCH2CH2), 1.27 (m, 32H, OCH2CH2CH2), 1.16–1.22 (m, 128H,
CH2), and 0.78 (t, 48H, J ¼ 7 Hz, CH3).
supported by
a Grant-in-Aid for Exploratory Research
(No. 15655053) from Japan Society for the Promotion of
Science.
References
1
a) J. K. Young, G. R. Baker, G. R. Newkome, K. F. Morris,
and C. S. Johnson, Jr., Macromolecules, 27, 3464 (1994). b) S. A.
Kuzdzal, C. A. Monnig, G. R. Newkome, and C. M. Moorefield,
J. Chem. Soc., Chem. Commun., 1994, 2139. c) J. C. M. van Hest,
Molecular Dynamics Calculations. The initial geometry of
the G3 dendron was obtained by building a molecular model with
the Chem 3D program package. The atom types and force field
parameters were taken from general Amber force field (GAFF).
The RESP (restrained ElectroStatic Potential fit) atomic charge15
were obtained by a quantum mechanical calculation (HF/6-
´
M. W. P. L. Baars, C. Elissen-Roman, M. H. P. van Genderen, and
E. W. Meijer, Macromolecules, 28, 6689 (1995). d) M. Niwa, T.
Higashizaki, and N. Higashi, Tetrahedron, 59, 4011 (2003). e) R.
Sadamoto, N. Tomioka, and T. Aida, J. Am. Chem. Soc., 118,
3978 (1996). f) N. Tomioka, D. Takasu, T. Takahashi, and T. Aida,
Angew. Chem., Int. Ed., 37, 1531 (1998). g) G.-D. Zhang, N.
Nishiyama, A. Harada, D.-L. Jiang, T. Aida, and K. Kataoka,
Macromolecules, 36, 1304 (2003).
ꢀ
31G ) with the Gaussian 98,16 followed by a treatment with the
ANTECHAMBER module of AMBER 7. Since the G3 dendron
was too large to calculate, the G1 and G1.5 dendron were calculat-
ed, and the RESP charges were transferred to the G3 dendron for
atoms with similar topology and offset by a constant value to main-
tain the neutrality of the whole molecule. The Amber topology and
parameter files were generated by the tLeap modele of AMBER 7.
For a simulation with an explicit solvent, a rectangular box was
2
a) D. J. Pesak, J. S. Moore, and T. E. Wheat, Macromole-
cules, 30, 6467 (1997). b) D. J. Pesak and J. S. Moore, Tetra-
hedron, 53, 15331 (1997).
´
a) C. J. Hawker, K. L. Wooley, and J. M. J. Frechet, J.
3
ꢀ
added, providng at least 10 A of CHCl3 around any atom. Molecu-
´
Chem. Soc., Perkin Trans. 1, 1993, 1287. b) A. Ritzen and T. Frejd,
Eur. J. Org. Chem., 2000, 3771. c) J. G. Weintraub, S. Broxer, N.
M. Paul, and J. R. Parquette, Tetrahedron, 57, 9393 (2001). d) M. J.
Laufersweiler, J. M. Rohde, J.-L. Chaumette, D. Sarazin, and J. R.
Parquette, J. Org. Chem., 66, 6440 (2001).
lar-dynamics calculations were carried out by using the SANDER
module of Amber 7.0 with SHAKE17 applied to all hydrogen atoms
and 2-fs time steps. Berendsen temperature coupling18 was used
with a time constant of 0.5 ps. The in-vacuo simulation protocol
involved three sequential phases: energy minimization for 500
steps, slow heating from 0 to 300 K over 10 ps, and production
run at 300 K for 100 ps. Generalized Born model19 with dielectric
4
R. C. van Duijvendobe, A. Rajanayagam, and G. J. Koper,
Macromolecules, 33, 46 (2000).
a) J. W. Leon, M. Kawa, and J. M. J. Frechet, J. Am. Chem.
´
5
ꢀ
constant of 1.0 was used, and a 20-A cutoff was applied. The sim-
Soc., 118, 8847 (1996). b) G. R. Newkome, C. D. Weis, C. N.
Moorefield, and I. Weis, Macromolecules, 30, 2300 (1997).
ulation with a solvent box involved seven sequential phases: ener-
ꢀ 2
gy minimization for 500 steps with constraints of 500 kcal/mol/A
on the solute atoms, two-stage initial equilibration of the solvent
6
a) M. Brettreich and A. Hirsch, Tetrahedron Lett., 39, 2731
(1998). b) A. Quaranta, D. J. McGarvey, E. J. Land, M. Brettreich,
S. Burghardt, H. Schonberger, A. Hirsch, N. Gharbi, F. Moussa, S.
ꢀ 2
with constraints of 10 kcal/mol/A on the solute atoms (stage 1:
¨
constant volume, heating from 100 to 300 K over 1 ps and keeping
at 300 K for 24 ps; stage: 24 ps at 300 K, constant pressure at 1
atm), two successive minimizations for 600 steps with constraints
Leach, H. Gottinger, and R. V. Bensasson, Phys. Chem. Chem.
¨
Phys., 5, 843 (2003).
7
a) P. R. Ashton, D. W. Anderson, C. L. Brown, A. N.
