Syntheses and self-assembly behaviors of the azobenzenyl modified β-cyclodextrins isomers

Article abstract of DOI:10.1021/jo800488f

(Figure Presented) A couple of analogues of azobenzenyl β-cyclodextrins 1 and 2 with self-locked and self-unlocked conformations have been synthesized via the Huisgen cycloaddition from the same reactants, but in distinct reaction conditions (i.e., the hy

Full text of DOI:10.1021/jo800488f

Syntheses and Self-Assembly Behaviors of the Azobenzenyl Modified
ꢀ-Cyclodextrins Isomers
Yu Liu,* Zi-Xin Yang, and Yong Chen
Department of Chemistry, State Key Laboratory of Elemento-Organic Chemistry, Nankai UniVersity,
Tianjin 300071, People’s Republic of China
yuliu@nankai.edu.cn
ReceiVed March 1, 2008
A couple of analogues of azobenzenyl ꢀ-cyclodextrins 1 and 2 with self-locked and self-unlocked
conformations have been synthesized via the Huisgen cycloaddition from the same reactants, but in distinct
reaction conditions (i.e., the hydrothermal synthesis and the “click” reaction, respectively), their
conformations were sufficiently proved by X-ray crystal structural analysis, molecular modeling study,
and 2D NMR spectroscopy, and their self-assembly behaviors in aqueous solution were also investigated
by NMR spectroscopy. Interestingly, the self-locked conformer 1, which could be regarded as a new
type of [1]rotaxane, self-assembled to a novel bimolecular capsule, where its azobenzene substituent
was included in both its own cavity and the counterpart’s cavity, in aqueous solution and in the solid
state. In contrast, by adjusting the conformation of 1 to a self-unlocked one, the resulting conformer 2
was found to self-assemble to a linear supramolecule. These studies have shown the stronger impact of
the reaction condition changing in cyclodextrin’s modification products and will provide a new access to
control the structure of supramolecular assemblies by tuning the conformation of building blocks.
Introduction
ane or self-included structure have been developed to study the
substituent’s machine-like movement with the CD cavity.4 For
example, Harada et al. introduced the PEG polymer chain with
different lengths or end groups to the CD rim to study their
The fabrication of nanometer-scaled supramolecular archi-
tectures with unique structures and functions is a key aspect in
supramolecular chemistry and nanoscience.1 Possessing the
capability of including various organic molecules within their
hydrophobic cavities, cyclodextrin (CD) has been used as the
basic building block to construct the supramolecular assemblies
to study their structure, function, and molecular motion.2
Superior to those from native CDs, supramolecular architectures
constructed from modified CDs with various covalently linked
substituents always exhibit more excited functions.3 In general,
there are two assembly modes for modified CDs, that is, the
intermolecular inclusion mode and intramolecular inclusion
mode. Recently, utilizing the intramolecular inclusion mode of
modified CDs, some simple molecular machines with [1]rotax-
(1) (a) Tian, H.; Wang, Q.-C. Chem. Soc. ReV. 2006, 35, 361–374. (b) Balzani,
V.; Credi, A.; Silvi, S.; Venturi, M. Chem. Soc. ReV. 2006, 35, 1135–1179. (c)
Saha, S.; Stoddart, J. F. Chem. Soc. ReV. 2007, 36, 77–92. (d) Lankshear, M. D.;
Beer, P. D. Acc. Chem. Res. 2007, 40, 657–668. (e) Green, J. E.; Choi, J. W.;
Boukai, A.; Bunimovich, Y.; Johnston-Halperin, E.; DeIonno, E.; Luo, Y.; Sheriff,
B. A.; Xu, K.; Shin, Y. S.; Tseng, H. R.; Stoddart, J. F.; Heath, J. R. Nature
2007, 445, 414–417. (f) Roithova, J.; Milko, P.; Ricketts, C. L.; Schroder, D.;
Besson, T.; Dekoj, V.; Belohradsky, M. J. Am. Chem. Soc. 2007, 129, 10141–
10148. (g) Miljanic, O. S.; Dichtel, W. R.; Khan, S. I.; Mortezaei, S.; Heath,
J. R.; Stoddart, J. F. J. Am. Chem. Soc. 2007, 129, 8236–8246. (h) Guidry, E. N.;
Li, J.; Stoddart, J. F.; Grubbs, R. H. J. Am. Chem. Soc. 2007, 129, 8944–8945.
