Molecular puzzle ring: Pseudo[1]rotaxane from a flexible cyclodextrin derivative

Article abstract of DOI:10.1021/ja806620z

A pseudo[1]rotaxane formed by a flexible cyclodextrin (CD) derivative (1-R) with a bulky end group has been investigated on kinetic quantitation. 1-Rs have the cinnamamide moiety as a guest and a bulky end group (R) as a rate-determining moiety of the threading process. The R groups play an important role for the formation of pseudo[1]rotaxane, and kinetics of the self-inclusion process was found to be controlled by the size and shapes of the R groups. 1-Ad and 1-Me derivatives, which have an adamantyl and methyl end group, respectively, formed self-inclusion complexes by threading of the arm moiety with a conformational conversion of altrose from ¹C₄ form to ⁴C₁ form. Flexibility of the altro-α-CD cavity resulted in an induced fit (from ¹C₄ to ⁴C ₁) to the arm moiety, and introducing a bulky end group allowed the stability of this pseudo[1]rotaxane to be enhanced.

Full text of DOI:10.1021/ja806620z

Published on Web 11/19/2008
Molecular Puzzle Ring: pseudo[1]Rotaxane from a Flexible
Cyclodextrin Derivative
Atsuhisa Miyawaki, Paul Kuad, Yoshinori Takashima, Hiroyasu Yamaguchi, and
Akira Harada*
Department of Macromolecular Science, Graduate School of Science, Osaka UniVersity,
Toyonaka, Osaka 560-0043, Japan
Received August 21, 2008; E-mail: harada@chem.sci.osaka-u.ac.jp
Abstract: A pseudo[1]rotaxane formed by a flexible cyclodextrin (CD) derivative (1-R) with a bulky end
group has been investigated on kinetic quantitation. 1-Rs have the cinnamamide moiety as a guest and a
bulky end group (R) as a rate-determining moiety of the threading process. The R groups play an important
role for the formation of pseudo[1]rotaxane, and kinetics of the self-inclusion process was found to be
controlled by the size and shapes of the R groups. 1-Ad and 1-Me derivatives, which have an adamantyl
and methyl end group, respectively, formed self-inclusion complexes by threading of the arm moiety with
a conformational conversion of altrose from 1C4 form to 4C1 form. Flexibility of the altro-R-CD cavity resulted
in an induced fit (from 1C4 to 4C1) to the arm moiety, and introducing a bulky end group allowed the stability
of this pseudo[1]rotaxane to be enhanced.
9
10
than the C1 form. However, Fujita et al. and Nolte et al.11
have reported that the 4C1 form can be stabilized in an inclusion
state with the proper guest even if the conformational change
4
1. Introduction
Aquaporins (AQPs) are nanosize pore channels which express
various functions necessary for the maintenance of living
systems. AQPs form tetramers as a channel. The size of the
pore in the channel is precisely controlled by their peptide
sequence and directly affects the molecular transport in the cell
membrane, with small pore sizes only allowing small molecules
such as water to pass through the pore.1,2
1
4
from C4 to C1 occurs with a high energy barrier. Herein, the
formation mechanisms of pseudo[1]rotaxane are investigated
by using cis-cinnamamide-altro-R-CDs (1-Rs) with an ada-
mantyl (Ad), methyl (Me), or trinitrobenzene (TNB) group on
the end position of the arms (Scheme 1). The end groups (R)
vary in volume with the following order: Ad > TNB . Me.12
The Ad and TNB groups are often employed as stoppers for
R-CD-rotaxane due to being too sterically bulky for the R-CD
cavity.13,14 TNB is the planar structure, but Ad and Me are the
spherical structures. The self-assembly and kinetic properties
of these compounds are discussed when focusing on the
structural properties of the R groups and flexibility of altro-
R-CD. Upon examination of the induced fit (from 1C4 to 4C1)
of the altro-R-CD cavity to the arm moiety and comparison
with the sizes or shapes of the R groups, both parameters
Cyclodextrins (CDs), which have a nanosize pore, are well-
known to form inclusion complexes with some organic mol-
ecules. Significant efforts have been devoted to the investigation
of the structures and inclusion properties of CDs.3-5 When the
size of the cavity and structural changes of the host molecules
are controlled by specific stimuli, it opens the way to novel
applications to new supramolecular systems6-8 and to AQPs
mimic systems where controlling the size of the cavity filters
the transported molecules. In an aqueous solution, 3-mono-altro-
R-CD has two equilibrium states (1C4 and 4C1) due to the altrose
1
moiety where the C4 form is thermodynamically more stable
(9) Fujita, K.; Ohta, K.; Ikegami, Y.; Shimada, H.; Tahara, T.; Nogami,
Y.; Koga, T.; Saito, K.; Nakajima, T. Tetrahedron. Lett. 1994, 35,
9577–9580.
(1) (a) Agre, P.; Preston, G. M.; Smith, B. L.; Jung, J. S.; Raina, S.; Moon,
C.; Guggino, W. B.; Nielsen, S. Am. J. Physiol. Renal. Physiol. 1993,
265, 463–476. (b) Agre, P. Angew. Chem., Int. Ed. 2004, 43, 4278–
4290.
(10) (a) Fujita, K.; Chen, W.-H.; Yuan, D.-Q.; Nogami, Y.; Koga, T.;
Fujioka, T.; Mihashi, K.; Immel, S.; Lichtenthaler, F. W. Tetrahedron:
Asymmetry 1999, 10, 1689–1696. (b) Chen, W.-H.; Fukudome, M.;
Yuan, D.-Q.; Fujioka, T.; Mihashi, K.; Fujita, K. Chem. Commun.
