Binding and catalytic properties of 2-0-and 3-o-permethylated cyclodextrins

Article abstract of DOI:10.1246/bcsj.82.196

Hexakis(3-0-methyl)-α,-cyclodextrin (3α) bound tom-and p-nitrophenolate ions more strongly, whereas hexakis(2-0-methy1)-α- cyclodextrin (2α) bound less strongly than native α-cyclodextrin. ROESY spectra showed that the 3-0-mefhyl groups of 3a interact with the guest protons, whereas 2-0-methyl groups of 2a do not. 3a accelerated and 2a decelerated the cleavage of m-nitrophenyl acetate in an alkaline solution, suggesting that the C(2)-OH of α-cyclodextrin is more catalytic than the C(3)-OH. However, the catalytic effect of 3a was much smaller than that of native a-cyclodextrin. Loss of hydrogen bonding between the C(3)-OH and C(2)-OH by 3-O-permethylation is responsible for the small catalytic effects of 3α. Similar results were obtained for β-cyclodextrin analogs.

Full text of DOI:10.1246/bcsj.82.196

1
96
Bull. Chem. Soc. Jpn. Vol. 82, No. 2, 196–201 (2009)
© 2009 The Chemical Society of Japan
Binding and Catalytic Properties of 2-O- and
3
-O-Permethylated Cyclodextrins
Takuya Nagata, Keisuke Yoshikiyo, Yoshihisa Matsui, and Tatsuyuki Yamamoto*
Faculty of Life and Environmental Science, Shimane University, 1060 Nishikawatsu, Matsue 690-8504
Received June 23, 2008; E-mail: tyamamot@life.shimane-u.ac.jp
Hexakis(3-O-methyl)-¡-cyclodextrin (3¡) bound to m- and p-nitrophenolate ions more strongly, whereas hexakis(2-
O-methyl)-¡-cyclodextrin (2¡) bound less strongly than native ¡-cyclodextrin. ROESY spectra showed that the 3-O-
methyl groups of 3¡ interact with the guest protons, whereas 2-O-methyl groups of 2¡ do not. 3¡ accelerated and 2¡
decelerated the cleavage of m-nitrophenyl acetate in an alkaline solution, suggesting that the C(2)ÍOH of ¡-cyclodextrin
is more catalytic than the C(3)ÍOH. However, the catalytic effect of 3¡ was much smaller than that of native ¡-
cyclodextrin. Loss of hydrogen bonding between the C(3)ÍOH and C(2)ÍOH by 3-O-permethylation is responsible for the
small catalytic effects of 3¡. Similar results were obtained for ¢-cyclodextrin analogs.
Cyclodextrins (CDs) are cyclic oligomers composed of
and heptakis(3-O-methyl)-¢-CD (3¢), on the cleavage of 3-
and 4-nitrophenyl acetates (m-NPA and p-NPA, respectively),
together with their complexation with 3- and 4-nitrophenols
(m-NP and p-NP, respectively), in an alkaline solution. It is
expected that the obtained results will afford some insight
into the role of the C(2)Í and C(3)ÍOHs of CDs in acyl-
transfer reactions. Procedures for the selective permethylation
at the C(2)Í and C(3)ÍOHs of CDs have already been
six (¡-CD), seven (¢-CD), or more ¡-D-glucopyranose units
and stereoselectively catalyze the cleavage of phenyl esters
in alkaline solutions.1 The reaction proceeds via the prior
inclusion of ester within the CD cavity, followed by the
nucleophilic attack of the alkoxide ion derived from a
secondary hydroxy group of CD to form the phenolate ion
and acylated CD. However, there has been controversy as to
whether the C(2)ÍOH or the C(3)ÍOH of CD is ionized and
attacks the ester carbonyl first. Bender et al. inferred at first that
the C(3)ÍOH appears to be ionized.1b Later, they suggested that
the C(2)ÍOH is responsible for the accelerated release of
7
developed by Ashton et al. and Bergeron, Meeley, and co-
8
worker respectively.
Results and Discussion
1
c
2
phenol. Breslow et al. found both C(2)- and C(3)-acylated
compounds in the reaction mixtures of aryl esters with ¢-CD
in alkaline solutions. They suggested that the alkoxide ion
derived from the C(2)ÍOH of ¢-CD attacks the substrates first,
followed by equilibration of sugar esters with neighboring
Binding Properties of the Permethylated CDs. Binding
constants (K ) were determined by UVÍvis spectrophotometry
a
for complexation of the permethylated CDs and the corre-
sponding native CDs with p-NP and m-NP in a carbonate
buffer solution (pH 10.60) at 298 K (Table 1). Changes in
absorbance of m-NP with the addition of ¢-CD, 2¢, and 3¢
2b
hydroxy groups under basic conditions. On the other hand,
3
Bergeron and Burton inferred that the C(2)ÍOH and C(3)ÍOH
were too small to determine the K values for these systems.