ꢀ 2
of 50 kcal/mol/A and 0 on the solute atoms, pre-equilibration of
Shipway, J. F. Stoddart, and M. S. Tolley, Chem.—Eur. J., 4,
781 (1998). b) Y. Wang, C. M. Cadona, and A. E. Kaifer, J. Am.
Chem. Soc., 121, 9756 (1999). c) C. M. Cardona, T. D. McCarley,
and A. E. Kaifer, J. Org. Chem., 65, 1857 (2000). d) C. M.
Cardona, J. Alvarez, A. E. Kaifer, T. D. McCarley, S. Pandey,
G. A. Baker, N. J. Bonzagni, and F. V. Bright, J. Am. Chem.
Soc., 122, 6139 (2000). e) C. M. Cardona, T. Wilkes, W. Ong,
A. E. Kaifer, T. D. McCarley, S. Pandey, G. A. Baker, M. N. Kane,
S. N. Baker, and F. V. Bright, J. Phys. Chem. B, 106, 8649 (2002).
the whole system at constant volume (heating from 100 to 300 K
over 1 ps and keeping at 300 K for 24 ps), and production run at
300 K, 1 atm for 100 ps. Long-range electrostatic forces were eval-
uated using the particle mesh Ewald method.20 A 9-A cutoff was
ꢀ
applied. A constant procedure was maintained with isotropic mole-
cule based scaling. All calculations were performed on an NEC
SX-7 computer.
Micellization Experiments. Aqueous NaOH (10 mg, 0.25
mmol in 0.3 mL) was added to a solution of G3(CO2H) (101
mg, 0.018 mmol) in CHCl3 (30 mL). The turbid mixture was stir-
red for 10 min. After removal of the solvent, G3(CO2Na) was dried
in vacuo. For DLS measurements, x mL of THF and (2 ꢂ x) mL of
water was added to 2.0 mg of G3(CO2Na) (0.33 mmol), and the
mixture was sonicated for 5 min. In order to calculate diameters,
8
a) S. C. Zimmerman, F. Zeng, D. E. C. Reichert, and S. V.
Kolotuchun, Science, 271, 1095 (1996). b) V. Percec, W.-D. Cho,
and G. Ungar, J. Am. Chem. Soc., 122, 10273 (2000). c) F. Zeng, S.
C. Zimmerman, S. V. Kolotuchun, D. E. C. Reichert, and Y. Ma,
Tetrahedron, 58, 825 (2002).
9
M. E. Piotti, F. Rivera, R. Bond, C. J. Hawker, and J. M. J.

1000 Bull. Chem. Soc. Jpn., 77, No. 5 (2004)
Dendrons with Internal Carboxyl Groups
´
Frechet, J. Am. Chem. Soc., 121, 9471 (1999).
´
M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas,
J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennussi, C.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P.
Y. Ayala, Q. Cu, K. Morokuma, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. Coslowski, J. V. Ortiz, A. G.
Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M.
A. Al-Laham, C. Y. Poeng, A. Nanayakkara, C. Gonzalez, M.
Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen, M. W.
Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle, and J. A.
Pople, ‘‘Gaussian 98 (Revision A.7),’’ Gaussian, Inc., Pittsburgh,
PA (1998).
17 J.-P. Ryckaert, G. Ciccotti, and H. J. C. Berendsen, J.
Comput. Phys., 23, 327 (1977).
18 H. J. C. Berendsen, J. P. M. Postma, W. F. van Gunsteren,
A. DiNola, and J. R. Haak, J. Chem. Phys., 81, 3684 (1984).
19 V. Tsui and D. A. Case, Biopolymers (Nucl. Acid. Sci.), 56,
275 (2001).
10 D. Felder, M. G. Nava, M. P. Carreon, J.-F. Eckert, M.
Luccisano, C. Schall, P. Masson, J.-L. Gallani, B. Heinrich, D.
Guillon, and J.-F. Nierengarten, Helv. Chim. Acta, 85, 288 (2002).
11 a) P. B. Rheiner and D. Seebach, Chem.—Eur. J., 5, 3221
(1999). b) B. Forier and W. Dehaen, Tetrahedron, 55, 9829 (1999).
12 D. R. Vutukuri, K. Sivanandan, and S. Thayumanavan,
Chem. Commun., 2003, 796.
13 F. Voglte and M. Fischer, Angew. Chem., Int. Ed., 38, 884
(1999).
¨
14 D. A. Case, D. A. Pearlman, J. W. Caldwell, T. E.
Cheatham, III, J. Wang, W. S. Ross, C. L. Simmerling, T. A.
Darden, K. M. Merz, R. V. Stanton, A. L. Cheng, J. J. Vincent,
M. Crowley, V. Tsui, H. Gohlke, R. J. Radmer, Y. Duan, J. Pitera,
I. Massova, G. L. Seibel, U. C. Singh, P. K. Weiner, and P. A.
Kollman, ‘‘Amber 7,’’ University of California, San Francisco
(2002).
15 C. I. Bayly, P. Cieplak, W. D. Cornell, and P. A. Kollman,
J. Phys. Chem., 97, 10269 (1993).
16 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A.
Montgomery, Jr., R. E. Stratmann, J. C. Burant, S. Dappricg, J.
20 T. Darden, D. York, and L. Pedersen, J. Chem. Phys., 98,
10089 (1993).
21 T. M. Aminabhavi and B. Gopalakrishna, J. Chem. Eng.
Data, 40, 856 (1995).