(i) Aucagne, V.; Berna, J.; Crowley, J. D.; Goldup, S. M.; Hanni, K. D.; Leigh,
D. A.; Lusby, P. J.; Ronaldson, V. E.; Slawin, A. M. Z.; Viterisi, A.; Walker,
D. B. J. Am. Chem. Soc. 2007, 129, 11950–11963. (j) Frampton, M. J.; Anderson,
H. L. Angew. Chem., Int. Ed. 2007, 46, 1028–1064.
5298 J. Org. Chem. 2008, 73, 5298–5304
10.1021/jo800488f CCC: $40.75 2008 American Chemical Society
Published on Web 06/13/2008

Azobenzenyl Modified ꢀ-Cyclodextrins Isomers
SCHEME 1. Syntheses of 1 and 2
self-threading and self-dethreading dynamics as well as the
thermal or photochemical control of conformational exchange
of substituents.5 Easton et al. constituted a molecular machine
from modified CDs, and the photoisomerization of its substi-
tutent changed the in/out conformation of CD and provided the
on/off switch of the machine.6 On the other hand, the intermo-
lecular inclusion mode of modified CDs was reported to result
in the formation of complicated assemblies, such as a hermaph-
rodite [2]rotaxane with a daisy chain structure7 and supramo-
lecular polymers through host-guest interactions.8 However,
these studies were mainly focused on the assembly behaviors
of modified CDs with either self-included or self-excluded
conformation; the comparative study on the self-assembly of
modified CDs with the same composition but disparate confor-
mations is still rare, to the best of our knowledge. In this work,
we prepared a couple of azobenzenyl-modified CDs with
different topological structures, i.e., a self-locked conformer 1
and its self-unlocked analogue 2. Significantly, the self-locked
1 presents a structural character of CD-based [1]rotaxane, and
this [1]rotaxane self-assembles to a bimolecular capsule in
aqueous solution, while its self-unlocked analogue 2 forms a
linear supramolecule. It is of our particular interest to elucidate
the factors to restrict the conformation of modified CDs and
the influences on the self-assembly behaviors of different CD
conformers. These studies will serve our deep understanding
of the mechanism of supramolecular aggregation and may
provide a new way to control the form of supramolecular
aggregates by tuning the conformation of building blocks.
Results and Discussion
Synthesis. 1 was synthesized by Huisgen 1,3-dipolar cy-
cloaddition in 69% yield with use of the hydrothermal synthesis
(Scheme 1) and easily purified by recrystallization in water,
while its self-unlocked analogue 2 was also prepared in 71%
yield by “click” chemistry, a CuI-catalyzed Huisgen 1,3-dipolar
cycloaddition. A possible mechanism for the formation of 1 may
be that, with the capability of ꢀ-CD to include various organic
molecules in its hydrophobic cavity, 3 first formed an inclusion
complex with the intermediate 4, and then the reaction between
the azido group in 3 and the ethynyl group in 4 gave the product
1. This procedure is similar to a reported cycloaddition
procedure induced by cucurbituril, where a regiospecific 1,4-
disubstituted product was obtained through the encapsulation
of the cationic substrates in the cavity of cucurbituril.9 Owing
to the steric effect of the CD cavity, the synthesis of the self-
locked 1 under hydrothermal condition exhibited good 1,5-
disubstituted regioselectivity, like the increase of regioselectivity
of nitrile oxide cycloaddition in a ꢀ-CD-scaffold.10 However,
it should be noted that the formation of inclusion complex is
an equilibrium process, and the formation of self-unlocked
isomers of 1 is also possible in the hydrothermal synthesis.
(2) (a) Nepogodiev, S. A.; Stoddart, J. F. Chem. ReV. 1998, 98, 1959–1976.
(b) Wenz, G.; Han, B.-H.; Muller, A. Chem. ReV. 2006, 106, 782–817. (c) Harada,
A. Acc. Chem. Res. 2001, 34, 456–464. (d) Cheetham, A. G.; Hutchings, M. G.;
Claridge, T. D. W.; Anderson, H. L. Angew. Chem., Int. Ed. 2006, 45, 1596–
1599. (e) Michels, J. J.; O’Connell, M. J.; Taylor, P. N.; Wilson, J. S.; Cacialli,
F.; Anderson, H. L. Chem. Eur. J. 2003, 9, 6167–6176. (f) Murakami, H.;
Kawabuchi, A.; Matsumoto, R.; Ido, T.; Nakashima, N. J. Am. Chem. Soc. 2005,
127, 15891–15899. (g) Wang, Q.-C.; Qu, D. H.; Ren, J.; Chen, K.-C.; Tian, H,
Angew. Chem., Int. Ed. 2004, 43, 2661–2665. (h) Qu, D.-H.; Wang, Q.-C.; Tian,
H. Angew. Chem., Int. Ed. 2005, 44, 5296–5299. (i) Park, J. S.; Wilson, J. N.;
Hardcastle, K. I.; Bunz, U. H. F.; Srinivasarao, M. J. Am. Chem. Soc. 2006,
128, 7714–7715. (j) Oshikiri, T.; Takashima, Y.; Yamaguchi, H.; Harada, A.