a
1
do not differ substantially in catalytic activity. Subsequent
However, H NMR signals of m-NP in D O containing
2
4
Õ3
computer simulation of phenyl ester cleavage by ¢-CD has
0.1 mol dm
Na CO were appreciably changed with the
2 3
indicated that the alkoxide ion derived from the C(2)ÍOH is
more inclined to the cavity interior than that derived from the
C(3)ÍOH, and only the reaction path leading to C(2)-acylation
shows an activation barrier lower than that for the correspond-
addition of these hosts, and the K values for these systems
were determined by NMR shift titration and are also shown in
a
Table 1. Interestingly, the K value for a 3¡Íp-NP system was
a
remarkably larger than that for a native ¡-CDÍp-NP system by
a factor of about 8, whereas that for a 2¡Íp-NP system was
smaller than that for an ¡-CDÍp-NP system. A similar tend-
ency was found for the corresponding ¢-CDÍp-NP and ¡-CDÍ
1
5
ing reference reaction in water. Furthermore, H NMR studies
of CDs in aqueous solutions showed that the C(3)ÍOH groups
are hydrogen donors and the C(2)ÍOH groups are the acceptors
in the formation of hydrogen bonds between them. These
findings suggest that the alkoxide ion derived from the C(2)Í
m-NP systems. The K value for a 3¢Ím-NP system was also
a
larger than that of a 2¢Ím-NP system, though it was slightly
smaller than that of a native ¢-CDÍm-NP system. These results
indicate that the 3-O-methyl groups of 3¡ and 3¢ contribute to
the stabilization of inclusion complexes with m-NP and p-NP,
whereas the 2-O-methyl groups of 2¡ and 2¢ do not. This
6
OH attacks ester first. On the other hand, Fukudome et al.
showed that C(3)Ímonothio-CDs are much more effective than
C(2)Ímonothio-CDs in promoting acyl transfer, suggesting that
the C(3)ÍOH is more reactive than the C(2)ÍOH.
1
The present work deals with the catalytic effects of hexakis-
presumption was confirmed by H NMR spectroscopy.
(
2-O-methyl)-¡-CD (2¡), hexakis(3-O-methyl)-¡-CD (3¡),
Figure 1 shows the ROESY (rotating-frame nuclear
Overhauser enhancement spectroscopy) spectra of mixtures of
and their ¢-CD analogs, heptakis(2-O-methyl)-¢-CD (2¢)
Published on the web February 13, 2009; doi:10.1246/bcsj.82.196

T. Nagata et al.
Bull. Chem. Soc. Jpn. Vol. 82, No. 2 (2009)
197
Table 1. Binding Constants (Ka), Differences (ž) in Molar
Absorbance between Complexed and Free Guests, and
Wavelength (­) Observed for Inclusion Complexes of
Native and Permethylated CDs with Nitrophenolates in a
Carbonate Buffer (pH 10.60) at 298 K, Together with Ka
CDÍm-NP is shown in Figure 1d. The 2-H and 4-H adjacent to
the nitro group of m-NP gave strong cross-peaks with the C(3)Í
H of ¡-CD, together with weak cross-peaks between the 5-H
and the C(3)ÍH and between the 2-H and the C(5)ÍH,
suggesting that the nitro group of m-NP also locates in the
vicinity of the C(5)ÍHs. Similar results were obtained in the
case of the 3¡Ím-NP system (Figure 1f). That is to say, the
system gave strong cross-peaks between the 2-H and 4-H of m-
NP and the C(3)ÍH of 3¡, together with weak cross-peaks
between the 2-H and 4-H of m-NP and the C(5)ÍH of 3¡,
suggesting that the nitro group of m-NP locates in the vicinity
of the C(5)ÍH of 3¡. However, the inclusion of m-NP within
the cavity of 3¡ is shallower than that of the 3¡Íp-NP or ¡-
CDÍm-NP system, since the 5-H of m-NP gave no cross-peak
with the C(3)ÍH of 3¡. In addition, all the protons of m-NP
gave strong or weak cross-peaks with the 3-O-methyl protons
of 3¡, suggesting that the 3-O-methyl groups are inclined to the
cavity interior and take part in complexation. On the other
hand, the 2¡Ím-NP system gave 2 cross-peaks between the 2-H
and 4-H of m-NP and the C(3)ÍH of 2¡, together with a very
weak cross-peak between the 2-H of m-NP and the C(5)ÍHs of
1
Values Determined by H NMR Shift Titration in 0.1
Õ3
mol dm Na2CO3/D2O at 298 K
Ka/molÕ1 dm3
ž/mmolÕ1 dm cm
3
Õ1
Host
­/nm
p-NP
¡
-CD
1860
1420
14200
683
33
1080
3.96
5.84
6.78
2.22
2.37
4.06
417
419
419
420
420
419
2¡
3¡
¢-CD
2
3
¢
¢
m-NP
¡
-CD
233
107
434
161
33
0.507
0.556
0.623
450
448
453
2
3
¡
¡
a)
¢-CD
2
¡ (Figure 1e). No cross-peak was observed between any
a)
2
3
¢
¢
protons of m-NP and 2-O-methyl protons of 2¡. These results
indicate that m-NP is shallowly included within the cavity
of 2¡ and also that 2-O-methyl groups of 2¡ are directed
toward the outside of the cavity, as was observed for the 2¡Íp-
NP system. Possible structures of the inclusion complexes of
m-NP with 2¡ and 3¡ are illustrated in Figures 2c and 2d,
respectively.
a)
138
1
a) The Ka values were determined by H NMR shift titration in
Õ3
0
.1 mol dm Na2CO3.
