.
Angewandte
Communications
position (Supporting Information, Figure S19). A more
accurate analysis by UV/Vis spectroscopy showed t1/2 = 102 s
(258C) and pseudo-first-order behavior (both reactants used
at c = 5 ꢁ 10ꢀ6 m). The two metal ion translocations are thus
kinetically independent of the concentration of the metal-free
nanoswitch, implying that the rate-limiting step involves an
intramolecular dissociation of the metal complex = state II.
Moreover, because the translocation of both Cu+ and Zn2+
ions proceeds within a few minutes, the communication is of
rather similar speed in both directions.
After these promising results, we checked for bidirec-
tional reversible redox-controlled copper ion translocation. A
solution (1:1:1) of 1, 2, and [Cu(CH3CN)4]B(C6F5)4 in
dichloromethane (c = 0.80 mm), thus of [Cu(1)]+ and 2,
served as the starting point.[15] A scan to anodic potential in
a cyclic voltammetry (CV) experiment revealed a small
oxidation wave at 0.25 VSCE for [Cu(2)]+/2+ (10%) and a large
oxidation current at circa 0.75 VSCE for [Cu(1)]+/2+ (Support-
ing Information, Figure S33). Unfortunately, no Cu2+ trans-
location from [Cu(1)]2+ to 2 generating [Cu(2)]2+ was
recognized on the CV time scale in the reverse sweep, even
when we used low scan rates, such as 50 mVsꢀ1. Equally, the
translocation did not show up when the CV was stopped at the
anodic switching potential for 60 s before starting the
reductive scan. Such finding is in agreement with the slow
rate of Zn2+ translocation taking place only within minutes
(see above), as most likely Cu2+ will behave similarly owing to
analogous coordination preferences. On the other side,
exclusive formation of [Cu(2)]2+ was detected in ESI-MS
investigations 10 minutes after adding switch 2 to [Cu(1)]2+
(Supporting Information, Figure S27). Because of the low
concentrations (0.80 mm) and inadequate time scale of CV
experiments, metal ion transfer is thus not detectable after
oxidation in a single CV scan.
Figure 3. ESI-MS signature (top) of a solution containing 1, 2, and Cu+
(1:1:1) as obtained after oxidation with one equiv of TBPA+C in CH2Cl2;
(bottom) after reduction of the above solution with one equiv of
dmfc.[14]
two-electron oxidation produces a diamagnetic species. In
CD2Cl2, a [Cu(1)]+/[Cu(2)]+ ratio of 88:12 was observed,
while in presence of 2% CD3CN the ratio was improved to
95:5 (Supporting Information, Figure S30).[20]
Unfortunately, all attempts to monitor the kinetics of Cu2+
translocation by UV/Vis spectroscopy were met with failure
owing to the fact that the absorbance changes are too small,
but the rate of the process should to be similar to that of the
Zn2+ translocation (see above).
In conclusion, we report herein on several examples of
highly selective communication between two nanomechanical
switches and their kinetics, as effected by intermolecular
metal ion translocation. Even at micromolar concentrations,
communication in both directions is fast occurring at t1/2 = 2–
3 min with monomolecular metal-ion–ligand dissociation
being the rate determining step. The metal ion translocation
results in a circa 20 ꢀ relocation of the toggle at both
nanoswitches and thus to major geometric and electronic
changes, similar as those in biological systems, that may allow
to even link communication with switching ON/OFF (orga-
no)catalysis.[21]
A reversible bidirectional communication is possible
using chemical or electrochemical oxidation/reduction.
Accordingly, oxidation of [Cu(1)]+ and 2 with TBPA+C leads
to Cu2+ translocation furnishing 1 and [Cu(2)]2+, while after
reduction with dmfc or BFD the translocation of Cu+ yields
the starting state = ([Cu(1)]+ and 2).
In summary, our work opens new vistas for systems
chemistry, such as how to interconnect molecular computer
and storage units or artificial synapses by chemical exchange
processes. Such intermolecular switching is distantly reminis-
cent of that found in neural networks, relying on electro-
chemical signals between regions of the brain and spinal cord.
Furthermore, communication between nanoswitches may be
coupled to chemical follow-up processes.
To probe the redox-initiated toggling on a slower time
scale, we decided to use chemical redox agents. For oxidation,
we chose tris(4-bromophenyl)aminium hexachloroantimo-
+
ꢀ
nate (TBPAC SbCl6 )[16] and for reduction, decamethylferro-
cene (dmfc)[17] or a 3-alkyl-1,1’-biferrocenylene (BFD).[18,19]
To a mixture of [Cu(1)]+ and 2 in dichloromethane (prepared
+
as above), one equivalent of TBPAC was added. After 3 min,
a reductive scan CV analysis that was started at + 1.0 VSCE
(Supporting Information, Figure S34) showed
a sharp
increase in current at 0.25 VSCE, suggesting increased forma-
tion of [Cu(2)]2+. In ESI-MS, the molecular ion peak at m/z =
939.1, assigned to the dicharged [Cu(2)]2+, shows up almost
exclusively (Figure 3, top). The resultant solution of [Cu(2)]2+
was now reduced by adding 1 equiv of dmfc (Supporting
Information, Figure S35). An oxidative scan started at
ꢀ0.9 VSCE 2 minutes after mixing exhibited a lower current
signal for [Cu(2)]+ and a larger signal for [Cu(1)]+ (includes
Cu+/2+ and por0/+C redox transitions) than in the initial state.
Thus, reduction of [Cu(2)]2+ with dmfc led to translocation of
Cu+ ions to 1. An ESI mass spectrum recorded 10 minutes
after mixing displayed peaks at m/z = 1799.2 (major) and
1876.2 (minor) (Figure 3, bottom), indicative of [Cu(1)]+ and
[Cu(2)]+, respectively, thus confirming return to the starting
state. For quantification, backward translocation was moni-
1
tored by H NMR using BFD as reducing agent because its
4
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!