2000, 7, 541–542.
(2) Gonen, T.; Walz, T. Q. ReV. Biol. 2006, 39, 361–396.
(3) (a) Szejtli, J. Chem. ReV. 1998, 98, 1743–1753. (b) Saenger, W.; Katrin
Gessler, J. J.; Steiner, T.; Hoffman, D.; Sanbe, H.; Koizumi, K.; Smith,
S. M.; Takaha, T. Chem. ReV. 1998, 98, 1787–1802.
(11) (a) Venema, F.; Nelissen, H. F. M.; Berthault, P.; Birlirakis, N.; Rowan,
A. E.; Feiters, M. C.; Nolte, R. J. M. Chem. Eur. J. 1998, 4, 2237–
2250. (b) Birlirakis, N.; Henry, B.; Berthault, P.; Venema, F.; Nolte,
R. J. M. Tetrahedron. 1998, 54, 3513–3522.
(4) (a) Connors, K. A. Chem. ReV. 1997, 97, 1325–1357. (b) Rekharsky,
M. V.; Inoue, Y. Chem. ReV. 1998, 98, 1875–1917.
(5) (a) Lipkowitz, K. B. Chem. ReV. 1998, 98, 1829–1873. (b) Mu¨ller,
A.; Wenz, G. Chem. Eur. J. 2007, 13, 2218–2223.
(12) Molecular volume and shape of R’s were calculated by Spartan’06
software. .
(6) Shinkai, S.; Inuzuka, K.; Miyazaki, O.; Manabe, O. J. Org. Chem.
1984, 49, 3440–3442.
(13) (a) Buston, J. E. H.; Young, J. R.; Anderson, H. L. Chem. Commun.
2000, 11, 905–906. (b) Tamura, M.; Ueno, A. Bull. Chem. Soc. Jpn.
2000, 73, 147–154.
(7) (a) Hirose, K.; Ishibashi, K.; Shiba, Y.; Doi, Y.; Tobe, Y. Chem. Lett.
2007, 36, 810–811. (b) Hirose, K.; Shiba, Y.; Ishibashi, K.; Doi, Y.;
Tobe, Y. Chem. Eur. J. 2008, 14, 3427–3433.
(14) (a) Harada, A.; Li, J.; Kamachi, M. Chem. Commun. 1997, 15, 1413–
1414. (b) Easton, J. C.; Lincoln, S. F.; Meyer, A. G.; Onagi, H.
J. Chem. Soc., Perkin Trans. 1999, 1, 2501–2506.
(8) Kikuzawa, A.; Kida, T.; Akashi, M. Org. Lett. 2007, 9, 3909–3912.
9
17062 J. AM. CHEM. SOC. 2008, 130, 17062–17069
10.1021/ja806620z CCC: $40.75
2008 American Chemical Society

pseudo[1]Rotaxane from a Flexible Cyclodextrin Derivative
A R T I C L E S
Scheme 1. Proposed pseudo[1]Rotaxane Structures and Chemical
Structures of 1-Rs
R. These observations suggest a self-inclusion structure ꢀ (Figure
2, right) where the aromatic moiety is included in the altro-R-
CD cavity but the adamantyl moiety is outside of the narrower
rim. Surprisingly, these observations indicate that an Ad group
thread through the altro-R-CD cavity. In order to understand
the detailed inclusion fashion of the aromatic moiety, resonances
of altro-R-CD were assigned (Figure 3). Drastic upfield and
downfield shifts of resonances C3(H) and C5(H) were observed,
indicating a strong shielding and deshielding effect produced
by the strict inclusion of the aromatic moiety facing its π
electron cloud to B and E glucopyranose units (Figure 3, right).
Furthermore, the coupling constants of altrose C1(H) in R and
ꢀ differed (Figure 1, inset). The J12 coupling constant assigned
to R (1RA) was J12 ) 6.0 Hz, whereas the constant assigned to
A
ꢀ (1ꢀ ) was J12 ) 1.5 Hz, indicating the conformational
conversion of the altrose from 1C4 to 4C1 through the inclusion
process in D2O.10,11 The computational investigations suggest
that this conformational conversion induces significant distortion
of the altro-R-CD cavity (see Supporting Information), leading
to an induced fit to the arm moiety.
2.2. Size Relations between Ad group and altro-r-CD
Cavity. A weaker hydrogen-bonding network at secondary
hydroxyl side of altro-R-CD caused by the introduction of
altrose may allow tumbling of the altrose unit of 1-Ad, leading
to the same pseudo[1]rotaxane structure described above.16
Therefore, to confirm the threading of an Ad group through the
altro-R-CD cavity, the size of an Ad group relative to that of
the altro-R-CD cavity was experimentally investigated using
an axis/wheel system. The axis molecule, both ends of which
were capped by the Ad groups, was prepared (Figure 4a).
Pyridinium ions were used to improve the solubility in water,
and the decamethylene moiety was chosen as a recognition site
for R-CD and 3-NH2-altro-R-CD. The axis molecule was mixed
with an excess amount of R-CD or 3-NH2-altro-R-CD in D2O.
After reaching the equilibrium of the complexation, the 1H NMR
spectrum of the mixture with 3-NH2-altro-R-CD showed both
splitting resonances of the axis protons and upfield and
downfield shifts of the Ad resonances, whereas the spectrum
of the mixture with R-CD only displayed resonance shifts of
the Ad protons. This result indicates that both wheel molecules
interact with the axis molecule. ROESY measurements were
performed to determine the supramolecular structures. The
mixture of 3-NH2-altro-R-CD and an axis molecule showed
were found to affect the kinetic property of the formation of
pseudo[1]rotaxanes.