¡
-CD, 2¡, and 3¡ with p-NP and m-NP in alkaline D O
2
solutions. For a native ¡-CDÍp-NP system (Figure 1a), the
ortho-H (2,6-H in the figure) of p-NP gave a cross-peak with
the C(3)ÍH of the host, and the meta-H (3,5-H in the figure),
cross-peaks with both the C(3)Í and C(5)ÍHs. Similar cross-
peaks were found in the case of a 3¡Íp-NP system (Figure 1c).
These results suggest that the p-NP ion deeply penetrates into
the CD cavities of ¡-CD and 3¡ in such a manner that the nitro
group locates in the vicinity of the C(5)ÍH. On the other hand,
the ortho- and meta-Hs of p-NP gave cross-peaks only with the
C(3)ÍH of 2¡ and no cross-peak with the C(5)ÍH (Figure 1b),
suggesting that p-NP is included within the cavity of 2¡ less
deeply than the cases of ¡-CD and 3¡. Interestingly, the 2-O-
methyl protons of 2¡ gave no cross peaks with the p-NP
protons, suggesting that 2-O-methyl groups of 2¡ are far apart
from the p-NP protons in an inclusion complex. In contrast, the
Catalytic Properties of the Permethylated CDs. Effects
of the permethylated CDs and native CDs on the rate of
cleavage of m-NPA and p-NPA were spectrophotometrically
examined in a carbonate buffer (pH 10.60) at 298 K. Figure 3
shows, for an example, the effects of 2¡ and 3¡ on the
cleavage of m-NPA. Interestingly, 2¡ retarded and 3¡
accelerated the cleavage. Such results were analyzed by a
nonlinear least-squares curve-fitting analysis of changes in rate
constants (k ) with the host concentrations (c ), based upon
obsd
0
an assumption of 1:1 complexation,1a to give the binding (Ka)
and rate constants (k ) for examined CDÍNPA complexes. The
c
obtained curves (as shown by solid lines in Figure 3) were
satisfactorily fitted to the data. The obtained K values and
a
ratios (k /k ) of k to the rate constants (k ) in the absence of
c
un
c
un
3
-O-methyl protons of 3¡ gave a clear cross-peak with the
CDs are summarized in Table 2. The K and k /k values for
a c un
ortho-H of p-NP, indicating that the 3-O-methyl groups of 3¡
are in close contact with the ortho-H of p-NP. Presumably, 2-O-
methyl groups of 2¡ are directed toward the outside of the
cavity, whereas 3-O-methyl groups of 3¡ are inclined to the
cavity interior. Possible structures of the inclusion complexes
of p-NP with 2¡ and 3¡ are illustrated in Figures 2a and 2b,
respectively. The structure of the ¡-CDÍp-NP complex should
be similar to that of the 3¡ÍpNP complex. van der Waals
interactions between the 3-O-methyl groups of 3¡ and the
ortho-H of p-NP, as well as an increase in hydrophobicity of
the ¡-CD cavity by 3-O-permethylation, should be responsible
¡- and ¢-CD complexes with m-NPA and p-NPA were roughly
1
a
equal to those reported by VanEtten et al. However, the
obtained Ka values for the nitrophenyl acetates were very
different from those for the corresponding nitrophenolate ions.
Similar large differences in the K values between various
a
phenyl acetates and the corresponding phenolates have been
1
c
reported thus far. The difference are brought about by
difference in the chemical properties and molecular structures
in phenyl acetate and the corresponding phenolates. Not only
3¡ but also 3¢ accelerated the cleavage of m-NPA, suggesting
that the alkoxide ion derived from the C(2)ÍOH of 3¡ and 3¢
are catalytically active. However, the k /k values for 3¡ and
for the observed large K values for a 3¡ÍNP system. 2-O-
a
c
un
Methyl groups of 2¡, directed to the outside of the cavity, are
not available for strengthening such van der Waals and
hydrophobic interactions. The ROESY spectrum of native ¡-
3¢ were much smaller than those for native ¡- and ¢-CDs,
respectively, though the K values were similar to each other. It
a
is known that the hydrogens of C(3)ÍOH groups of native CDs

1
98
Bull. Chem. Soc. Jpn. Vol. 82, No. 2 (2009)
3,5-H 2,6-H
2-O- and 3-O-Permethylated Cyclodextrins
4-H 2-H 5-H
6-H
a
d
H5
H3
H5
H3
3
,5-H
2,6-H
4-H 2-H 5-H
6-H
b
e
OMe
OMe
H3
H5
H3
3
,5-H
2,6-H
4-H 2-H 5-H
6-H
c
f
H3
OMe
H3
OMe
H5
H5
Õ3
Figure 1. ROESY spectra of p-NP or m-NP in the presence of ¡-CD, 2¡, or 3¡ in D2O containing 0.1 mol dm Na2CO3 at 298 K.
Õ3
Õ3
Õ3
Õ3
(
a) 10.2 mmol dmÕ3 p-NP + 42.0 mmol dm ¡-CD, (b) 10.1 mmol dm p-NP + 13.8 mmol dm 2¡, (c) 10.1 mmol dm
Õ3
Õ3
Õ3
Õ3
Õ3
p-NP + 12.9 mmol dm 3¡, (d) 10.2 mmol dm m-NP + 31.6 mmol dm ¡-CD, (e) 10.2 mmol dm m-NP + 16.9 mmol dm
¡, (f) 10.2 mmol dm m-NP + 13.4 mmol dm 3¡.