J. Am. Chem. Soc. 2005, 127, 12186–12187. (k) Onagi, H.; Carrozzini, B.;
Cascarano, G. L.; Easton, C. J.; Edwards, A. J.; Lincoln, S. F.; Rae, A. D. Chem.
Eur. J. 2003, 9, 5971–5977.
(3) (a) Nelson, A.; Belitsky, J. M.; Vidal, S.; Joiner, C. S.; Baum, L. G.;
Stoddart, J. F. J. Am. Chem. Soc. 2004, 126, 11914–11922. (b) Ooya, T.; Eguchi,
M.; Yui, N. J. Am. Chem. Soc. 2003, 125, 13016–13017. (c) Liu, Y.; Yu, L.;
Chen, Y.; Zhao, Y.-L.; Yang, H. J. Am. Chem. Soc. 2007, 129, 10656–10657.
(d) Liu, Y.; Wang, H.; Chen, Y.; Ke, C.-F.; Liu, M. J. Am. Chem. Soc. 2005,
127, 657–666. (e) Onagi, H.; Blake, C. J.; Easton, C. J.; Lincoln, S. F. Chem.
Eur. J. 2003, 9, 5978–5988.
(4) (a) Han, Y.; Cheng, K.; Simon, K. A.; Lan, Y.; Sejwal, P.; Luk, Y.-Y.
J. Am. Chem. Soc. 2006, 128, 13913–13920. (b) Deng, W.; Onji, T.; Yamaguchi,
H.; Ikeda, N.; Harada, A. Chem. Commun. 2006, 4212–4214. (c) Kuwabara, T.;
Shiba, K.; Nakajima, H.; Ozawa, M.; Miyajima, N.; Hosoda, M.; Kuramoto,
N.; Suzuki, Y. J. Phys. Chem. A 2006, 110, 13521–13529. (d) Park, J. W.; Lee,
S. Y.; Song, H. J.; Park, K. K. J. Org. Chem. 2005, 70, 9505–9513. (e) Ma, X.;
Qu, D.-H.; Ji, F.-Y.; Wang, Q.-C.; Zhu, L.-L.; Xu, Y.; Tian, H. Chem. Commun.
2007, 1409–1411. (f) Ma, X.; W, Q.-C.; Tian, H. Tetrahedron lett. 2007, 48,
7112–7116.
(8) (a) Deng, W.; Yamaguchi, H.; Takashima, Y.; Harada, A. Angew. Chem.,
Int. Ed. 2007, 46, 5144–5147. (b) Harada, A. J. Polym. Sci., Part A 2006, 44,
5113–5119. (c) Miyauchi, M.; Hoshino, T.; Yamaguchi, H.; Kamitori, S.; Harada,
A. J. Am. Chem. Soc. 2005, 127, 2034–2035. (d) Liu, Y.; Fan, Z.; Zhang, H.-
Y.; Yang, Y.-W.; Ding, F.; Liu, S.-X.; Wu, X.; Wada, T.; Inoue, Y. J. Org.
Chem. 2003, 68, 8345–8352.
(5) (a) Inoue, Y.; Kuad, P.; Okumura, Y.; Takashima, Y.; Yamaguchi, H.;
Harada, A. J. Am. Chem. Soc. 2007, 129, 6396–6397. (b) Inoue, Y.; Miyauchi,
M.; Nakajima, H.; Takashima, Y.; Yamaguchi, H.; Harada, A. J. Am. Chem.
Soc. 2006, 128, 8994–8995. (c) Inoue, Y.; Miyauchi, M.; Nakajima, H.;
Takashima, Y.; Yamaguchi, H.; Harada, A. Macromolecules 2007, 40, 3256–
3262.
(9) (a) Mock, W. L.; Irra, T. A.; Wepsiec, J. P.; Adhya, M. J. Org. Chem.
