Mechanistic evaluation of motion in redox-driven rotaxanes reveals longer linkers hasten forward escapes and hinder backward translations

Article abstract of DOI:10.1021/ja5013596

Mechanistic understanding of the translational movements in molecular switches is essential for designing machine-like prototypes capable of following set pathways of motion. To this end, we demonstrated that increasing the station-to-station distance wil

Full text of DOI:10.1021/ja5013596

Article
pubs.acs.org/JACS
Mechanistic Evaluation of Motion in Redox-Driven Rotaxanes Reveals
Longer Linkers Hasten Forward Escapes and Hinder Backward
Translations
Sissel S. Andersen,†,‡ Andrew I. Share,‡ Bjørn La Cour Poulsen,† Mads Kørner,† Troels Duedal,†
Christopher R. Benson,‡ Stinne W. Hansen,† Jan O. Jeppesen,*,† and Amar H. Flood*,‡
†Department of Physics, Chemistry, and Pharmacy, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark
‡Department of Chemistry, Indiana University, 800 East Kirkwood Avenue, Bloomington, Indiana 47405, United States
S
* Supporting Information
ABSTRACT: Mechanistic understanding of the translational move-
ments in molecular switches is essential for designing machine-like
prototypes capable of following set pathways of motion. To this end,
we demonstrated that increasing the station-to-station distance will
speed up the linear movements forward and slow down the
movements backward in a homologous series of bistable rotaxanes.
Four redox-active rotaxanes, which drove a cyclobis(paraquat-p-
phenylene) (CBPQT4+) mobile ring between a tetrathiafulvalene
(TTF) station and an oxyphenylene station, were synthesized with
only variations to the lengths of the glycol linker connecting the two stations (n = 5, 8, 11, and 23 atoms). We undertook the first
mechanistic study of the full cycle of motion in this class of molecular switch using cyclic voltammetry. The kinetics parameters
(k, ΔG⧧) of switching were determined at different temperatures to provide activation enthalpies (ΔH⧧) and entropies (ΔS⧧).
Longer glycol linkers led to modest increases in the forward escape (t1/2 = 60 to <7 ms). The rate-limiting step involves
movement of the tetracationic CBPQT4+ ring away from the singly oxidized TTF+ unit by overcoming one of the thiomethyl
(SMe) speed bumps before proceeding on to the secondary oxyphenylene station. Upon reduction, however, the return
translational movement of the CBPQT4+ ring from the oxyphenylene station back to the neutral TTF station was slowed
considerably by the longer linkers (t1/2 = 1.4 to >69 s); though not because of a diffusive walk. The reduced rate of motion
backward depended on folded structures that were only present with longer linkers.
modulate movements in molecular machines and motors.
Despite their importance, there are still only a few
studies11,18,24−28 aiming to elucidate the microscopic details of
the pathways of motion in bistable interlocked molecules. There
are even fewer studies21,22,29 that characterize a complete cycle
both forward and backward; none of which involve the systems of
tetrathiafulvalene (TTF) and cyclobis(paraquat-p-phenylene)
(CBPQT4+).
A number of strategies have been investigated to understand
and control the rates of motion in rotaxanes6,7 and catenanes.30,31
One approach has involved changing and tuning the stations11,18
where deeper potential wells slow molecular movements. Such
modifications often have the undesired effect of influencing the
thermodynamics and, consequently, the Boltzmann-weighted
change in populations during switching. More direct efforts22 to
control kinetics have utilized large bulky groups to completely
block the pathway between two stations. Steric-based speed
bumps,17,24,25,33 redox-switchable electrostatic speed bumps,27,34
photoswitchable gates and dynamic foldameric linker regions28
have also been employed to control the kinetics of motion. In an
INTRODUCTION
■
Biological molecular machines rely upon the regulation of
chemical processes,1 among others, to ensure a specific
sequence2 of events is followed in order for work done in one
movement not be undone in the next. Consequently, the precise
timing of each step in the cycle of motion is vital for performing
work and achieving unidirectional movements.3 Synthetic
molecular switches and machines have been synthesized and
studied3 as a means to mimic these underlying principles.4 Yet,
the vast majority of these investigations have focused on
identifying ways to control the thermodynamics of motion.3,5
Considering rotaxanes as exemplary,6−9 this outcome has been
highly successful and is now characterized by the almost
predictable5,10−13 control over the reversible linear motion of a
ring component between primary and secondary stations along
the dumbbell, a property called bistability. The thermodynamic
character of the resulting linear movements, however, is
inherently path independent. One way to better mimic the
behaviors of biomachines is to instead develop an understanding
of path-dependent properties;14−17 such as establishing knowl-
edge of how molecular structure impacts the kinetics of
motion.11,18 Mechanistic studies, therefore, are essential19−23
for developing and testing strategies on how to control, gate, and
Received: February 8, 2014
Published: April 18, 2014
© 2014 American Chemical Society
6373
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
analogous manner, modulating the size of the mobile ring has
been found to have a significant effect on the kinetics of switching
in rotaxanes.29,35,36 An alternative approach involves modifica-
tions to the length of the native linker26,37,38 that connects the
two different stations together.