Õ3
Õ3
2
are hydrogen bonded to the oxygens of the C(2)ÍOH groups of
3¡ and 3¢ are much smaller than those for native ¡- and ¢-
CDs. Different reasoning is also possible: The permethylation
will bring about a change in structure of the CD cavity, and the
geometries for inclusion complexes of 3¡ and 3¢ will differ
9
adjacent glucopyranose units not only in the crystal state but
5
also in an aqueous solution, and the hydrogen bonds stabilize
the alkoxide ion formed by the acid dissociation of a CD
1
b
3
hydroxy group to give a relatively low pK value (pK =
from those of native ¡- and ¢-CDs, respectively. In fact, the
a
a
1
b
10
1
2.1, 12.33 for ¡-CD). Since the C(3)ÍOH groups are
ROESY spectra described in the previous section showed that
the 3-O-methyl groups of 3¡ are inclined to the cavity interior,
whereas the C(3)ÍOH groups of native ¡-CD are directed to the
oxygens of the C(2)ÍOH groups of adjacent glucopyranose
5
hydrogen donors in native CDs, permethylation of the C(3)Í
OH groups breaks up such hydrogen bonding to destabilize the
alkoxide ion derived from the C(2)ÍOH. As a result the pKa
values of 3¡ and 3¢ increase and the reactivities of their C(2)Í
OHs decrease. This is one of reasons why the k /k values for
5
units. Thus, the distance between the catalytic site (the
alkoxide ion of host) and the reaction site (the carbonyl carbon
c
un

T. Nagata et al.
Bull. Chem. Soc. Jpn. Vol. 82, No. 2 (2009)
199
a)
-
Table 2. Equilibrium (K ) and Kinetic (k /k ) Parameters
O
a
c
un
-
for Complexes of Native and Permethylated CDs with
Nitrophenyl Acetates in a Carbonate Buffer (pH 10.60) at
O
H
H
H
H3CO
OCH3
OCH3
H CO
3
H
H
H
H
2
98 K
H3
H3
H3
H3
H
p-NPA
m-NPA
NO2
H5
H5
H5
H5
NO2
Host
-CD
Ka
kc/kun
Ka
kc/kun
/
molÕ1 dm3
/molÕ1 dm3
a
b
¡
120
30
140
160
250
50
2.9
17
280
13
40
180
47
420
0.49
9.8
260
0.46
9.5
2¡
0.41
0.66
9.3
0.68
5.1
H
-
O
H
3¡
¢-CD
H
-
H
H
O
H3CO
OCH3
OCH3
H3
H3CO
2¢
H3
H
H
H3
H3
3¢
H
NO2
a) kun = (5.00 « 0.05) © 10 sÕ1 (p-NPA) and (3.03 «
.05) © 10
Õ3
NO2
H5
H5
H5
H5
Õ3 Õ1
0
s
(m-NPA).
c
d
On the contrary to 3¡ and 3¢, 2¡ and 2¢ retarded the
cleavage of m-NPA. As mentioned above, the C(3)ÍOH of
native CD forms a hydrogen bond with the C(2)ÍOH. Similar
formation of hydrogen bonds will be possible, even after
permethylation at the C(2)ÍOH groups. If the C(3)ÍOH groups
are stabilized by hydrogen bonding, the acid dissociation of the
C(3)ÍOH to form the reactive alkoxide ion will be suppressed,
and the C(3)ÍOH groups become catalytically inert. Further-
more, the nucleophilic attack of hydroxide ion in bulk solution
on the substrate included in the cavity of 2¡ or 2¢ will be
retarded by steric hindrance. As shown in Table 2, the Ka
values for m-NPA complexes with 2¡ and 2¢ are significantly
larger than those for m-NPA complexes with 3¡ and 3¢,
respectively, suggesting that m-NPA is more deeply included
within the cavities of 2¡ and 2¢ than within those of 3¡
and 3¢ to be effectively protected from nucleophilic attack of
hydroxide ion in bulk solution.
Figure 2. Possible structures of the inclusion complexes
of 2¡ and 3¡ with p-NP and m-NP in alkaline solutions.
a) 2¡Íp-NP complex, (b) 3¡Íp-NP complex, (c) 2¡Ím-
NP complex, (d) 3¡Ím-NP complex.
(
b
a
Similar results were obtained for p-NPA systems, except 3¡.
Thus, the cleavage of p-NPA was decelerated by 2¡ and 2¢,
whereas accelerated by 3¢. Unexpectedly, 3¡ decelerated the
cleavage of p-NPA, though it accelerated the cleavage of
m-NPA. The Ka value for a 3¡Íp-NPA complex was about
-
3
[
CD] / mmol dm
Figure 3. Plots of the difference (Åkobsd) between the rates
of the cleavage of m-NPA in the presence (kobsd) and
in the absence (kun) of the permethylated CDs vs. CD
concentration in a carbonate buffer (pH 10.60) at 298 K.
Permethylated CDs added are 2¡ (a) and 3¡ (b),
respectively.