1989, 54, 5302–5308. (b) Tuncel, D.; Steinke, J. H. G. Chem. Commun. 1999,
1509–1510. (c) Carlqvist, P.; Maseras, F. Chem. Commun. 2007, 748–750.
(10) (a) Barr, L.; Lincoln, S. F.; Easton, C. J. Chem. Eur. J. 2006, 12, 8571–
8580. (b) Meyer, A. G.; Easton, C. J.; Lincoln, S. F.; Simpson, G. W. J. Org.
Chem. 1998, 63, 9069–9075.
(6) Coulston, R. J.; Onagi, H.; Lincoln, S. F.; Easton, C. J. J. Am. Chem.
Soc. 2006, 128, 14750–14751.
(7) Onagi, H.; Easton, C. J.; Lincoln, S. F. Org. Lett. 2001, 3, 1041–1044.
J. Org. Chem. Vol. 73, No. 14, 2008 5299

Liu et al.
FIGURE 1. ROESY spectrum of (a) 1 (1.9 × 10-3 mol dm-3) and (b) 2 (1.9 × 10-3 mol dm-3) in DMSO-d6 at 298 K with a mixing time of 300
ms. Possible structures of 1 (c) and 2 (d) in DMSO.
Therefore, we took several methods to avoid, as much as
possible, the influence of the isomerization. First, we performed
the reaction in water-ethanol solution to maintain the hydro-
phobic interactions between the ꢀ-CD cavity and 4 to some
extent. Then, during the purification, the crude products were
washed successively with acetone and dissolved in hot water
followed by filtration, which would remove the unreacted 4 and
possible isomers of 1, because 4 was soluble in acetone while
the possible isomers were poorly soluble in water (control
experiments have demonstrated that the water solubility of the
self-unlocked 2 obtained from the click reaction was less than
200 µg/mL at room temperature). Finally, because the water
solubility of 1 was lower than that of 3, the unreacted 3 could
be efficiently removed by recrystallization. On the other hand,
2 could be easily purified by sufficiently washing the crude
product with acetone and hot water to remove the unreacted 3
and 4.
Structural Analysis of Isomers. The structural analysis of
modified CD is very important to understand their molecular
assembly behavior. It is well-known that the elucidation of
crystal structure is one of the most convincing methods of
unequivocally illustrating the geometrical conformation of CD
5300 J. Org. Chem. Vol. 73, No. 14, 2008

Azobenzenyl Modified ꢀ-Cyclodextrins Isomers
FIGURE 2. ROESY spectrum of (a) 1 (0.9 × 10-3 mol dm-3) in D2O at 298 K with a mixing time of 300 ms. (b) Possible structure of 1 in water.
derivatives. Unfortunately, our repeated attempts only gave the
single crystal of 1 that was suitable for X-ray crystallography.
Therefore, we tried to compare the conformations of 1 and 2
through a combinational analysis based on 2D NMR spectros-
copy and molecular modeling study. 2D NMR spectroscopy is
an efficient method to study the conformation of CD derivatives
since one can conclude that two protons are closely located in
space (<5 Å) if an NOE cross-peak is detected between the
relevant protons in the NOESY or ROESY spectrum. Therefore,
it is possible to determine the orientation of the 4-(phenylazo)-
phenyloxymethyl[1,2,3]triazolyl substituent related to the ꢀ-CD
cavity by using the assigned NOE correlations, because if the
substituent is self-included in the ꢀ-CD cavity, the NOE
correlations between the protons of the substituent and the
J. Org. Chem. Vol. 73, No. 14, 2008 5301

Liu et al.
between the Ha protons of the substituent and the H3/H5/H6
protons of ꢀ-CD, which were absent in the ROESY spectrum
of 1 in DMSO. These correlations indicated that, besides the II
ring, the III ring of the substituent, which was located outside
the ꢀ-CD cavity in DMSO, was also embedded in the ꢀ-CD
cavity in water. Moreover, Figure 2a also showed a particular
NOE correlation (cross-peak h) between the Ha protons and the
Hc protons of the substituent that could not be observed in the
corresponding ROESY spectrum in DMSO. A simple calculation
by MM2 indicated that the distance between the Ha and the Hc
protons in a single azobenzene moiety was longer than 5 Å.
