Journal of the American Chemical Society
Communication
linkage is fully orthogonal to the pyridine→zinc(II) porphyrin
(Npy→ZnPor) binding but never in combination with a
pseudorotaxane-type complexation motif.44 Whereas the
pseudorotaxane unit has been the basis of many intricate
assemblies45 and interlocked molecular machines,46−48 its utility
in multicomponent molecular devices is still unexplored. Here,
the pseudorotaxane motif serves not only as a dynamic corner of
the five-component rotor along with the HETTAP and Npy→
ZnPor complexation but also as an anchor point for its out-of-
equilibrium formation.
Prior to designing detailed components of the rotor, three
fully orthogonal binding motifs were required with at least one
being responsive to an acid fuel. In this regard, we identified
complexes C1, C2, and C3 as promising candidates (Figure 2)
the ammonium moiety, shifted downfield from 4.95 and 4.00 to
5.51 and 5.27 ppm, respectively. The threading was also
confirmed by the emergence of multiple sets of cyclic ether
protons in the aliphatic region. UV−vis, fluorescence, and ESI-
further proved the formation of pseudorotaxane [(2·H+)(3)]
(log K = 5.58 0.03, Supporting Information, Figure S69).
The bent geometry of [(2·H+)(3)] with its terpyridyl and
pyridyl terminals allowed two connections to ligand 4 after
addition of Zn(OTf)2, one via a HETTAP linkage and the other
via Npy→ZnPor binding. The zinc(II) ion set up a strong
HETTAP43 complex between the central phenanthroline of 4
and the terpyridyl unit of the pseudorotaxane, whereas the
pyridine terminal of rotator [(2·H+)(3)] formed an Npy→ZnPor
coordination. The pyridine was expected to move quickly
between the two identical ZnPor binding sites, giving rise to
rotor 5 (Figure 3a). As expected from the clean self-sorting of
Figure 3. (a) Synthesis of rotor 5. (b) Comparison of partial 1H NMR
spectra (CD2Cl2, 400 MHz) of pseudorotaxane [(2·H+)(3)], 4, and
rotor 5.
Figure 2. Complexation motifs used in the five-component
pseudorotaxane rotor.
the basic building blocks, rotor 5 was furnished quantitatively, as
evident from 1H NMR, 1H−1H COSY, DOSY, and electrospray
mass spectroscopy (ESI-MS). For instance, ESI-MS peaks at m/
z = 1908.4 (doubly charged) and 1222.7 (triply charged)
attested the formation of [Zn(2·H+)(3)(4)]2+ (Supporting
Information, Figure S65). 1H NMR analysis showed the
characteristic upfield shifts of proton signals 10-H from 10.34
to 10.21 ppm and 15-H from 6.38 to 5.78 ppm, proving the
existence of both Npy→ZnPor and HETTAP complexation
(Figure 3b). Similarly, the pyridyl protons a′/b′-H were
characteristically shifted upfield from 8.54/7.25 ppm to 2.21/
5.36 ppm due to coordination at the ZnPor. The pseudorotax-
ane linkage was indicated by the signals of protons 21-H and 20-
H, adjacent to the ammonium moiety, that appeared at 5.43 and
5.15 ppm. Finally, a single set of 1H-DOSY signals (D = 4.5 ×
10−10 m2 s−1, r = 11.8 Å, in CD2Cl2) supported the formation of
the assembly (Supporting Information, Figure S47).
To measure the exchange frequency in rotor 5, we analyzed
the signal (Figure 4a) of proton 10-H, which showed up as a
sharp singlet (10.21 ppm) at 25 °C in the variable temperature
1H NMR. The single set of protons for both porphyrin units
attested fast rotation on the 1H NMR time scale. Diagnostically,
at −50 °C, it separated into two singlets (1:1) at 10.30 and 10.19
ppm. The signal at 10.19 ppm was assigned to the pyridine-
coordinated ZnPor station, whereas the upfield signal at 10.30
ppm referred to the freely rotating ZnPor. The kinetic analysis
provided the exchange frequency k at different temperatures,
with k298 = 15.4 kHz. From the Eyring equation, the activation
due to their interference-free self-sorting. As shown in Figure 2,
mixing of stoichiometric amounts of (7·H+)PF6 , 8−12, and
−
Zn(OTf)2 (1:1:1:1:1:1:1) generated the pseudorotaxane C1 =
[(7·H+)(11)]PF6 (log K = 4.74 0.20, Supporting Informa-
−
tion, Figure S68) and the strong HETTAP complex C2 =
[Zn(8)(10)]2+ (log β ≈ 15.0, Supporting Information, Figure
S70) together with the well-established Npy→ZnPor complex-
ation motif C3 = [(9)(12)] (log K = 4.31 0.12).49
On the basis of this insight, we designed the five components
of the rotor: protons, ligands 2, 3, and 4, and zinc(II) ions
(Figure 1). First, ligand 4 that contains two zinc(II) porphyrin
stations at both terminals and a phenanthroline station in the
middle was synthesized according to a known procedure.50 The
dibenzo-24-crown-8 and terpyridyl motifs were combined in
ligand 3, which acts as an axis of rotor 5. The rotator arm 2 was
conceived as a secondary amine with an anthracene stopper
carrying additionally a pyridine terminal. For the full synthesis of
When ligands 2, 3, and TFA were mixed (1:1:1) in DCM, the
ammonium moiety of the linear component 2·H+ (referred to as
thread) slid into the macrocycle of ligand 3. As expected, the 1H
NMR spectrum showed the typical upfield shifts of the crown’s
aromatic proton signals m-H, l-H, and n-H, while the peak at
6.89 ppm corresponding to o-H and p-H split in two sets
(Supporting Information, Figure S51). The signals of
anthracenyl and benzyl protons 21-H and 20-H, adjacent to
5320
J. Am. Chem. Soc. 2021, 143, 5319−5323