Scheme 1. Overall Cycle of Voltage-Driven Switching in
a
TTF−CBPQT4+ Rotaxanes
An exemplary study by Leigh26 made use of time-resolved
vibrational spectroscopy39 to watch the departure and arrival of a
mobile ring between the two stations in a series of amide-derived,
hydrogen-bonded bistable rotaxanes. The probability of the ring
moving forward to the secondary station was found to be split
into two kinetically resolved processes: First, escape from the
primary station required thermal activation. Second, the
probability of the ring stepping forward or backward along the
alkyl linker region with n = 5, 9, 12, or 16 methylene units was
based on a diffusive, yet biased, random walk along a one-
dimensional pathway. When longer linkers (n = 16) were used,
the net probability of the ring moving to the second station
became smaller, resulting in slower kinetics of switching for this
step and for the entire process. The slowing of the rates with
increasing linker length has also been observed in the threading
and dethreading of a macrocycle along a polymer chain bearing a
viologen station.37 While the diffusive walks in these cases
contributed to the observed rates, it has also been proposed that
the thermally activated escape from the initial station will
dominate26 the rate behavior in interlocked molecules. Thus,
there exist two alternative ideas regarding the rate-limiting steps
(RLSs) of translational movements in rotaxanes that can best be
distinguished using mechanistic studies.
a
The overall switching cycle starting from rotaxane R4+ (top) involves
oxidation to R5+ that is followed by switching to SR5+ which is almost
always oxidized immediately to SR6+ when the rotaxanes are under
voltage control. The switched rotaxane SR6+ is reduced in two steps to
SR4+ which allows molecular movements of the ring (blue) to reinstate
rotaxane R4+.
One of the most studied and reliable set of molecular switching
gear is based on the interaction between the redox active TTF
unit and the CBPQT4+ ring.6,11,24,30,32,40 Flexible ethylene glycol
linkers are almost exclusively employed to connect the TTF unit
to other stations on account of their ability to aid41 in the
template-directed synthesis of rotaxanes. These linkers also help
stabilize the CBPQT4+ ring once it is localized at a station, i.e.,
they participate in the thermodynamic stabilization of one station
over another.11 Few studies,23,42 however, have been directed at
examining their role in the mechanisms of redox-driven motion.
Furthermore, within the class of TTF−CBPQT4+ rotaxanes, the
mechanisms of motion for both the forward and backward
switching are usually studied in separate compounds; the
mechanism of a complete cycle has yet to be examined across
a homologous series.
terized structurally using NMR spectroscopy on control
compounds also as a function of chain length. Cyclic
voltammetry (CV) was recorded at fast and slow scan rates
and then simulated to characterize kinetics (k) and activation
parameters (ΔG⧧, ΔH⧧, ΔS⧧) of switching. The effects of the
linker’s length on the kinetics gave evidence for the mechanism of
motion in this class of rotaxane. In contrast to Leigh’s findings,26
the movement of the ring in the forward direction was hastened
modestly when the linker was extended in length while the return
motion was slowed considerably. We observe rates that are
largely limited by the thermally activated escape from either the
primary or secondary stations. Unlike Leigh’s findings, net rates
of motion forwards or backwards are too small to reveal any
biased random walks. Nevertheless, these movements still
display a dependence on the length of the linker. Overall, these
findings indicate that the linker’s length can influence the escape
energy and thus the microscopic details of motion just as it can in
a diffusive random walk between stations.