Therefore, the cross-peak h might not be assigned to an
intramolecular NOE correlation, but to the intermolecular NOE
correlation between the Ha protons in an azobenzene moiety
and the Hc protons in another azobenzene moiety. According
to these NOE correlations, we could deduce that the self-locked
substituent of 1 may also penetrate into the cavity of another
ꢀ-CD to form a tail-to-tail bimolecular capsule, where the tail
referred to the wide opening of the ꢀ-CD cavity, as the possible
structure illustrated in Figures 2b. This structural transformation
in different solvents was visualized in Figure 3.
Significantly, the tail-to-tail dimer of 1 was also unambigu-
ously verified by the X-ray crystal diffraction in the solid state.
The crystal structures were illustrated in Figure 4. As seen in
Figure 4a, the 4-(phenylazo)phenyloxymethyl[1,2,3]triazolyl
substituent threaded through the ꢀ-CD cavity with the II ring
of the substituent entirely embedded in the ꢀ-CD cavity but
the I ring and the III ring located outside the narrow and wide
opening of the ꢀ-CD cavity, respectively. The much longer
distance between the phenyl oxygen and the H-4 atom of the
III ring (ca. 11.4 Å) than the inner diameter of the ꢀ-CD cavity
(6.0-6.5 Å) as well as the rigidity of the substituent linking in
the ꢀ-CD rim jointly prevented the dethreading of the substituent
from the ꢀ-CD cavity. These results were in good agreement
with the deduced self-locked conformation of 1 in aqueous
solution. Figure 4b displayed a clear image of the bimolecular
capsule of 1 in the solid state. It should be noticed that the
bimolecular capsule in aqueous solution is more compact than
that in solid. Further investigations showed that there existed
several types of intramolecular and intermolecular interactions
within the bimolecular capsule. The intramolecular interactions
contained a C-H· · ·π interaction (d ) 2.810 and 2.638 Å for
two units of 1) between the C-6 atom of ꢀ-CD and the I ring of
substituent and six hydrogen bond interactions between the O-2
atom of each glucose unit and the O-3 atom of the adjacent
glucose unit (O · · ·O distance 2.7-3.2 Å), while the intermo-
lecular interactions included a hydrogen bond between the
hydroxyl groups on the wide opening of the counterpart ꢀ-CD
(O· · ·O distance 2.745 Å) and two π-π interactions between
the II (III) ring and the III′ (II′) ring in a face-to-face
arrangement of two substituents (centroid distance: II ring-III′
ring 3.816 Å, III ring-II′ ring 3.922 Å). These interactions
jointly stabilized the capsule-like structure of 1.
FIGURE 3. Structural transformation of 1 by changing solvents.
interior protons of ꢀ-CD cavity should be observed. A prelimi-
nary molecular dynamics simulation study demonstrated that,
because the substituent in 1 or 2 had a relatively rigid structure
and a longer skeleton length (ca.15.6 Å) than the diameter of
the ꢀ-CD cavity (6.0-6.5 Å),11 the conversion between the self-
locked conformer and its self-unlocked analogue was unavailable
even after increasing the temperature of the system once the
substituent was linked to the rigid ꢀ-CD rim. Figure 1a
illustrated the ROESY spectrum of 1 in DMSO, which displayed
the clear NOE correlations between the H3 protons of ꢀ-CD
and the Hb/Hc protons of the azobenzene moiety (cross-peaks
A and C), as well as the NOE correlations between the H5
protons of ꢀ-CD and the Hc protons of the azobenzene moiety
(cross-peak B). These correlations indicated that the II ring of
the substituent was embedded into the ꢀ-CD cavity. Moreover,
no NOE correlation peaks between the interior protons (H3/H5
protons) of ꢀ-CD and the Ha protons of substituent could be
found, which indicated that the III ring of the substituent was
threaded out of the ꢀ-CD cavity. The cross-peaks D and E were
assigned to the intramolecular NOE correlations between the
Hb protons and the Hc/Ha protons of the azobenzene moiety. In
contrast, the ROESY spectrum of 2 showed no appreciable NOE
correlations between interior protons of the ꢀ-CD cavity and
the protons of the substituent, indicating that the substituent
group of 2 was entirely located out of the ꢀ-CD cavity.
According to these NOE signals, we could deduce the possible
structures of 1 and 2 as illustrated in parts c and d of Figure 1.