1
0 times larger than that for a 3¡Ím-NPA complex. In this
connection, the K value for a 3¡Íp-NP complex was also
a
about 8 times larger than that for a native ¡-CDÍp-NP complex
(Table 1). Thus, p-NPA is so deeply and so rigidly trapped
within the 3¡ cavity that the included p-NPA has difficulty
11
of guest) in an inclusion complex will be changed by the
permethylation, so that the activation energy of the reaction is
changed. At the present stage of investigation, it is difficult to
judge which is true or both. However, binding constants K for
the m-NP complexes with 3¡ and 3¢ were not very different
from those for the corresponding complex with native ¡- and
reaching the transition-state binding in which the alkoxide ion
derived from the C(2)ÍOH is linked to the acyl carbon. In
contrast, 3¢ accelerated the cleavage of p-NPA. The cavity of
3¢ is larger that of 3¡, and p-NPA included within the 3¢
cavity will have more freedom of motion and can more easily
reach the transition state than that in 3¡.
a
¢
-CDs, respectively (Table 1), and the K values for m-NPA
In conclusion, hexakis(3-O-methyl)-¡-CD (3¡) accelerated
the cleavage of m-NPA, whereas hexakis(2-O-methyl)-¡-CD
(2¡) decelerated the reaction in alkaline solution, suggesting
that the C(2)ÍOH of ¡-CD is more catalytic than the C(3)ÍOH.
Similar results were observed for their ¢-CD analogs. However,
the catalytic effects of 3¡ and 3¢ were much smaller than those
of native ¡- and ¢-CDs, respectively. Loss of hydrogen bonding
between the C(3)ÍOH and C(2)ÍOH by 3-O-permethylation are
responsible for the small catalytic effects of 3¡ and 3¢.
a
with 3¡ and 3¢ were also not very different from those for the
corresponding complex with native ¡- and ¢-CDs, respectively
(Table 2), suggesting that the change in structure of the CD
cavity with 3-O-permethylation is not so large. Hence, it is
more plausible that a decrease in reactivities of the C(2)ÍOH
with 3-O-permethylation is mainly responsible for the observed
low catalytic effects of 3¡ and 3¢ compared with those of
native ¡- and ¢-CDs.

2
00
Bull. Chem. Soc. Jpn. Vol. 82, No. 2 (2009)
Experimental
2-O- and 3-O-Permethylated Cyclodextrins
CDCl3): ¤ Õ5.1, Õ4.8, Õ3.8, 18.3, 18.5, 26.0, 26.3, 57.5, 62.6,
7
3.3, 81.2, 95.3.
Materials. The ¡- and ¢-CDs were supplied by Ensuiko Sugar
2 (2.52 g, 1.04 mmol) was dissolved in a THF solution
Õ3
Refining Co., Ltd. They were dried overnight in vacuo at 110 °C.
t-Butylchlorodimethylsilane was purchased from Shin-Etsu Chem-
ical Co., Ltd. Sodium hydride (50Í72% in oil), tetrabutylammo-
nium fluoride, 4-(N,N-dimethylamino)pyridine, and allyl bromide
were purchased from Wako Pure Chemical Industries, Ltd. Methyl
iodide and potassium t-butoxide were purchased from Tokyo
Kasei Kogyo Co., Ltd., and Kanto Chemical Co., Inc., respec-
tively. Silica gel 60 (Merck, 0.040Í0.063 mm), Sephadex G-25
containing tetrabutylammonium fluoride (1 mol dm , 16 mL),
and the resulting solution was refluxed overnight. The reaction
mixture was concentrated by evaporation, and the residue was
dissolved in water (7 mL). The aqueous solution was washed with
CH Cl (5 © 30 mL) to remove the tetrabutylammonium salt, and
2
2
chromatographed with an octadecyl-bonded silica gel column and
aqueous methanol (20, 40, and 60% methanol) as eluants to give
1
2¡ (0.56 g, 0.53 mmol, 51%): H NMR (400 MHz, D2O): ¤ 3.36
(
Amersham Biosciences), and octadecyl-bonded silica gel (Organo
(dd, J = 3.4, 10.2 Hz, 6H), 3.54 (s, 18H), 3.57 (t, J = 9.3 Hz, 6H),
3.78 (m, 6H), 3.86 (dd, J = 4.3, 12.6 Hz, 6H), 3.89 (dd, J = 1.8,
12.6 Hz, 6H), 4.00 (t, J = 10.0 Hz, 6H), 5.24 (d, J = 3.4 Hz, 6H);
Co.) were used for column chromatography. Organic solvents and
other reagents were also commercially available.
13
Apparatus.
The UVÍvis spectra were recorded using a
2
C NMR (100 MHz, D O): ¤ 61.4, 63.1, 74.4, 74.7, 83.6, 84.5,
Shimadzu UV-2100 UV/Vis spectrophotometer equipped with
a temperature-controlled cell holder. The NMR spectra were
recorded on a JEOL Model JNM-A400 FT NMR spectrometer
101.4.
2¢ was prepared in a similar manner: 1H NMR (400 MHz,
D O): ¤ 3.38 (dd, J = 3.4, 10.0 Hz, 7H), 3.56 (s, 21H), 3.61 (t,
2
(
400 MHz) with a sample tube of 5.0 mm diameter at 298 K. The
J = 9.5 Hz, 7H), 3.81 (m, 7H), 3.86 (d, J = 3.2 Hz, 14H), 3.99 (t,
phase-sensitive ROESY spectra were acquired with a mixing time
of 1000 ms and 512 © 256 data points, followed by zero-filling,
for the inclusion complexes of ¡-CD and permethylated CDs
J = 9.5 Hz, 7H), 5.26 (d, J = 3.7 Hz, 7H); 13C NMR (100 MHz,
D O): ¤ 61.8, 62.9, 74.0, 74.7, 83.8, 84.0, 101.7.