Herein, we characterize the impact of the ethylene glycol linker
lengths on the translational motion forward and backward within
a series of rotaxanes. We have measured the kinetics of motion
4+
for a complete cycle (Scheme 1) using bistable rotaxanes RS
,
RM4+, RL4+, and RXL4+ (Figure 1) with short (S), medium (M),
long (L), and extra long (XL) glycolic linkers composed of 5, 8,
11, and 23 atoms, respectively. In this class of rotaxanes, it is
known43 that oxidation of the TTF unit (green) to TTF+ is
sufficient to repel the tetracationic CBPQT4+ ring (blue) causing
it to move to a secondary station (red). Subsequent oxidation to
TTF2+ generates the switched rotaxane, SR6+. Full reduction
back to the neutral TTF station removes all electrostatic
repulsions and allows the macrocycle to return to its initial
position (Scheme 1). Herein, switching was characterized by 1H
NMR spectroscopy to confirm the location of the CBPQT4+ ring
in the switched state for each rotaxane. During these
investigations it was also discovered that oxyphenylenes located
on the stoppers act as secondary stations. The location of the
mobile rings and the associated conformations of the switched
rotaxane SR6+ and metastable intermediate SR4+ were charac-
RESULTS AND DISCUSSION
■
Molecular Design. Studies, presented herein, revealed the
two identical oxyphenylenes (red, Figure 1) act as secondary
stations to complement the primary redox-active TTF station.
The outcome is a three-station, bistable rotaxane capable of
switching left or right with equal probability. Within each
rotaxane, thiomethyl groups (SMe, purple, Figure 1) were
incorporated on either side of the central TTF station to slow the
movements down such that the rates forward (<60 ms) and
backward (>1 s) could both be quantified with electrochemical
methods by using a range of temperatures (263−333 K) and scan
rates (166−0.2 V s−1). The linkers are extended by one ethylene
glycol across the short, medium, and long rotaxanes, while they
double in length from short to long to extra long rotaxanes. Using
electrogenerated TTF+, we can watch the escape of the mobile
ring in the forward direction, however, we cannot observe its
6374
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
4+
Figure 1. Molecular structure and cartoon representations of [2]rotaxanes RS4+, RM4+, RL4+, and RXL
.
1
arrival at the secondary station by electrochemical means. By
contrast, the times of departure and arrival can be reliably
measured for the return movement.
anes had similar H NMR spectra (Figure S1), indicating they
have similar structures. Following oxidation, the observations of
significant shifts to all of the protons in the spectra were
consistent with switching.49−52 The new location of the
CBPQT4+ ring along the dumbbell of each rotaxane was readily
determined from the strong shielding felt by the oxyphenylene
protons when they are encircled by the ring (e.g., Ha and Hb,
Figure 2).
Synthesis. The four different dumbbells that are precursors
4+
to RS4+, RM4+, RL4+, and RXL were all synthesized in a similar
fashion from the cis/trans mixture of the same TTF derivative,44
the same bulky tetraaryl stopper,45 and four glycols of different
lengths with 5, 8, 11, and 23 atoms. The TTF unit served as a
template for the formation of the tetracationic CBPQT4+ ring
around the dumbbell from the dicationic precursor46 and 1,4-
bis(bromomethyl)-benzene. The dumbbells and rotaxanes were
characterized by NMR spectroscopy, mass spectrometry, and
elemental analyses (see Supporting Information).47
In the 1H NMR spectrum of the switched rotaxane bearing the
short linker, SRS6+ (Figure 2c), the protons associated with the
SMe speed bumps are shifted downfield (δ = 2.92 and 2.87 ppm)
relative to the position at which they are found to resonate in the
4+
parent rotaxane, RS (δ = 2.58 ppm, Figure 2b). Previous
1
Chemical Switching Studies using H NMR Spectros-
investigations have shown that such behavior is entirely
consistent with the formation of the TTF2+ dication49,50 and,
therefore, of the oxidized rotaxane. The inequivalence in their
positions is indicative of the lower symmetry in SRS6+ in which
the CBPQT4+ has moved away from TTF2+. Evidence for the
linear translation of the ring to encircle one of the O-linked
copy. In order to investigate the location of the ring in the
switched state, SR6+, 1H NMR spectroscopy was used to monitor
the changes to the hydrogen environments upon addition of 10
equiv of the chemical oxidant tris-p-bromophenylammoniumyl
hexachloroantimonate (TBPASbCl6).48 Initially all four rotax-
6375
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
Figure 2. (a) The structure of SR6+ with hydrogen labels. Partial 1H NMR spectra (400 MHz, 298 K, CD3CN, 2.0 mM) of (b) RS4+, (c) SRS6+, the signal
6+
from one of the SMe groups (i.e., δ = 2.87 ppm) is obscured under the H2O signal (broad signal around ∼2.8 ppm), (d) SRM6+, (e) SRL6+, and (f) SRXL
.