Generally, the [1]rotaxane is defined as a type of supramo-
lecular system consisting of a rod-like molecule (axle compo-
nent) covalently linking to a cyclic molecule (wheel component),
where the axle molecule threads through the cavity of the wheel
component, and the wheel component is trapped by bulky ends
(stopper) to prevent the unthreading.2b On the other hand, it is
important to note that CDs were reported to barely include the
guest molecule in DMSO due to the solvation and binding of
DMSO molecules by CDs,12 and the exclusion of the self-
included substituent of a CD derivative will inevitably take place
in a DMSO solution if this conformation conversion is available
from the viewpoint of geometry. Therefore, the clear NOE
correlations in the ROESY spectrum of 1 in DMSO, along with
the molecular modeling study, jointly indicated that 1 could be
regarded as a unique [1]rotaxane, although it had no stopper.3e,4e,f
Self-Assembly of 1. Interestingly, 1 gave the more compli-
cated NOE signals in aqueous solution than in DMSO. As shown
in Figure 2a, besides the NOE correlations (cross-peaks a, b, c,
and d) between the Hb/Hc protons of substituent and the H3/
H5/H6 protons of ꢀ-CD, the ROESY spectrum of 1 in D2O
displayed clear NOE correlations (cross-peaks e, f, and g)
It is well-documented that there exist four crystal structures
of monomodified CDs based on the spatial relationship between
the substituent and the CD cavity.13 The first one is the self-
included type, where the substituent is included in its own cavity.
The second is the layer type, where the substituent is located
out of any CD cavity. The third is the mutual-locked dimer type
(11) Szejtli, J. Chem. ReV. 1998, 98, 1743–1754.
(12) (a) Fujita, K.; Ueda, T.; Imoto, T.; Tabushi, I.; Toh, N.; Koga, T. Bioorg.
Chem. 1982, 11, 72–84. (b) Liu, Y.; Zhao, Y.-L.; Zhang, H.-Y.; Fan, Z.; Wen,
G.-D.; Ding, F. J. Phys. Chem. B 2004, 108, 8836–8843. (c) Shikata, T.;
Takahashi, R.; Onji, T.; Satokawa, Y.; Harada, A. J. Phys. Chem. B 2006, 110,
18112–18114.
(13) (a) Harata, K.; Takenaka, Y.; Yoshida, N. J. Chem. Soc., Perkin Trans.
2001, 2, 1667–1673. (b) Lindner, H. J.; Yuan, D.-Q.; Fujita, K.; Kubo, K.;
Lichtenthaler, F. W. Chem. Commun. 2003, 1730–1731.
5302 J. Org. Chem. Vol. 73, No. 14, 2008

Azobenzenyl Modified ꢀ-Cyclodextrins Isomers
FIGURE 4. (a) X-ray crystal structure of 1 with H atoms and solvent water molecules omitted for clarity; the molecules are colored by atom type:
green, carbon atoms in the substituent; gray, carbon atoms in the ꢀ-CD; red, oxygen atoms; and blue, nitrogen atoms. (b) Stereodrawing of the
structure of the bimolecular capsule.
1
DMSO. However, a comparison of the H NMR spectra of 2
and 4 in D2O-DMSO (v:v ) 9:1) solution still gave useful
information about the assembly behavior of 2. As shown in
Figure 5, the δ value of the Ha protons of 2 showed a clear
broadened and upfield shift (ca. 0.1 ppm), while those of Hb
and Hc protons showed no appreciable shift, as compared with
the corresponding protons of 4.
Generally, the chemical shift values of the guest protons tend
to show appreciable changes if the guest molecules are included
in the CD cavities.14 Therefore, Figure 6 indicated that the III
ring of the substituent moiety of 2 partly entered the CD cavity.
Considering the self-unlocked structural feature of 2, where the
substituent was located out of its own ꢀ-CD cavity, we could
deduce that the III ring of 2 shallowly penetrated into the cavity
of an adjacent ꢀ-CD. But there may exist two possible self-
assembly modes of 2: a linear aggregation or a dimer with
mutual-insertion from the first face of the ꢀ-CD cavity. The
FTICR-MS (Fourier Transform Ion Cyclotron Resonance Mass
Spectrometry) of 2 clearly showed the peak of [3M2 + K +
H]2+ (2114.18). Judging from the structural feature of 2, the
formation of the 3M2 aggregate indicated that the aromatic
substituent of 2 must be intermolecularly included into the
hydrophobic cavity of another ꢀ-CD from the secondary
hydroxyl side. According to this intermolecular inclusion mode,
the formation of the linear supramolecule of 2 in the solid state
should be a natural process,15 although the existence of the
dimeric structure could not be rigorously ruled out. Just as many
aromatic modified CDs which have been demonstrated to form
the linear supramolecules (channel aggregations) in aqueous
FIGURE 5. Partial 1H NMR spectra of 2 (bottom) and 4 (top) in D2O-
DMSO-d6 (v:v ) 9:1) at 298 K with DSS as the standard.