2
Preparation of 3¡ and 3¢. Hexakis(3-O-methyl)-¡-CD (3¡)
Õ3
8
with m-NP and p-NP in D2O containing 0.1 mol dm Na2CO3.
was prepared according to a procedure reported by Bergeron et al.
12
Methanol (¤ 3.343 ) was used as an internal reference for
In a typical run, ¡-CD (6.04 g, 6.21 mmol) was allowed to react
with a large excess of allyl bromide (70 g, 0.58 mol) in a mixture of
dimethyl sulfoxide (DMSO, 100 mL) and DMF (100 mL) contain-
ing barium oxide (30 g, 0.2 mol) and barium hydroxide octahydrate
1
H NMR measurements in D2O.
Preparation of 2¡ and 2¢. Hexakis(2-O-methyl)-¡-CD (2¡)
7
was prepared according to a procedure reported by Ashton et al. In
a typical run, ¡-CD (9.92 g, 10.2 mmol) was allowed to react with
t-butylchlorodimethylsilane (28.00 g, 185.8 mmol) in a solution of
N,N-dimethylformamide (DMF, 100 mL) and pyridine (60 mL)
containing a trace amount of 4-(N,N-dimethylamino)pyridine
(30 g, 0.095 mol) at room temperature under N for 3 days. The
reaction mixture was treated with an aqueous ammonia solu-
2
tion (25%, 50 mL), and precipitates were filtered off. The filtrate
was concentrated under vacuum. The residue was applied to a
SiO2 column and eluted with chloroform/ethyl acetate solu-
tions (chloroform:ethyl acetate = 10:0, 9:1, and 8:2 by volume).
Hexakis(2,6-di-O-allyl)-¡-CD (3, 3.79 g, 2.61 mmol, 42% yield)
was obtained from an eluate of 8:2 chloroform/ethyl acetate:
(
21 mg) at 100 °C under nitrogen for 18 h. The reaction mixture
was concentrated by evaporation, and the residue was partitioned
between water (150 mL) and CH2Cl2 (150 mL). The organic layer
Õ3
was washed with KHSO4 (0.5 mol dm , 150 mL), dried over
1
CaSO4, and concentrated by evaporation. The concentrate was
chromatographed on a SiO2 column with CH2Cl2 as an eluent
to give hexakis[2,6-bis(O-t-butyldimethylsilyl)]-¡-CD (1, 17.95 g,
H NMR (400 MHz, CDCl3): ¤ 3.45 (dd, J = 3.2, 9.8 Hz, 6H), 3.55
(t, J = 9.3 Hz, 6H), 3.70 (d, J = 2.7 Hz, 12H), 3.86 (d, J = 10.0
Hz, 6H), 4.00Í4.08 (m, 12H), 4.12 (t, J = 9.3 Hz, 6H), 4.22 (dd,
J = 7.0, 12.6 Hz, 6H), 4.44 (dd, J = 5.4, 12.7 Hz, 6H), 4.72
(s, 6H), 4.92 (d, J = 3.4 Hz, 6H), 5.20 (dd, J = 10.5, 15.4 Hz,
1
7
1
0
6
.66 mmol, 75% yield): H NMR (400 MHz, CDCl3): ¤ 0.038 (s,
8H), 0.043 (s, 18H), 0.14 (s, 18H), 0.15 (s, 18H), 0.88 (s, 54H),
.91 (s, 54H), 3.48 (t, J = 9.3 Hz, 6H), 3.55 (dd, J = 2.9, 9.8 Hz,
H), 3.70 (d, J = 11.2 Hz, 6H), 3.73 (d, J = 12.0 Hz, 6H), 3.95
1
3
12H), 5.28 (dd, J = 13.7, 15.9 Hz, 12H), 5.93 (m, 12H); C NMR
(100 MHz, CDCl3): ¤ 68.8, 70.4, 72.6, 73.2, 73.6, 78.6, 83.5,
101.4, 117.3, 118.7, 134.2, 134.8.
(
4
dd, J = 2.7, 11.2 Hz, 6H), 3.99 (t, J = 9.3 Hz, 6H), 4.37 (s, 6H),
.79 (d, J = 2.9 Hz, 6H); C NMR (100 MHz, CDCl3): ¤ Õ5.2,
1
3
The allyl ether 3 (3.50 g, 2.41 mmol) was dissolved in DMF
(100 mL). Sodium hydride (2.63 g) was added to the solution at
room temperature under N2. To the mixture, a solution of methyl
iodide (8.56 g, 60 mmol) in DMF (30 mL) was slowly added and
allowed to stand overnight. After the addition of water (10 mL), the
reaction mixture was taken up in chloroform (500 mL) and washed
5 times with 100 mL of water. The organic layer was dried over
Õ4.7, Õ4.6, 18.4, 18.8, 25.9, 26.2, 62.2, 72.0, 72.3, 74.6, 82.5,
1
02.6.
