Communications
wheel rearranges and assumes a more-stiffened cone struc-
At this point a question came apparent: are these results
sufficient to exclude the threading of some of the axles 5a or
5b from the lower rim, or is this direction of threading
prevented because of the insufficient size of the calix[6]arene
annulus for these guests? Acetonitrile should decrease the
extent of ion pairing of the axle and hence limit the magnitude
of the pivoting role exerted by the ureido groups of the wheel
during threading.
ture. The downfield shift of almost Dd = 2 ppm of the
resonances of the six NH protons indicates their involvement
in hydrogen bonding with the two tosylate anions of 5a, as
also shown by X-ray structural data.[13] The position of the
calixarene along the axle was established by looking at the
variation in the chemical shifts of the protons of the dumbbell
portion between C1 and C1’ relative to those of 6 (see
Scheme 2 for numbering). In particular, the protons from C2
to C6’ in 8 undergo a substantial upfield shift, while those
from C5’ to a’ are downfield shifted to a much smaller extent.
The relative orientation of the dumbbell C6 spacers with
respect to the calixarene rims were established through
ROESY experiments, which showed cross peaks between
protons 4 and 5 of the axle and protons m, n, and o of the
wheel (see Supporting Information).
The lack of ROE cross peaks between the two different
stoppers and the corresponding hexyl chains, or with protons
of the wheel, precluded the existence of equivalent matching
spacer—stopper groups. Nevertheless, the presence of only
one signal for proton a at d = 5.24 ppm in the NMR spectrum
of 8[12] enables it to be assumed that the diphenylacetyl
stopper is positioned at the upper rim of the wheel, while the
dicyclohexyl one points toward the lower rim of the calix-
arene.
Pseudorotaxane formation in CD3CN was thus studied by
1H NMR spectroscopy by equilibrating 1 and 5a in a 1:1
molar ratio. In the spectrum of the solution thus obtained,
most of the signals appeared as broad peaks and this prevents
accurate assignment. However, the presence of three singlets
at d = 5.11, 5.13, and 5.14 ppm (relative ratio 9:59:32) in the
region where the proton a of the diphenylacetyl moiety
resonates suggested that this proton existed in three different
environments. In particular, the signal at d = 5.11 ppm was
easily assigned to the free axle while the other two were
tentatively assigned to the two isomeric pseudorotaxanes 7b
and 7a. A similar distribution of products was also observed
by dissolving the oriented pseudorotaxane 7a that had
formed in C6D6 in CD3CN. This observation shows that the
oriented pseudorotaxane 7a can dissociate into its compo-
nents in acetonitrile and that the threading/dethreading
processes can occur from both calix[6]arene rims, thus
yielding both pseudorotaxanes isomers (see the Supporting
Information). This result clearly indicates that the calix[6]-
arene annulus is also large enough to thread on to the axle
from the lower rim, and that indeed, in acetonitrile, the
control elements governing the exclusive threading from the
upper rim operate—if at all—only to a limited extent. Both
rotaxanes 8 and 10 were obtained in 24% yield as unique
chromatographic fractions by treating 1, axle 5a, and
dicyclohexylacetyl chloride in acetonitrile (Scheme 3). Their
1H NMR spectra in C6D6 correspond exactly to the sum of
rotaxanes 8 and 10, and thus indirectly demonstrate that only
To confirm the structure of 8, and also to exclude the
presence of other rotaxanes in the reaction mixture, the
rotaxane isomer 10 (see Scheme 2) was also synthesized (in
18% yield) by employing the new axle 5b, which is
characterized by the presence of a dicyclohexylacetic stopper
1
(see Scheme 1). As expected, the H NMR spectrum of 10
shows that the conformation of the calix[6]arene is almost
exactly the same as that observed in 8. However, the protons
of the dumbbell portion between C3’ to C6 in 10 are
substantially shifted upfield, while those from C5 to C1 are
shifted downfield. The presence of only one singlet at d =
5.23 ppm strongly suggests that the
diphenylacetyl stopper is in proximity
to the lower rim in 10, while the dicy-
clohexyl stopper is oriented toward the
upper rim of the calixarene (see the
Supporting Information).
These data can be tentatively
explained if it is considered that axles
5a and 5b are present as tight ion pairs
during the initial formation of the pseu-
dorotaxane in toluene and as such can
not fit into the calixarene cavity. These
conditions maximize the hydrogen-
bonding ability of the ureido NH
groups toward the axle counteranion(s)
that thus pivot the cationic portion of the
axles to thread through the wheel exclu-
sively from the upper rim. The high
kinetic stability of the oriented pseudo-
rotaxane that forms favors the stopper-
ing of the OH group that protrudes from
the lower rim of the wheel, without any
detectable axle–wheel isomerization.
Scheme 3. a) CH3CN, reflux; b) Cy2CHCOCl.
280
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 278 –281