2. Results and Discussion
2.1. Assembly Structure of 1-Ad. The 1H NMR spectrum of
1-Ad in D2O differed from that in DMSO-d6 (Figure 1). In D2O,
two equilibrating species (R and ꢀ) were observed in both the
aromatic and altro-R-CD regions. On the other hand, a single
species was observed in DMSO-d6. These results indicate that
1-Ad forms a hydrophobicity-driven self-assembly with a
complexation rate slower than the NMR time scale at 30 °C in
D2O. The hydrodynamic radii (Rh)15 of R and ꢀ, which were
estimated in D2O using pulsed field gradient spin-echo NMR
measurements, were 0.86 (R) and 0.76 (ꢀ) nm, suggesting an
intramolecular complexation. The NOESY spectrum in D2O
showed correlations between the inner protons of altro-R-CD
and the protons of the arm moiety assigned to ꢀ (Figure 2). A
strong correlation between aromatic moiety (protons cꢀ and dꢀ)
and C5(H), which is located in the narrower rim, was observed,
as well as between olefin resonance (bꢀ) and C3(H), which is
located in the wider rim. In contrast, a correlation could not be
discerned between altro-R-CD and the arm moiety assigned to
Figure 1. 1H NMR spectra of 1-Ad at 1 mM in DMSO-d6 (upper) and D2O (lower) at 30 °C. Red and blue colors identify resonances assigned to R and
ꢀ, respectively. Sugar units are labeled clockwise from A to F (viewed from the wider rim), beginning with the altrose unit.
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 17063

A R T I C L E S
Miyawaki et al.
Figure 2. Partial NOESY NMR spectrum of 1-Ad at 0.5 mM in D2O (τm ) 600 ms).
Figure 3. Assignment of the altro-R-CD resonances of 1-Ad based on COSY, TOCSY, ROESY, and HMQC measurements (left) in D2O, and proposed
inclusion fashion of the aromatic moiety (right). Sugar units are labeled clockwise from A to F (viewed from the wider rim), beginning with altrose.
correlations between the inner protons of 3-NH2-altro-R-CD and
the protons of both the Ad group and methylene linker (Figure
4b). On the other hand, only a correlation between the inner
protons of R-CD and the protons of Ad groups was observed
in the mixture with R-CD (Figure 4c). These observations show
that 3-NH2-altro-R-CD included both the Ad group and meth-
ylene linker moiety, whereas R-CD included only Ad moieties.
Conversely, the Ad group can thread through the 3-NH2-altro-
R-CD cavity but cannot thread through the R-CD cavity. In
conclusion, altro-R-CD has a cavity large enough to allow
threading of the Ad group due to the introduction of the altrose
moiety.
2.3. Assembly Structures of 1-Me and 1-TNB. The self-
assembly properties of 1-Me and 1-TNB were also investigated
1
in D2O. The arrangement of the H NMR resonances of 1-Me
was similar to that of 1-Ad because two equilibrating species
(R and ꢀ) with different J12 values (J12(R) ) 6.0 and J12(ꢀ) 1.5
Hz) were observed (Figure 5b). The spectral property of 1-Me,
such as Rh values (Rh(R) ) 7.8 nm and Rh(ꢀ) ) 7.3 nm),
anisotropic effects to C3(H) and C5(H) protons, and ROESY
analysis, are similar to those of 1-Ad, suggesting the intramo-
lecular complexation and strict inclusion of the aromatic moiety.
On the other hand, 1-TNB showed a single-species resonance
(15) The Rh values were calculated by using the Stokes-Einstein equation.
(16) (a) Chen, Z.; Bradshaw, S, J.; Yi, G.; Pyo, D.; Black, R, D.;
Zimmerman, S. S.; Lee, L. M. J. Org. Chem. 1996, 61, 8949–8955.
(b) Yamada, T.; Fukuhara, G.; Kaneda, T. Chem. Lett. 2003, 32, 534–
535. (c) Nishiyabu, R.; Kano, K. Eur. J. Org. Chem. 2004, 24, 4985–
4988.
1
with a J12 value of 6.3 Hz, indicative of the C4 form (Figure
9
17064 J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008

pseudo[1]Rotaxane from a Flexible Cyclodextrin Derivative
A R T I C L E S
Figure 4. Chemical structure of the axis molecule (a). Partial ROESY spectrum of the mixture of axis molecule (1.25 mM) and 3-NH2-altro-R-CD (10 mM)
at τm ) 200 ms in D2O (b), and the mixture of axis molecule (1.25 mM) and R-CD (10 mM) at τm ) 200 ms in D2O (c), respectively.
Figure 5. 1H NMR spectra of (a) 1-TNB, (b) 1-Me, and (c) 1-Ad in D2O at 30 °C.
5a). No correlation could be discerned between altro-R-CD and
the arm moiety in any mixing time (τm ) 150, 200, and 600
ms) in ROESY analysis, suggesting the noninclusion form. On
the basis of these results, it was concluded that 1-Me forms a
self-inclusion complex similar to that formed by 1-Ad, whereas
1-TNB cannot form such a complex. Hence, we hypothesize
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 17065

A R T I C L E S
Miyawaki et al.
1
Figure 6. Time course H NMR spectra of 1-Ad (5 °C, 1 mM, D2O) 210 s (a), 1590 s (b), 3279 s (c), and saturated state (d).
