4
G. Berton et al. / Tetrahedron Letters xxx (xxxx) xxx
Scheme 4. Heterochiral meso and homochiral racemic M3L2 assemblies obtained
from the achiral 3 triphenylene pyridine ligands.
an expected result, considering the smaller size of the former and
its more spherical shape. When 5 and m-3 were both present in a
3:2 M ratio, two effects can be highlighted: 1) the diffusion coeffi-
cient of both species is identical; 2) the common diffusion coeffi-
cient is lower than both the two corresponding free species,
which is a strong indication of the formation of a new coordination
species. Based on the flat non-spherical shape of the ligand and the
similar but thicker shape of the expected M3L2 assembly, the
decrease of the diffusion coefficient observed agrees with the pos-
sible formation of a trigonal bipyramidal cage. Considering the
broadening of the resonances of the coordination species, which
is an indication of an ongoing exchange process, it was not possible
to extrapolate an indication of the MW of the aggregate in solution
[30]. Nevertheless, the value of the diffusion coefficient is not dras-
tically different with respect to that of the single triphenylene
ligand m-3 in solution, thus excluding the formation of polymeric
assemblies and of very large aggregates characterized by the same
M3L2 stoichiometry but eventually composed of multiple of these
units.
Fig. 4. 19F NMR spectra during the titration of m-3 with increasing amounts of 5
(each spectrum after 0.1 eq. of metal complex) (top); plot of the 19F NMR chemical
shift of the triflate anion vs the molar fraction of ligand (bottom).
The triphenylene ligand m-3 was also investigated with other
metal corners, such as [Pt(dppe)(OTf)2] 6, in order to investigate
the effect of the different metal center characterized by a generally
lower rate of exchange of the ligands. The titration of m-3 with 6
was monitored by multinuclear NMR experiments, showing a ser-
ies of spectra which were very similar to those observed for the
metal complex 5. 19F NMR spectroscopy was again the most effi-
cient experiment to determine the stoichiometry of the assembly,
observing a drastic change in chemical shift for the triflate anion
when a 0.4 ligand molar fraction was reached, corresponding to
the formation of a cage with M3L2 stoichiometry (see ESI, Titration
4). We also investigated the effect of the temperature on the NMR
spectra, either in CD3CN at 300 K and 348 K or in CD2Cl2 at 193 K
and 233 K, observing in all cases complicated spectra (see ESI). In
particular, only at high temperature in CD3CN were the resonances
of the assembly slightly sharper, even though the number of reso-
nances was high. The pseudo 2D-DOSY NMR spectra of the M3L2
assembly in CD3CN at 348 K, reported in Fig. 7, shows how the sig-
nals of the ligand and m-3 and 5 have the same diffusion coeffi-
cient at high temperature.
The interaction between m-3 and the [Pd(dppp)(OTf)2] 7 was
considered to investigate the effect of a larger bite angle on the
Pd(II) corners. The results of the titration are reported in the ESI
(Titration 5), from which in this case the 19F NMR experiments sug-
gest a 1:1 ratio between the tridentate ligand and the metal. The
pseudo-2D-DOSY experiments with this combination of metal
complex and ligand supported the formation of polymeric aggre-
gates in solution rather than a discrete M3L2 cage, due to the obser-
vation of very broad cross-peaks in the DOSY spectrum,
corresponding to a rather small and uncertain diffusion coefficient
for the 1:1 mixture of 7 and m-3. The comparison of this diffusion
coefficient with that of the free triphenylene ligand m-3 suggested
Fig. 5. 1H NMR spectra of 5 with m-3 in a M3L2 ratio in CD3CN at 347 K, 332 K and
302 K (from top to bottom).
the original spectrum, indicating that the system is stable and
reversible in the range of temperatures investigated.
The exchange process that involves the metal-ligand assemblies
in solution could also be explained by considering that two mole-
cules of 3, facing each other through coordination with the metal,
form two diastereomeric assemblies, due to the combination of the
two enantiotopic faces of triphenylene 3. Therefore, two possible
diastereoisomers M3L2 are possible: one C3-symmetric between
homochiral ligands (as a racemate) and the C3h-symmetric meso
with heterochiral ligands, thus causing the doubling of all reso-
nances (Scheme 4).
To further investigate the formation of the coordination cage in
solution, a series of pseudo 2D-DOSY NMR spectra were recorded
for the free ligand m-3, the free complex 5 and the M3L2 assembly
in CD2Cl2 solution at 300 K (Fig. 6). The comparison between these
experiments showed that the diffusion coefficient in solution of the
metal complex 5 is larger than the free triphenylene m-3, which is
Please cite this article as: G. Berton, T. Lorenzetto, G. Borsato et al., Triphenylene based metal-pyridine cages, Tetrahedron Letters, https://doi.org/10.1016/j.