Sodium hydride (4.40 g) was added to a solution of 1 (4.65 g,
.98 mmol) in tetrahydrofuran (THF, 120 mL) with cooling by ice-
1
water under N2 bubbling. To the solution, methyl iodide (6.5 mL)
was added and stirred overnight. Methanol (15 mL) was added to
the reaction mixture to decompose remaining sodium hydride. The
reaction mixture was concentrated by evaporation, and the residue
was partitioned between water (130 mL) and CH2Cl2 (250 mL).
The organic layer was washed with water (100 mL) and aqueous
NaCl solutions (2 © 100 mL), dried over CaSO4, and concentrated
by evaporation. The concentrate was chromatographed with a
SiO2 column and hexane/acetone (99:1, v/v) as an eluent to
CaSO , filtrated, and concentrated under vacuum to give a crude
4
product of hexakis(2,6-di-O-allyl-3-O-methyl)-¡-CD (4, 5.37 g):
1
H NMR (400 MHz, CDCl3): ¤ 3.31 (dd, J = 3.1, 9.5 Hz, 6H), 3.57
(t, J = 9.3 Hz, 6H), 3.62 (t, J = 9.3 Hz, 6H), 3.63 (s, 18H), 3.71 (d,
J = 9.8 Hz, 6H), 3.78Í3.84 (m, 12H), 4.00 (dd, J = 5.4, 12.9 Hz,
6H), 4.06Í4.10 (m, 12H), 4.27 (dd, J = 5.6, 12.9 Hz, 6H), 4.98 (d,
J = 3.2 Hz, 6H), 5.13Í5.17 (m, 12H), 5.23Í5.33 (m, 12H), 5.94
give hexakis[2-O-methyl-3,6-bis(O-t-butyldimethylsilyl)]-¡-CD
(m, 12H); 13C NMR (100 MHz, CDCl ): ¤ 62.1, 69.0, 71.2, 71.4,
3
1
(
2, 3.57 g, 1.47 mmol, 74% yield): H NMR (400 MHz, CDCl3):
72.3, 80.0, 81.4, 82.4, 100.9, 116.4, 117.0, 135.1, 135.6.
¤
5
0.03 (s, 36H), 0.10 (s, 18H), 0.12 (s, 18H), 0.89 (s, 54H), 0.90 (s,
The crude compound 4 (5.37 g) was allowed to react with
potassium t-butoxide (5.40 g) in DMSO (60 mL) at 100 °C for 6 h
under nitrogen. The reactants were taken up in 400 mL of diethyl
4H), 2.94 (d, J = 7.1 Hz, 6H), 3.32 (s, 18H), 3.66 (br, 12H), 3.88
1
3
(
br, 6H), 4.11 (br, 12H), 5.17 (br, 6H); C NMR (100 MHz,

T. Nagata et al.
Bull. Chem. Soc. Jpn. Vol. 82, No. 2 (2009)
201
ether and washed with water (5 © 100 mL). The ether layer was
dried over CaSO4, filtered, and evaporated to give a crude product
of hexakis(3-O-methyl-2,6-di-O-prop-1-enyl)-¡-CD (5, 4.12 g):
¢-CD, 2¢, and 3¢, respectively. The Ka values were determined
by a nonlinear least-squares curve-fitting analysis of changes in
the chemical shift with the host concentration, based upon an
assumption of 1:1 complexation. The thus-calculated curves were
well-fitted to the observed data with correlation coefficient greater
than 0.999.
1
H NMR (400 MHz, CDCl3): ¤ 1.56 (d, J = 6.6 Hz, 18H), 1.64
(
(
d, J = 6.8 Hz, 18H), 3.50Í3.63 (m, 18H), 3.65 (s, 18H), 3.74
t, J = 3.5 Hz, 6H), 3.89 (d, J = 10.5 Hz, 12H), 4.18 (dd, J = 4.0,
1
1.8 Hz, 6H), 4.39 (m, 12H), 5.13 (d, J = 3.4 Hz, 6H), 5.93 (d,
Kinetics.
The rate of ester cleavage was measured by
1
3
J = 4.6 Hz, 6H), 6.09 (d, J = 4.6 Hz, 6H); C NMR (100 MHz,
CDCl3): ¤ 9.4, 9.5, 41.0, 61.6, 70.8, 71.3, 80.2, 81.6, 99.1, 101.1,
1
following the appearance of the absorption of the corresponding
phenoxide anion at 390 nm for m-NPA or at 410 nm for p-NPA in a
sodium carbonate buffer (pH 10.60) at 298 K. In a typical run,
2.00 mL of base solutions containing CDs were pipetted into a pair
of 1.00-cm quartz cells, one of which was used as a reference cell
and the other, as a sample cell. After thermal equilibrium at 298 K
01.4, 145.3, 145.8.
The crude product 5 (4.12 g) was allowed to react with mercuric
oxide (2.22 g) and mercuric chloride (2.22 g) in 50 mL of water and
5 mL of acetone by shaking for 10 min at room temperature. The
2
Õ3
mixture was centrifuged, and the supernatant was evaporated in
vacuo. The residue was dissolved in 100 mL of water and washed
with diethyl ether (3 © 50 mL). The aqueous layer was evaporated
in vacuo, and the residue was applied to a Sephadex G-25 column.
