Communication
At this point, we set out to study the driving force of the
self-assembly further. The related crown ether 1’ was synthe-
sized featuring one diacetylene group in the place of one
naphthalene unit (Supporting Information, Scheme S7). How-
ever, complexation with the NDI unit was not indicated by the
resulting NMR spectra (Supporting Information, Figure S6). This
observation highlights the importance of the aromatic donor–
acceptor interactions. In addition, as shown in Figure 1c, NMR
studies of the effects of different solvents showed that the
pseudorotaxane structure is more stable in methanol, a protic
solvent, rather than in chloroform. We concluded that the ob-
served p–p stacking, as the key driving force, could be further
described as a synergic combination of aromatic donor–ac-
ceptor interactions and solvophobic effects.[14]
Based on the above results concerning the dynamic nature
of the assemblies, the tectons 2a and 2b were also obtained
by an alternative method, whereby 2a’ and 2b’ were treated
with 1 directly (Yields: 2a: 58%, 2b: 53%). Although pure 2a
and 2b could be separated from uncatenaned mixtures by
careful recrystallization, there is little significance to separate
them considering their dynamic nature in solution. The pre-
pared binuclear tectons 2a and 2b were then used in situ to
construct architectures with structures of higher order.
Figure 1. Selected characterization methods of pseudorotaxane tectons.
a) Molecular structure of complex 2a. Counter anions and hydrogen atoms
are omitted (N=blue, O=red, C=gray, Rh=orange). b) ESI-MS spectrum of
2a (left) and 2b (right). c) 1H NMR spectra of i) crown ether 1 in CDCl3, ii) un-
catenaned tecton 2a’’ in CD3OD/CDCl3 (1:3), iii) complexation in CD3OD/
CDCl3 (1:3, 12.5 mm), iv) complexation in CD3OD (12.5 mm).
Initially, pyrazine, one of the simplest ditopic donors, was
used as a linker for the self-assembly of [2]catenanes. The
[2]catenane 3a was prepared by mixing the binuclear tecton
2a, the uncatenaned tecton 2a’, and pyrazine in a ratio of
1:1:2. Compound 3b was obtained by using a related proce-
dure with Cp*Ir-based tectons (Yields: 3a: 86%, 3b: 84%).
Direct mixture of the crown ether 1 and the metalla-rectangles
3a’ or 3b’ was found to be an alternative method to prepare
the catenanes (Yields: 3a: 85%, 3b: 83%).
concentrated solution in MeOH/DMF/CHCl3. X-ray diffraction
analysis unambiguously revealed the molecular structure of
the 1:1 inclusion complex 2a, as shown in Figure 1a. The struc-
ture is stabilized by parallel face-centered D–A–D (D=donor,
A=acceptor) stacking interactions with interplanar distances
of approximately 3.40 ꢁ, whereas no significant hydrogen
bonds were observed. We therefore assume that p–p stacking
interactions are the key driving force for the self-assembly. The
prepared pseudorotaxane features a trans conformation in the
solid state and is end-capped with two Cp*Rh fragments, each
of which binds one weakly coordinating DMF molecule.
1H NMR studies of both 3a and 3b were subsequently em-
ployed to confirm our design. Unexpectedly, due to the iso-
1
merization of the tectons in solution, the H NMR spectra were
too obscure to assign or integrate all of the resonances. Alto-
gether five isomers are possible for 3a or 3b, which result
from the two conformations of the dinuclear tectons—the cis-
form and the trans-form—and the orientation of the binuclear
tectons when assembling the tetranuclear complexes. Despite
the spectra being complicated, some crucial information could
still be extracted. The resonance of the Cp* groups splits from
a single peak for 3a’ into two peaks for 3a. The ratio of inte-
gration of the peaks assigned to the Cp* groups is exactly 1:1,
which is in accordance with the [2]catenane structure. In the
case of the iridium complex 3b’, the resonance of the Cp*
groups splits into two peaks due to isomerization on the NMR
timescale, and into four peaks after catenation of 1 (Support-
ing Information, Figure S14). Furthermore, new well-resolved
resonances in the “shielded” region (d=6.0–5.0 ppm) indicated
The behaviors of 2a and 2b in solution were studied by ESI
mass spectrometry (ESI-MS) and 1H NMR spectroscopy. The
prominent peaks corresponding to 2a at m/z=1557.30 and
2b at m/z=1737.41 confirmed the existence of the catenaned
1
complexes in solution. However, the H NMR spectra obtained
from the complexes showed that despite the bulky Cp*M (M=
Rh, Ir) stoppers, the crown ether still partially slips out. This is
most probably due to the dynamic nature of the labile coordi-
nation bonds. Therefore, an equilibrium was observed in solu-
tion. Furthermore, as shown in Figure 1c, the considerable
broadening of the resonances also indicated the dynamic
nature of the complexes. In addition, the spectra suggested
the existence of isomers of 2b on the NMR timescale at ambi-
ent temperature (Supporting Information, Figure S4). Hence
the binuclear tectons 2a and 2b are best described as pseu-
dorotaxanes rather than rotaxanes.
1
the catenation for both 3a and 3b. H DOSY experiments for
3a and 3b also supported these results. The corresponding
peaks displayed a single diffusion coefficient, suggesting one
stoichiometry for the assembly.
The structure of the [2]catenane 3a was confirmed by
single-crystal X-ray diffraction analysis, as shown in Figures 2a
and b.[15] We prepared single crystals of 3a by vapor diffusion
&
&
Chem. Eur. J. 2019, 25, 1 – 6
2
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!