Figure 7. Time course UV-vis spectra of 1-Ad (left) and its absorbance decay at 280 nm (right) in pH ) 7 (a) and in pH > 12 (b).
that the end group plays an important role for self-complexation
and that the planarity of the TNB group is not suitable for
threading through the altro-R-CD cavity.
to ꢀ increased with time, whereas that of the resonances assigned
to R decreased, indicating the interconversion between nonin-
clusion (R) and inclusion (ꢀ). This result suggests that the initial
structure was mainly in its noninclusion form and that the ratio
of the self-inclusion form gradually increased as a function of
time (s) in aqueous solutions.
To determine the rate constant (k) of such a slow intercon-
version, time course UV-vis measurements for 1-Ad were
performed in water. A UV-vis experiment is suitable for the
quantitative investigation of its kinetics at room temperature.
The absorbance of 1-Ad was measured as soon as the compound
prepared by precipitation from methanol to acetone was
dissolved in water until equilibrium was reached (Figure 7a,
left). Figure 7a, right shows the absorbance decay at 280 nm
2.4. Kinetic Investigation. The formation mechanism of these
assemblies was kinetically investigated. 1-Ad and 1-Me formed
self-inclusion complexes by way of the threading of the arm
moiety with a conformational conversion of the altrose. Neither
a coalescence tendency of resonances nor chemical exchange
signals were observed between the R and ꢀ resonances of 1-Ad
in the NMR analysis, indicating that 1-Ad did not interconvert
on the NMR time scale (10-2 < k < 102 s-1) at 30 °C. The
initial and final structures of 1-Ad in this interconversion were
1
confirmed by the time course H NMR experiment at low
temperature (Figure 6). The integral of the resonances assigned
9
17066 J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008

pseudo[1]Rotaxane from a Flexible Cyclodextrin Derivative
A R T I C L E S
3. Conclusions
1-Rs having the cinnamamide moiety as a guest and bulky
end groups as a rate-determining moiety of the threading process
have been prepared, and their supramolecular structures and
kinetic properties have been quantitatively examined. Kinetics
of the self-inclusion process was found to be controlled by the
size and shape of the R groups. These results suggest that the
self-inclusion process does not derive from tumbling of the
altrose unit bearing the arm moiety, but by threading of the
arm moiety through the CD cavity.
It is concluded that 1-Ad forms a pseudo[1]rotaxane with a
q
higher ∆Gon,off than 1-Me. The Ad group is an important
Figure 8. Results of EXSY measurement for 1-Me. Plots of ln[(r + 1)/(r
- 1)] as a function of mixing time (τm ) 100, 200, 300, 400, 500, 700, and
900 ms).
component for this assembly process, and modification of
pyranose from glucopyranose to altropyranose enables the Ad
group to thread through the altro-R-CD cavity. We speculate
that the flexibility of altro-R-CD results in an induced fit, which
enhances the stability of this pseudo[1]rotaxane with the Ad
group, and leads to this unexpected result.
and its saturation after 600 s. The time course UV-vis
experiment for 1-Rs also reveals that no other processes
(hydration, local heat effect of dissolving sample) produce
spectral change in the experimental time scale. Therefore, we
concluded that this absorbance decay indicates the environmental
variations around the aromatic moiety such as hydration and
electron density change related to the inclusion of aromatic
moiety into the altro-R-CD cavity. The k values for the
interconversion between noninclusion and self-inclusion of 1-Ad
were determined to be 5.9 × 10-3 s-1 using a first-order
opposing equation, which corresponds to a free energy of
activation (∆G on,off‡) of 87.2 kJ mol-1 at 30 °C. This k value is
reasonable because this rate constant is out of the range of EXSY
experiments. On the other hand, 1-Me did not show such UV
decay because of a faster interconversion (see Supporting
Information). Therefore, rate constants for the interconversion
of 1-Me were estimated by using chemical exchange signals
between the R and the ꢀ resonances in EXSY measurements
(Figure 8).
Table 1 lists the determined k values from UV-vis and EXSY
analysis. It should be noted that the k values of 1-Ad are
100-200 times smaller than those of 1-Me. These differences
in the free energy of activation (∆[∆Gon,offq ] ) 12 and 13.4 kJ
mol-1) imply an energy barrier for the puzzle ring structure of
1-Ad derived from the relation between the size of an Ad group
and the flexibility of the altro-R-CD cavity. In order to further
understand the flexibility contribution of altro-R-CD in the
assembly process of 1-Ad, the k values were determined under
deprotonated conditions of the hydroxyl groups because CDs
have rigid structures aided by intramolecular hydrogen bonding
between the hydroxyl groups3 (Figure 7b). The ∆Gonq value in
basic conditions is listed in Table 1, and shows a decrement of
free energy of activation (∆[∆Gonq] ) 1.6 kJ mol-1). The free
energy (-∆Go) estimated by van’t Hoff plots also shows
thermodynamic stabilization of pseudo[1]rotaxane under depro-
tonated conditions. These results suggest that the flexibility of
the cyclodextrin framework is advantageous to form
pseudo[1]rotaxane from both kinetic and thermodynamic points
of view, indicating the effect of intramolecular hydrogen
bonding on the flexibility of altro-R-CD.