Fractions containing polysaccharide were combined and concen-
trated under vacuum to give 3¡ (0.80 g, 0.76 mmol, 32% yield
had been reached, 50 µL of 20 mmol dm m-NPA in methanol or
10 µL of 20 mmol dm p-NPA in methanol was added to the
Õ3
sample cell, and the change in absorbance was followed. The
results were treated by the least-squares curve-fitting analysis
according to the ordinary first-order rate equation. All the reactions
examined obeyed good first-order kinetics with respect to substrate
in both the absence and presence of CDs. The results were
analyzed by a nonlinear least-squares curve-fitting analysis of
changes in the rate constants with the host concentrations, based
1
based on 3): H NMR (400 MHz, D2O): ¤ 3.66 (s, 18H), 3.71 (t,
J = 9.3 Hz, 6H), 3.72 (dd, J = 2.7, 10.0 Hz, 6H), 3.82 (t, J = 10.0
Hz, 6H), 3.87 (s, 6H), 3.88 (s, 12H), 5.04 (d, J = 3.4 Hz, 6H);
13
1a
C NMR (100 MHz, D2O): ¤ 61.6, 63.2, 74.0, 74.9, 80.9, 84.9,
upon an assumption of 1:1 complexation, to give the binding and
1
03.6.
3
rate constants for CDÍNPA complexes.
¢ was prepared in a similar manner: 1H NMR (400 MHz,
D2O): ¤ 3.67Í3.80 (m, 28H), 3.69 (s, 21H), 3.88 (s, 14H), 5.08 (d,
J = 2.7 Hz, 7H); 13C NMR (100 MHz, D2O): ¤ 62.3, 63.0, 74.6,
References
7
4.7, 80.0, 85.2, 103.4.
1
a) R. L. VanEtten, J. F. Sebastian, G. A. Clowes, M. L.
Spectrophotometric Determination of Binding Constants for
Bender, J. Am. Chem. Soc. 1967, 89, 3242. b) R. L. VanEtten,
G. A. Clowes, J. F. Sebastian, M. L. Bender, J. Am. Chem. Soc.
1967, 89, 3253. c) D. W. Griffiths, M. L. Bender, Adv. Catal. 1973,
23, 209.
m-NP and p-NP Complexes with CDs. The binding constants
Ka) for the complexation of m-NP and p-NP with CDs were
determined by UVÍvis spectroscopy in a sodium carbonate buffer
(
(
0
2
pH 10.60) at 298 K. The concentrations of m-NP and p-NP were
2
a) R. Breslow, M. F. Czarniecki, J. Emert, H. Hamaguchi,
.207 and 0.051 mmol dmÕ3, respectively. In a typical run,
.00 mL of base solutions were pipetted into a pair of 1.00-cm
J. Am. Chem. Soc. 1980, 102, 762. b) G. L. Trainor, R. Breslow,
J. Am. Chem. Soc. 1981, 103, 154.
quartz cells, one of which was used as a reference cell and the
other, as a sample cell. After thermal equilibrium at 298 K had
been reached, aliquots of a CD solution containing m-NP or p-NP
were added stepwise to the sample cell. In each step, a UVÍvis
spectrum was recorded from 600 to 300 nm. The complexation was
observed at a wavelength where the largest change in absorbance
arose (448Í453 nm for m-NP and 417Í420 nm for p-NP). The Ka
values were determined by a nonlinear least-squares curve-fitting
analysis of changes in the absorbance with the host concentra-
tion, based upon an assumption of 1:1 complexation. The thus-
calculated curves were well-fitted to the observed data with
correlation coefficient greater than 0.999.
3
3664.
4
R. J. Bergeron, P. S. Burton, J. Am. Chem. Soc. 1982, 104,
V. Luzhkov, J. ¡qvist, J. Am. Chem. Soc. 1998, 120, 6131.
S. Bekiroglu, L. Kenne, C. Sandström, J. Org. Chem. 2003,
5
68, 1671.
6
M. Fukudome, Y. Okabe, D.-Q. Yuan, K. Fujita, Chem.
Commun. 1999, 1045.
7
P. R. Ashton, S. E. Boyd, G. Gattuso, E. Y. Hartwell, R.
Koeniger, N. Spencer, J. F. Stoddart, J. Org. Chem. 1995, 60, 3898.
8
R. J. Bergeron, M. P. Meeley, Y. Machida, Bioorg. Chem.
1976, 5, 121.
9
W. Saenger, in Environmental Effects on Molecular
Determination of Binding Constants for m-NP Complexes
with ¢-CDs by Means of NMR Shift Titration. The Ka values
for the complexation of m-NP with ¢-CD, 2¢, and 3¢ were
Structure and Properties, ed. by B. Pullman, D. Reidel Pub.,
Dordrecht, 1976, pp. 256Í305.
10 R. I. Gelb, L. M. Schwartz, J. J. Bradshaw, D. A. Laufer,
Bioorg. Chem. 1980, 9, 299.
11 O. S. Tee, C. Mazza, X.-X. Du, J. Org. Chem. 1990, 55,
3603.
12 Y. Matsui, S. Tokunaga, Bull. Chem. Soc. Jpn. 1996, 69,
2477.
1
determined by H NMR shift titration in D2O containing
Õ3
1
0
.1 mol dm Na2CO3 at 298 K. H NMR spectra were recorded
Õ3
for sample solutions containing 4.08 mmol dm m-NP and various
concentrations of ¢-CD, 2¢, or 3¢. 4-H, 2-H, and 5-H of m-NP
showed the largest changes in chemical shift with the addition of