The switched rotaxanes SRi6+ (i = S, M, L, and XL) are all generated by adding 10 equiv of the oxidant TBPASbCl6. The signals being crossed out arise
from traces of DMF present in the samples after the synthesis.
oxyphenylene stations is made obvious from the dramatic
changes (Figure 2c) of the chemical shift position of that
particular O-linked oxyphenylene’s Ha and Hb protons. In the
starting rotaxane RS4+, the resonances appear at δ = 7.10 ppm
(Ha) and δ = 6.77 ppm (Hb). Upon oxidation, the signals for one
of the oxyphenylene stations experience large upfield shifts to δ =
6.22 ppm (Ha) and δ = 2.48 ppm (Hb) on account of the
anisotropic shielding effect (Figure 2c) that occurs when the
CBPQT4+ ring encircles that station. The significant upfield shift
(Δδ = 4.29 ppm) of the oxyphenylene’s Hb protons can most
likely be explained by the fact that they participate in C−H···π
interactions with the CBPQT4+ ring by pointing toward the faces
of the p-xylene units of CBPQT4+. The signals for the other
oxyphenylene station remain almost at the original positions (δ =
7.18 and 6.80 ppm for Ha′ and Hb′, respectively). Examination of
differences in the chemical shift positions of the protons
associated with the oxyphenylene station and neighboring
glycolic protons indicate that the CBPQT4+ ring is centralized
on the oxyphenylene station to differing degrees across the
switched rotaxanes (Figure 2c−f). Among the rotaxanes, the one
6+
with the shortest linker, SRS (Figure 2c), shows the largest
effect of shielding on the oxyphenylene protons concomitant
with the smallest shielding on the neighboring glycolic chain.
These observations indicate that the ring is closer to the
oxyphenylene station in the short rotaxane. In the extra long
6+
rotaxane, SRXL (Figure 2f), however, greater shielding effects
are seen on the glycolic chain, while the oxyphenylene protons
are less shielded than in the short rotaxane SRS6+. Thus, the
CBPQT4+ ring spends more of its time surrounding the
oxyphenylene protons when the linker is short. Consistently,
NMR data of the rotaxanes with the medium and long linkers
show chemical shifts that are positioned between the two
extremes.
1
6+
the H correlation spectroscopy (COSY) spectrum for SRS
recorded in CD3CN at 298 K (Figure S3) clearly shows the
through-bond scalar coupling between the two shielded protons
in the oxyphenylene station. Consistently, the protons associated
with the two methylene groups (−O−CH2−CH2−) attached
The preference of the CBPQT4+ ring for the oxyphenylene
unit in SRS6+ when compared to the longer rotaxanes SRL6+ and
SRXL6+ could arise from stronger electrostatic repulsions between
the dicationic TTF2+ and the tetracationic CBPQT4+ ring on
account of the shorter distance between like charges. In order to
investigate this hypothesis, a set of rotaxanes (Figure 3a) was
designed to serve as controls in which the thioether is changed
from SMe to a larger SEt.
6+
directly to the encircled oxyphenylene station in SRS are
similarly shielded (Figure 2c) and display through-bond scalar
coupling (Figure S3). Taken all together, these observations
unambiguously show that the CBPQT4+ ring has localized on the
oxyphenylene stations upon formation of the TTF2+ dication.
Similar observations (Figure 2d−f) were made for rotaxanes
RM4+, RL4+, and RXL4+ following oxidation and switching.