4. Experimental Section
4.1. Materials. R-Cyclodextrin (R-CD) was obtained from Junsei
Chemical Co., Ltd. trans-p-Aminocinnamic acid, 4-aminopyridine,
and 1,10-diiododecane were obtained from Tokyo Kasei Kogyo,
Co., Ltd. N,N′-Dicyclohexylcarbodiimide (DCC), 1-hydroxyben-
zotriazole (1-HOBt), acetic anhydride, and 2,4,6-trinitrobenzene-
sulfonic acid sodium salt were obtained from Nacalai Tesque Inc.
Dimethyl sulfoxide-d6 (DMSO-d6), and D2O used as solvents for
NMR measurements were obtained from Merck Ltd. 3A-Amino-
3A-deoxy-(2AS,3AS)-R-CD (3-NH2-altro-R-CD) was prepared
according to the method reported previously.17
4.2. Synthesis. 4.2.1. Preparation of 1-Ad and 1-Me. 1-Rs
(1-Ad and 1-Me) were prepared by photo irradiation of the trans-
cinnamamide-altro-R-CD derivatives (trans1-Ad and trans1-Me)
using xenon light source and optical filters (center wavelength: 300
nm, half-bandwidth: 10 nm) with further purification by HPLC.
4.2.1.1. 1-Ad. trans1-Ad was prepared according to the method
reported previously18 and was dissolved in methanol (100 mg/50
mL). After the photo irradiation (λ ) 300 nm), the solution was
evaporated. The precipitate was purified by HPLC to give 1-Ad in
32% yield.
1H NMR (DMSO-d6, 600 MHz): δ 9.16 (s, 1H, NH), 8.03 (d, J
) 9.0 Hz, 1H, NH), 7.73 (d, J ) 8.7 Hz, 2H of Ph), 7.60 (d, J )
8.7 Hz, 2H of Ph), 6.54 (d, J ) 13.0 Hz, 1H, PhCHd), 5.86 (d, J
) 13.0 Hz, 1H, dCHCO), 5.64-5.16 (m, 11H, O(2)H, and O(3)H
of R-CD), 4.89 (t, J ) 5.2 Hz, 2H, C(6′)H of R-CD), 4.85-4.63
(m, 5H, C(1)H of R-CD), 4.64 (d, J ) 6.7 Hz, 1H C(1′)H of R-CD),
4.56-4.44 (m, 10H, O(6)H of R-CD), 4.16 (m, 1H, C(3′)H of
R-CD), 3.94 (m, 1H, C(5′)H of R-CD), 3.78-3.55 (m, 12H, C(4′)H,
C(3)H, C(5)H and C(2′)H of R-CD), 3.47-3.24 (m, 10H, C(4)H
and C(2)H of R-CD), 2.01(s, 3H, -CH-), 1.90(s, 6H,
-CH-CH2-C-), 1.69 (s, 6H, -CH-CH2-CH-). Positive ion
MALDI-TOF Mass m/z ) 1303 [M + Na]+. Elemental Anal. Calcd
for C56H82N2O31 ·7.2H2O: C, 47.74; H, 6.90; N, 1.99. Found: C,
47.38; H, 6.53; N, 2.00.
4.2.1.2. 1-Me.
4.2.1.2.1. p-Methylamidecinnamic
acid
(MeAmCiOH). To a solution of trans-p-aminocinnamic acid (1.0
g, 6.1 mmol) in methanol was added acetic anhydride (2 mL). After
stirring the methanol solution at room temperature for 3 h, the
Table 1. Kinetic Parameters of 1-Ad and 1-Me.
∆Gon [kJ mol-1
]
koff [s-1
]
∆Goff [kJ mol-1
]
K
∆Go [kJ mol-1
]
q
q
1-Rs
pH
method
τ-1 [s-1
]
kon [s-1
]
1-Ad
1-Ad
1-Me
7
UV
UV
NMR
1.17 × 10-2
1.34 × 10-2
1.89
5.9 × 10-3
1.1 × 10-2
1.2
87.2
85.6
73.8
5.9 × 10-3
2.4 × 10-3
6.8 × 10-1
87.2
89.4
75.2
1.0
4.5
1.8
-0.1
3.8
1.4
>12
7
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 17067

A R T I C L E S
Miyawaki et al.
solution became turbid gradually. The precipitate was collected by
filtration and washed with NaOH aqueous solution (pH ) 11) to
give MeAmCiOH in 79% yield.
acid. The precipitate was dissolved in ethyl acetate and washed
with hydrochloric acid. The ethyl acetate phase was evaporated to
give cis-transTNBAmCiOH in 63% yield.
1H NMR (DMSO-d6, 500 MHz): δ 12.20 (s, 1H COOH), 10.09
(s, 1H, NH), 7.60 (s, 4H of Ph), 7.50 (d, J ) 16.0 Hz, 1H
Ph-CH)), 6.39 (d, J ) 16.0 Hz, 1H, )CHCO), 2.05 (s, 3H,
-COCH3).
1H NMR (DMSO-d6, 500 MHz): δ 12.28 (s, 1H COOH), 10.21
(s, 1H transNH), 10.19 (s, 1H cisNH), 8.96 (s, 2H of transPh),
8.95 (s, 2H of cisPh), 7.63 (d, J ) 8.6 Hz, 4H of cis- and transPh),
7.53 (d, J ) 16.0 Hz, 1H transPh-CHd), 7.16 (d, J ) 8.6 Hz, 2H
of transPh), 7.12 (d, J ) 8.6 Hz, 2H of cisPh), 6.83 (d, J ) 12.8
Hz, 1H cisPh-CHd), 6.46 (d, J ) 16.0 Hz, 1H trans ) CHCO),
5.91 (d, J ) 12.8 Hz, 1H cis ) CHCO).