FIGURE 6. Possible self-assembly structure of 2.
(the “yin-yang” type), where a CD dimer is formed with the
modified rims closely located, and the substituents are accom-
modated in each other’s cavities. The fourth, which is frequently
observed in the CD crystals, is the channel type, where the
substituent of a modified CD unit penetrates in the cavity of
the adjacent CD unit to form a polymeric supramolecule.
Interestingly, the self-assembly structure of 1 was found to not
belong to any of the above types, because its azobenzene
substituent was included in both its own cavity and the
counterpart’s cavity.
(14) Schneider, H.-J.; Hacket, F.; Rudiger, V.; Ikeda, H. Chem. ReV. 1998,
Self-Assembly of 2. Because of its poor water solubility, the
self-assembly behavior of 2 in water is difficult to investigate
by using the ROESY spectrum even in the presence of 10%
98, 1755–1786.
(15) Liu, Y.; Fan, Z.; Zhang, H.-Y.; Diao, C.-H. Org. Lett. 2003, 5, 251–
254.
J. Org. Chem. Vol. 73, No. 14, 2008 5303

Liu et al.
solution and the solid state,8d,15,16 the self-unlocked 2 showed
many appropriate linear aggregations in its high-resolution TEM
image (see the Supporting Information) And the XRD of 2
exhibited a strong peak at 2θ ) 18.380° (d ) 4.8230 Å, I/I0 )
100), which was reported to be a characteristic peak of channel
packing of CDs.17
crystals were collected by filtration and washed with acetone (3 ×
50 mL), and then recrystallized from boiling water. The precipitate
was collected by filtration to give 1 (0.48 g, yield 69%). 1H NMR
(300 MHz, DMSO-d6, ppm) δ 7.80-8.00 (m, 5H, H of alkene and
Ph), 7.45-7.65 (m, 3H, H of Ph), 7.20-7.30 (m, 2H, H of Ph),
5.60-6.00 (m, 14H, O-2,3 H of ꢀ-CD), 5.40 (d, 2H, H of
methylene), 4.76-4.90 (s, 7H, C-1 H of ꢀ-CD), 4.40-4.66 (m,
6H, O-6 H of ꢀ-CD), 3.00-4.00 (m, 42H, C-2, C-3, C-4, C-5, C-6
H of ꢀ-CD). ESI-MS m/z 1396.38 (M + H+), 1418.53 (M + Na+).
Elemental Anal. Calcd for C57H81O35N5 ·8H2O: C 44.44, H 6.35,
N 4.55. Found: C 44.57, H 6.34, N 4.41.
Conclusions
In summary, a couple of modified ꢀ-CD analogues with a
long and rigid substituent have been designed and synthesized
via the Huisgen cycloaddition in either the hydrothermal
synthesis condition or the “click” reaction condition and present
the distinct topological structure as self-locked or self-unlocked.
The self-locked 1 has a novel [1]rotaxane structure without the
stopper part. Interestingly, this [1]rotaxane self-assembles to a
unique bimolecular capsule. In contrast, its self-unlocked
analogue 2 exhibits a linear supramolecular structure. This is a
rare report on the differently restricted conformation of modified
CDs with the same composition, and these new observations
show clearly the influence of the conformational factors on the
self-assembly behaviors of modified CDs. These results will
be useful for our deep understanding of the supramolecular
aggregation phenomena, and may open a new way to obtain
the versatile materials from the controlled supramolecular self-
assembly.
Crystal data for 1: C114H224N10O101, M ) 3351.03, monoclinic,
space group P21, a ) 15.6068(19) Å, b ) 20.806(3) Å, c )
23.361(3) Å, R ) 90°, ꢀ ) 91.114(2)°, γ ) 90°, V ) 7584.4(16)
Å3, Z ) 2, Dc ) 1.467 g ·cm-3, λ(Mo-KR) ) 0.71070 Å, T )
113(2) K, F(000) ) 3572, µ ) 0.130 mm-1, approximate crystal
dimensions 0.32 × 0.28 × 0.26 mm3, θ range ) 1.30-26.00°,
reflections collected/unique 82797/28740 (Rint ) 0.0605), final R
indices [I > 2σ(I)] R1 ) 0.0763, wR2 ) 0.2010, R indices (all data):
R1 ) 0.0826, wR2 ) 0.2079, goodness of fit on F2 ) 1.033.