It is known25 that the electrostatic repulsion provided by an
oxidized TTF2+ provides additional impetus for the tetracationic
CBPQT4+ to pass over the SEt speed bump to prepare SEt−SR6+
(see Figures S5 and S6). However, the SEt barrier is too big for
Location of the Ring in the Switched States SR6+ and
SR4+ as a Function of Linker Length. While switching is
clearly evident from the 1H NMR spectra of all oxidized species,
6376
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
(SRS6+, Figure 4c). The only exception lies with the SMe signals
now being replaced by resonances from the SEt groups (Figure
4b). After reduction using Zn powder, the neutral rotaxane SEt−
SRS4+ is produced. The resulting 1H NMR spectrum (Figure 4d)
has a greater similarity to the spectrum of the oxidized rotaxane
6+
SEt−SRS (Figure 4b) than it has to its parent rotaxane SEt−
4+
RS (Figure 4a) present at the beginning. For instance, the
chemical shift position (Figure 4d) of the oxyphenylene protons
Ha (δ = 6.25 ppm) and Hb (δ = 2.53 ppm) and the two methylene
groups (−O−CH2−CH2−) at ca. 3.5 and 1.5 ppm attached
4+
directly to the encircled oxyphenylene station in SEt−SRS
almost completely retain their shielded positions (Figure 4d).
The hydrogen resonances on the −O−CH2−CH2− linker
attached directly to the oxyphenylene station move slightly more
4+
upfield (Figure 4d) in SEt−RS than either of the oxidized
rotaxanes. We attribute this observation to the proximity of the
neutralized TTF rather than any movement of the CBPQT4+ ring
toward the −O−CH2−CH2− group. The SEt−based rotaxanes
with medium and long linkers show identical outcomes. These
observations are consistent with electrostatics only playing a
small role in positioning CBPQT4+ in the oxidized, switched
rotaxanes.
Figure 3. (a) Molecular structure describing [2]rotaxane SEt−RS4+ (n =
1), SEt−RM4+ (n = 2), and SEt−RL4+ (n = 3). (b) Generalized cartoon
representations illustrating the preparation of SEt−SR4+ starting from
SEt−R4+.
Returning to the fact that the CBPQT4+ appears to remain on
the oxyphenylene station in the SEt-based rotaxanes after the
TTF unit is returned to its neutral form suggests that the
CBPQT4+ ring could have some small affinity for the
oxyphenylene station. This idea is consistent with the fact that
the oxyphenylene’s Hb protons seem to be engaged in C−H···π
interactions with the CBPQT4+ ring (vide supra) and can also be
explained by the chemical similarity between an oxyphenylene
unit and a hydroquinone (HQ) moiety, which is known54 to act
as a station for CBPQT4+. A binding study between CBPQT4+
and the bulky stopper unit bearing the long glycol linker (see
Scheme S2 and Figure S7) was conducted. The free energy
obtained ΔG° = −1.65 kcal mol−1 (Ka = 16 M−1) for the
complexation in CD3CN at 298 K confirms that CBPQT4+ binds
a little to the oxyphenylene unit, and this finding could explain
the CBPQT4+ ring’s positioning in the switched rotaxanes.
Mechanistic Studies of the Switching Reactions Using
CV. Electrochemistry has been widely used to study switching
processes in mechanically interlocked systems.27,34,55−57 In
the CBPQT4+ ring to return again after reduction to the neutral
4+
TTF unit, and consequently, SEt−SRS is produced (Figure
3b). These control rotaxanes allow us to test if the tetracationic
CBPQT4+ ring alters its position in the switched state when the
neutral TTF unit is present (e.g., SEt−SR4+) and therefore
represents the situation when electrostatic repulsions from
TTF2+ dication can be excluded. The rotaxanes were synthesized
in a similar fashion to the title compounds (see Supporting
Information).53
The 1H NMR spectrum (Figure 4) of the oxidized SEt-based
rotaxane with the short linker (SEt−SRS6+) shows features
consistent with linear translation of the CBPQT4+ ring to one of
the O-linked oxyphenylene stations. This fact is made obvious
1
from the H NMR spectrum (Figure 4b), which is almost
identical to the one observed for the similar SMe-based rotaxane
Figure 4. Partial 1H NMR spectra (400 MHz, 298 K, CD3CN, 2.0 mM) of (a) SEt−RS4+, (b) SEt−SRS6+, (c) SRS6+, and (d) SEt−SRS4+. The switched
rotaxanes (SEt−SRS6+ and SRS6+) are generated by adding 10 equiv of the oxidant TBPASbCl6. SEt−SRS4+ is generated by adding Zn powder to SEt−
SRS6+. The signals being crossed out arise from traces of DMF present in the samples after the synthesis.