4.2.1.2.2. trans1-Me. To a solution of 3-NH2-altro-R-CD (200
mg, 0.21 mmol) in 20 mL DMF was added MeAmCiOH (51 mg,
0.25 mmol). After the solution was cooled down below 0 °C, DCC
(51 mg, 0.25 mmol) and 1-HOBt (33 mg, 0.25 mmol) were added.
The resulting mixture was stirred at room temperature for 5 days.
After insoluble materials were removed by filtration, the filtrate
was poured into acetone (250 mL). The precipitate was purified
by preparative reversed phase chromatography (elution: water-
acetonitrile) to give trans1-Me in 46% yield.
4.2.2.3. cis1-TNB. To a solution of 3-NH2-altro-R-CD (0.57 g,
0.6 mmol) in 50 mL DMF was added cis-transTNBAmCiOH (0.22
g, 0.6 mmol). After the solution was cooled down below 0 °C,
DCC (0.15 mg, 0.72 mmol) and 1-HOBt (95 mg, 0.72 mmol) were
added. The resulting mixture was stirred at room temperature for
5 days. After insoluble materials were removed by filtration, the
filtrate was poured into acetone (500 mL). The precipitate was
purified by preparative reversed phase chromatography (elution:
water-acetonitrile) to give cis1-TNB in 13% yield.
1H NMR (DMSO-d6, 500 MHz): δ 10.05 (s, 1H, -NH), 8.10 (d,
J ) 9.3 Hz, 1H, -NH), 7.71 (d, J ) 8.6 Hz, 2-H of Ph), 7.55 (d,
J ) 8.6 Hz, 2-H of Ph), 7.34 (d, J ) 15.6 Hz, 1H, Ph-CHd),
6.44 (d, J ) 15.6 Hz, 1H, dCH-CO), 5.94-5.15 (m, 11H, O(2)H,
and O(3)H of R-CD), 4.87 (t, J ) 5.2 Hz, 2H, C(6′)H of R-CD),
4.85-4.77 (m, 5H, C(1)H of R-CD), 4.65 (d, J ) 6.6 Hz, 1H,
C(1′)H of R-CD), 4.53-4.41 (m, 10H, O(6)H of R-CD), 4.16 (m,
1H, C(3′)H of R-CD), 3.88 (m, 1H, C(5′)H of R-CD), 3.81-3.56
(m, 12H, C(4′)H, C(3)H, C(5)H and C(2′)H of R-CD), 3.45-3.20
(m, overlaps with HOD), 2.05 (s, 3H, -CH3). Positive ion MALDI-
TOF Mass m/z ) 1154 [M + Na]+.
1H NMR (DMSO-d6, 500 MHz): δ 10.19 (s, 1H, NH), 8.94 (s,
2H of Ph), 8.06 (d, J ) 8.7 Hz, 1H, NH), 7.77 (d, J ) 8.5 Hz, 2H
of Ph), 7.06 (s, 2H of Ph), 6.58 (d, J ) 12.7 Hz, 1H, PhCHd),
5.92 (d, J ) 12.7 Hz, 1H, dCHCO), 5.84-5.17 (m, 11H, O(2)H,
and O(3)H of R-CD), 4.87 (t, J ) 5.3 Hz, 2H, C(6′)H of R-CD),
4.85-4.71 (m, 5H, C(1)H of R-CD), 4.64 (d, J ) 6.8 Hz, 1H C(1′)H
of R-CD), 4.55-4.43 (m, 10H, O(6)H of R-CD), 4.16 (m, 1H,
C(3′)H of R-CD), 3.94 (m, 1H, C(5′)H of R-CD), 3.75-3.55 (m,
12H, C(4′)H, C(3)H, C(5)H and C(2′)H of R-CD), 3.46-3.39 (m,
10H, C(4)H and C(2)H of R-CD). Positive ion MALDI-TOF Mass
4.2.1.2.3. cis1-Me. trans1-Me was dissolved in methanol (0.11
g/30 mL). After the photo irradiation (λ ) 300 nm), the solution
was evaporated. The precipitate was purified by HPLC to give 1-Me
in 61% yield.
m/z
) 1322 [M +
Na]+. Elemental Anal. Calcd for
C51H69N5O36 ·6H2O: C, 42.65; H, 5.68; N, 4.88. Found: C, 42.25;
H, 5.49; N, 4.92.
1H NMR (DMSO-d6, 500 MHz): δ 9.98 (s, 1H, NH), 8.00 (d, J
) 9.3 Hz, 1H, NH), 7.72 (d, J ) 8.9 Hz, 2H of Ph), 7.50 (d, J )
8.9 Hz, 2H of Ph), 6.55 (d, J ) 13.0 Hz, 1H, PhCHd), 5.87 (d, J
) 13.0 Hz, 1H, dCHCO), 5.84-5.14 (m, 11H, O(2)H, and O(3)H
of R-CD), 4.87 (t, J ) 5.3 Hz, 2H, C(6′)H of R-CD), 4.85-4.66
(m, 5H, C(1)H of R-CD), 4.64 (d, J ) 6.7 Hz, 1H C(1′)H of R-CD),
4.54-4.41 (m, 10H, O(6)H of R-CD), 4.16 (m, 1H, C(3′)H of
R-CD), 3.93 (m, 1H, C(5′)H of R-CD), 3.78-3.55 (m, 12H, C(4′)H,
C(3)H, C(5)H and C(2′)H of R-CD), 3.48-3.35 (m, 10H, C(4)H
and C(2)H of R-CD), 2.03 (s, 3H, -CH3). Positive ion MALDI-
TOF Mass m/z ) 1154 [M + Na]+. Elemental Anal. Calcd for
C47H70N2O31 ·7H2O: C, 43.93; H, 6.59; N, 2.18. Found: C, 44.06;
H, 6.31; N, 2.18.