Synthesis of Mono-6-deoxy-6-{4-[(E)-4-(phenylazo)phenyl-
oxymethyl][1,2,3]triazolyl}-ꢀ-CD (2). 3 (0.58 g) and 4 (0.14 g)
were dispersed in a water-ethanol solution (12 mL, v:v ) 2:1),
and sodium ascorbate (0.08 g) and CuSO4 ·5H2O (0.05 g) were
added. The resulting mixture was stirred at room temperature for
8 h. The solvent was removed under reduced pressure, and the
residue was washed with acetone (3 × 100 mL), 15% ammonia (3
× 10 mL), and hot water (3 × 40 mL), then dried in vacuum to
give 4 as a yellow powder (0.50 g, yield 71%). 1H NMR (300 MHz,
DMSO-d6, ppm) δ 8.21 (s, 1H, H of alkene), 7.80-7.95 (m, 4H,
H of Ph), 7.48-7.60 (m, 3H, H of Ph), 7.18-7.30 (m, 2H, H of
Ph), 5.58-6.00 (m, 14H, O-2,3 H of ꢀ-CD), 5.20 (d, 2H, H of
methylene), 4.70-5.00 (s, 7H, C-1 H of ꢀ-CD), 4.40-4.60 (m,
6H, O-6 H of ꢀ-CD), 3.00-4.00 (m, 42H, C-2, C-3, C-4, C-5, C-6
H of ꢀ-CD). ESI-MS m/z 1396.51 (M + H+), 1418.64 (M + Na+).
FTICR-MS m/z 717.74 (M + K+ + H+), 1396.55 (M + H+),
1416.02 (2M + K+ + H+), 1434.50 (M + K+), 2114.18 (3M +
K+ + H+). Elemental Anal. Calcd for C57H81O35N5 ·5H2O: C 46.06,
H 6.17, N 4.71. Found: C 45.91, H 6.23, N 4.92.
Experimental Section
Synthesis of 4-Propargylazobenzene (4). To a solution of
acetone (30 mL) containing 4-hydroxyazobenzene (1.0 g) and NaH
(0.5 g) was added propargyl bromide (2.5 mL), and the resulting
mixture was stirred at room temperature for 3 h. The solvent was
removed under reduced pressure, and the residue was recrystallized
from petroleum ether to give 4 (0.9 g, yield 75%) as a salmon pink
1
crystall. H NMR (300 MHz, CDCl3, ppm) δ 7.80-8.00 (m, 4H,
H of Ph), 7.40-7.60 (m, 3H, H of Ph), 7.10-7.15 (m, 2H, H of
Ph), 4.77-4.78 (d, 2H, H of methylene), 2.56 (t, 1H, H of alkyne).
ESI-MS m/z 259.14 (M + Na+).
Acknowledgment.Wethankthe973Program(2006CB932900)
and NNSFC (Nos. 90306009, 20421202, and 20772062) for
financial support. We also thank the reviewers for their valuable
comments and suggestions regarding the revision.
Synthesis of Mono-6-deoxy-6-{5-[(E)-4-(phenylazo)phenyl-
oxymethyl][1,2,3]triazolyl}-ꢀ-CD (1). 3 (0.58 g) and 4 (0.14 g)
were dispersed in a water-ethanol solution (12 mL, v:v ) 2:1),
and the resulting mixture was stirred for 8 h. Then, the turbid
solution was put in a Teflon-lined stainless bomb, which was sealed
and then heated at 358 K under hydrothermal conditions for 3 days.
When the solution was cooled to room temperature, orange-red
crystals available for X-ray crystallography were obtained. The
Supporting Information Available: X-ray crystallographic
data of 1 in CIF format (CCDC no. 662572), general experi-
mental methods, 1H NMR spectra of 1-4, ESI spectra of 1 and
2, TEM images of 2, and the FTICR-MS spectrum of 2. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(16) Eliadou, K.; Giastas, P.; Yannakopoulou, K.; Mavridis, I. M. J. Org.
Chem. 2003, 68, 8550–8557.
(17) Huang, L.; Tonelli, A. E. J. Macromol. Sci., ReV. Macromol. Chem.
Phys. 1998, C38, 781–837.
JO800488F
5304 J. Org. Chem. Vol. 73, No. 14, 2008