6377
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
particular, CV is an electrochemical method that allows for the
addition and removal of electrons by applying specific potentials
to the working electrode. Current peaks (Ip) in the CV response
at specific voltages (V) correspond to oxidation and reduction
events. All peak positions for anodic oxidations and cathodic
reductions, Epa and Epc, respectively, will be distinguished as
peaks, and all half wave potentials, (i.e., E1/2 = (Epa + Epc)/2) will
be called waves. Some of the oxidation events that are observed in
the forward sweep (0 → +1 V versus Ag/AgCl) also serve as
triggers to stimulate translational movements in bistable
rotaxanes, while the reduction peaks during the return sweep
(+1 → 0 V) remove the stimulus. The time between addition and
removal of the stimulus is available for the ring’s movement.
Changing the sweep rate of the CV can easily vary this time
period. The extent of molecular switching can then be analyzed
by using the return CV sweep on account of the fact that peak
heights, Ip, are proportional to concentration according to the
Randles−Sevcik relation.58 For this reason, the return sweeps are
often used to analyze the products of redox-driven switching.
The redox responses of the dumbbells were studied as
controls. The short dumbbell, DS (for its structure, see Scheme
S1), showed two redox waves (E1/2, as above) at +0.46 and +0.75
V (Figure 5a) corresponding to the first and second oxidation of
TTF, respectively, which is consistent with previous studies.18
The redox potentials of both the initial and switched forms of
each rotaxane, e.g., R4+ and SR6+, were measured by utilizing fast
scan rates (>10 V s−1) in order to minimize any switching. Under
these conditions, each rotaxane had similar redox potentials
(Table 1) irrespective of the linker’s length. Variable scan rates
were used to temporally resolve the switching speeds. Each of the
CV responses show that all the rotaxanes follow the same well
documented mechanism of switching according to the sequence,
oxidation-move-oxidation (ox-move-ox, Scheme 1) for the
forward switching from R4+ to SR6+. The redox response of the
short rotaxane RS4+ is described in depth and is representative of
the features seen with all four rotaxanes
Switching Rates in the Short RotaxaneEscape in the
4+
Forward Direction. Rotaxane RS displayed a one-electron
oxidation peak in the forward scan of the CV at +0.86 V (Figure
5b, red trace), which is ∼400 mV more positive than the first
oxidation of the dumbbell DS. This observation is consistent with
prior studies18,59 and follows from the fact that more energy is
required to oxidize the TTF unit when it is encircled by the
tetracationic ring. Scanning further out to ∼+1.4 V at fast scan
rates (10 V s−1) and at 273 K reveals (Figure S12) the second
5+
6+
oxidation of the rotaxane RS to RS that occurs prior to the
ring’s movement away from the monocationic TTF+ unit. In a
separate experiment, holding the potential above +1 V for a
sufficient period of time (10 s) ensures generation of the
switched rotaxane, SR6+, at the electrode surface within the
diffusion layer. Performing a fast sweep back to 0 V allows the
redox potentials of the switched rotaxane to be identified (Figure
5c, red trace). The switched rotaxane shows two redox waves in
this forward sweep at +0.42 and +0.65 V that have very similar
potentials to the redox waves of the free dumbbell (Figure 5a),
indicating the CBPQT4+ ring no longer encircles the TTF unit.
These data are in line with the interpretations from the chemical
oxidation studies described above.
Figure 5. Representative CVs illustrating that motion occurs by
comparing the (a) dumbbell DS (0.25 mM, 0.1 M TBAPF6, 2:1
CH2Cl2:MeCN, CH2Cl2 added for solubility), (b) rotaxane RS4+ (2 mM,
0.1 M TBAPF6, MeCN, 273 K) starting with R4+ showing changes in
relative peak positions (arrows) with decreasing scan rate (red trace: 10
4+
V s−1 − blue trace: 0.2 V s−1), and (c) rotaxane RS (2 mM, 0.1 M
TBAPF6, MeCN, 298 K) starting with the switched rotaxane SR6+
showing changes in relative peak positions (arrows) with decreasing
scan rate (red: 10 V s−1 − blue: 0.25 V s−1). Glassy carbon working
electrode.