4.2.3. Preparation
of
the
Axis
Molecule.
4.2.3.1. 4-Adamantylamidepyridine (AdPy). To a solution of
4-aminopyridine (1.0 g, 10.6 mmol) and triethylamine (1.0 g, 9.9
mmol) in 80 mL of THF was added 1-adamantanecarbonyl chloride
(1.7 g, 8.5 mmol) in 15 mL of THF solution under ice cooling.
The resulting mixture was stirred at room temperature overnight.
After the precipitate was removed, the solution phase was concen-
trated and dissolved in ethylacetate. The solution was washed with
hydrochloric acid (pH ) 3) and was evaporated to give AdPy in
62% yield.
1H NMR (DMSO-d6, 500 MHz): δ 9.44 (s, 1H, NH), 8.38 (d, J
) 6.3 Hz, 2H of Py), 7.68 (d, J ) 6.3 Hz, 2H of Py), 2.01 (s, 3H,
-CH-), 1.90 (s, 6H, -CH2-), 1.70 (s, 6H, -CH2-).
4.2.2. Preparation of 1-TNB. 4.2.2.1. cis-trans Mixture of
p-Aminocinnamic Acid (cis-trans-AmCiOH). trans-p-Amino-
cinnamic acid was dissolved in methanol (205 mg/25 mL). After
the photo irradiation (λ ) 300 nm), the solution was evaporated.
4.2.3.2. Bis(4-adamantylamidepyridine)decane ([AdPy-C10-
PyAd]2+ ·2Cl-). The acetonitrile solution of AdPy (0.80 g, 3.1
mmol) and 1,10-diododecane (0.74 g, 1.9 mmol) was stirred under
reflux for 3 days. The precipitate was collected and washed with
acetonitrile to give [AdPy-C10-PyAd]2+ ·2I- in 31% yield. A
saturated aqueous solution of [AdPy-C10-PyAd]2+ ·2I- was poured
into an aqueous solution of hexafluorophosphate ammonium, and
the solution was stirred for 2 days at room temperature. The
precipitate was collected by centrifugal separation and washed with
water to give [AdPy-C10-PyAd]2+ ·2PF6- in 90% yield. An acetone
solution of [AdPy-C10-PyAd]2+ ·2PF6-was poured into an acetone
solution of tetraethylammonium chloride, and the solution was
stirred for 2 days at room temperature. The precipitate was collected
by centrifugal separation and washed with acetone to give [AdPy-
C10-PyAd]2+ ·2Cl- in 89% yield.
1
The ratio of the isomerization was determined by using H NMR
measurement (cis/trans ) 0.85).
1H NMR (DMSO-d6, 500 MHz): δ 11.80 (s, 2H cis- and
transCOOH), 7.59 (d, J ) 8.6 Hz, 2H of cisPh), 7.40 (d, J ) 15.8
Hz, 1H transPh-CHd), 7.32 (d, J ) 8.6 Hz, 2H, transPh), 6.63
(d, J ) 13.0 Hz, 1H cisPh-CH)), 6.54 (d, J ) 8.6 Hz, 2H,
transPh), 6.50 (d, J ) 8.6 Hz, 2H of cisPh), 6.11 (d, J ) 15.8 Hz,
1H trans ) CHCO), 5.67 (s, 2H transNH2), 5.57 (s, 2H cisNH2),
(s, 3H, -COCH3), 5.52 (d, J ) 13.0 Hz, 1H cis ) CHCO).
4.2.2.2. cis-trans Mixture of p-Trinitrophenylaminocinnamic
Acid (cis-trans-TNBAmCiOH). To
a solution of cis-
transAmCiOH (0.15 g, 0.94 mmol) in 25 mL of NaOH aqueous
solution was added trinitrobenzene sulfonic acid sodium salt (0.40
g, 1.1 mmol). After the solution was stirred at room temperature
for 5 h, the solution pH was adjusted to 2-3 by using hydrochloric
1H NMR (D2O, 500 MHz): δ 8.50 (d, J ) 7.5 Hz, 4H of Py),
8.09 (d, J ) 7.5 Hz, 4H of Py), 4.65 (t, J ) 7.1 Hz, 4H, R-methylene
in decamethylene), 2.00 (s, 6H, methyne in Ad), 1.87 (s, 12H,
methylene in Ad), 1.85 (m, 4H, ꢀ methylene in decamethylene),
1.68 (q, J ) 11.6 Hz and 22.9 Hz, 12H, methylene in Ad),
1.21-1.14 (m, methylene in decamethylene). Elemental Anal. Calcd
for C42H60N4O2Cl2 ·2.8H2O: C, 65.15; H, 8.54; N, 7.24. Found: C,
65.19; H, 8.26; N, 7.15.
(17) (a) Takahashi, K.; Hattori, K.; Toda, F. Tetrahedron Lett. 1984, 25,
3331–3334. (b) Ikeda, H.; Nagano, Y.; Du, Y.; Ikeda, T.; Toda, F.
Tetrahedron Lett. 1990, 31, 5045–5048.
(18) Miyawaki, A.; Miyauchi, M.; Takashima, Y.; Yamaguchi, H.; Harada,
A. Chem. Commun 2008, 4, 456–458.
9
17068 J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008

pseudo[1]Rotaxane from a Flexible Cyclodextrin Derivative
A R T I C L E S
4.3. Measurements. The NMR spectra were recorded on a JEOL
JNM LA-500 and VARIAN UNITY600 NMR spectrometer. The
preparative reversed phase chromatography was carried out with a
Waters Delta 600 system (column: SunFireTM Prep C18 19 mm
× 150 mm). Positive-ion matrix assisted laser desorption ionization
time-of-flight (MALDI-TOF) mass spectrometry experiments were
performed using a Shimadzu/KRATOS Axima CFR Ver.2.2.3 mass
spectrometer calibrated by R-cyano-4-hydroxycinnamic acid and
insulin. Photo irradiation was performed using a ASAHI SPECTRA
xenon light source MAX-301 with a band-pass filter HQBP300-
UV φ25.
4.3.1. PFGSE NMR Experiments. The PFGSE NMR diffusion
measurements were carried out by using the bipolar pulse pair
stimulated echo (BPPSTE).19 The strength of the pulsed gradients
was increased from 6.36 × 10-1 to 61.7 gauss/cm. The time
separation of pulsed field gradients and their duration were 0.10
and 1.0 × 10-3 s. The sample was not spun, and the airflow was
disconnected. The shape of the gradient pulse was rectangular, and
its strength varied automatically during the course of the experi-
ments. The D values which represent diffusion coefficient were
determined from the slope of the regression line ln (I/Io) versus
G2, according to following equation.
was obtained from the Eyring equation
k ) (kBT ⁄ p) exp(-∆Gq ⁄ RT)
where kB is the Boltzmann constant, T is the absolute temperature,
p is Plank’s constant, ∆Gq is the free energy of activation, and R
is the gas constant.
4.3.3. UV-Vis Experiments. UV-vis measurements were
carried out at room temperature. The absorbance at 280 nm was
measured for 10 min after dissolving the sample in water at 0.1
mM. The noninclusion (R state) sample for these experiments was
prepared by precipitation from methanol to acetone. The k values,
which represent rate constant, were determined by fitting the
absorbance decay in following equations.
kon
AaB
koff
k ) kon + koff
τ ) 1 ⁄ (kon + koff)
Abs ) r(1 - exp(-t ⁄ τ))
where K is self-assembly constant, Abs is the absorbance at 280
nm, t is time, τ is relaxation time, and r is defined by the following
equation
ln(I ⁄ Io) ) -γ2G2δ2(∆-δ ⁄ 3-τ ⁄ 2)D
where (I/Io) is (observed spin echo intensity/intensity without
gradients), G is gradient strength, ∆ is delay between the midpoints
of the gradients, δ is gradient length, and τ is 90°-180° pulse
r ) (kona - koffb) ⁄ (kon + koff)
where a is initial concentration of A and b is initial concentration
of B. In the case of self-complexation for 1-Ad, the K value was
determined to be K ) 1 from the 1H NMR spectrometry measure-
ment; hence
distance. The calibration of the gradients was carried out by a
20
diffusion measurement of H2O (DH2O ) 2.30 × 10-9 m2 s-1
25 °C.
)
at
4.3.2. EXSY Experiments. The EXSY measurements21 were
carried out at 303 K with a phase-sensitive NOESY pulse sequence.
The mixing time was increased from 100 to 900 ms. The k values,
which represent rate constants, were determined by using the
following equations.
τ ) 1 ⁄ (2kon)
4.4. Calculated Structures. Quantum-chemical calculations
were performed by employing the Spartan ’06 package on a 32-bit
multiprocessor computer. Semiempirical (AM1) method geometrical
optimization was carried out for calculating molecular volume of
substituents. The Hartree-Fock (HF) geometrical optimization was
carried out with the 3-21G basis set for determination of 3-NH2-
kon
AaB
koff
4
altro-R-CD (1C4 and C1), and R-CD structures. To account for
k ) kon + koff
solvent effects, calculations were performed in the presence of water
molecules.
k ) (1 ⁄ τm) ln(r + 1 ⁄ r - 1)
Acknowledgment. We thank Dr. A. Hashidzume and Mr. Seiji
Adachi, Osaka University for his helpful comments and NMR
experiments. This work has been partially supported by global COE
program and Grant in-Aid No. A19205014 for Scientific Research.
A.M. and P.K. acknowledge JSPS for financial support (20 • 637
and 19 • P07064).
where τm is the mixing time and r is defined by the following
equation
r ) 4XAXB(IAA + IBB) ⁄ (IAB + IBA) - (XA - XB)2
where XA and XB are the mole fraction of A and B, respectively,
IAB and IBA are the intensities of the cross peaks between the signals
A and B, respectively, and IAA and IBB are the intensities of the
diagonal signals. The free energy of activation of the interconversion
Supporting Information Available: UV-vis, PFGSTE NMR,
1H NMR, ROESY, EXSY, van’t Hoff plot, and calculated
structures.This material is available free of charge via the
Internet at http://pubs.acs.org.
(19) (a) Stejskal, O. E; Tanner, J. E. J. Chem. Phys. 1965, 42, 288–292.
(b) Stilbs, P. Prog. in NMR Spectrosc. 1987, 19, 1–45.
(20) Weingartner, H. Z. Phys. Chem. (Leipzig) 1982, 132, 129–149.
(21) Perrin, C. L.; Dwyer, T. J. Chem. ReV. 1990, 90, 935–967.
JA806620Z
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 17069