electron peak in the forward sweep at +0.86 V is observed to
increase in magnitude as the second oxidation takes place. These
increases arise from the ox-move-ox sequence (Scheme 1). First,
RS4+ is oxidized by one electron to RS5+, which initiates the ring’s
movement to the oxyphenylene station, generating SRS5+. After
the CBPQT4+ ring has moved, the potential at the working
electrode provides a thermodynamically favored environment to
To characterize the kinetics of the forward switching reaction,
the potential was swept from 0 to +1.1 V and back again using
different scan rates (Figure 5b). Slower scan rates allow more
time for the rotaxanes to switch. Starting from 10 V s−1 (red trace,
Figure 5b) and slowing down the scan rate, the initial one-
6378
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
a
Table 1. Electrochemical Data for the Initial and Switched Rotaxanes
E1/2 [V]
Epa [V]
4+
4+
4+
4+
4+
4+
4+
4+
SRS
SRM
SRL
SRXL
RS
RM
RL
RXL
b
b
b
b
Ox1
Ox2
+0.48
+0.71
+0.47
+0.70
+0.45
+0.68
+0.49
+0.71
+0.86
+0.84
+0.83
+0.80
a
Conditions for the CV: 0.1 M TBAPF6, MeCN, c = 2.0 mM, glassy carbon working electrode, Pt counter electrode, scan rate >10 V s−1, E1/2 and Epa
b
values vs Ag/AgCl, errors 0.01 V. Anodic oxidation peak.
5+
immediately oxidize SRS to SRS6+. During the subsequent
chemical switching sequence (Scheme 1 and Supporting
Information). The rate constants (Table 2) for the movement
of the CBPQT4+ ring away from the TTF+ unit (R5+ → SR5+)
were optimized to fit the scan-rate-dependent shapes of the CVs;
compare experimental and simulated CVs in Figure S13.
analysis sweeps (from +1.1 to 0 V), the temporal resolution of
the movement away from the TTF+ in the ox-move-ox sequence
is also made obvious. The peak at +0.75 V, which is assigned to
the reduction of the unswitched rotaxane RS5+ (red trace, Figure
5b), is most intense using faster scan rates and is then observed to
decrease in size as the scan rate slows down. Concomitantly, the
switched rotaxane, SRS6+, is formed, and its associated peaks at
+0.65 and +0.42 V (blue trace, Figure 5b) that correspond to
sequential reduction of SRS6+ and SRS5+, are observed to increase
in intensity. The voltage window scanned from the initial
oxidation at +0.86 V out to the return voltage (+1.1 V) and back
to the first reduction at ∼+0.7 V amounts to a 0.65 V span, which
4+
Surprisingly, when the linker was doubled in length, RS and
RL4+, the time taken for the CBPQT4+ ring to escape from the
TTF+ monocation over the barrier did not increase but instead
decreased by a factor of 3 with t1/2 = 60 and 21 ms, respectively.
This result is in contrast with previous studies26,37 where longer
linkers slowed the overall rates of reaction. The CV data recorded
4+
on the rotaxane with extra long ethylene glycol linkers (RXL
)
are consistent with the same sequence of steps, however, the rates
of switching were outside the range of the present experimental
setup. The return switching reaction (Table 3) showed longer
linkers slowing the movements more dramatically from t1/2 = 1.4
to 17 s (293 K); a factor of 10 or more when the linker’s length is
doubled.
can give rise to a 65 ms (10 V s−1) to 3.25 s (0.2 V s−1) time
window to resolve the escape motion associated with RS
5+
→
5+
SRS
.
Switching Rates in the Short Rotaxane: Return Move-
ment. The return switching can be characterized in a similar
manner (Figure 5c). Initially, the potential is held above +1 V for
10 s to generate SRS6+ in the diffusion layer, and thereafter, the
potential is swept to 0 V and back again at various scan rates.
Upon scanning from +1 to 0 V, reduction to SRS4+ takes place at
+0.42 V. After it is generated, CBPQT4+ moves to the neutral
TTF to generate RS4+ such that the return sweep in the CV (0 →
+1 V) represents a reduction-reduction-move sequence (Scheme
1). During this sweep the relative amounts of the metastable
Escape in the Forward Direction: Activation Enthalpy
and Entropy. Motivated by the unexpected rate enhancement
seen for escape in the forward direction, variable-temperature
studies of the three rotaxanes (RS4+, RM4+, and RL4+) were
conducted to determine the activation parameters in a bid to
shed light on the nature of the RLS. Variable-temperature data
(263−303 K) for the forward switching was used in an Eyring
analysis (Figure 6a). Even though the rates of the forward
4+
4+
4+
switched rotaxane SRS and the product rotaxane RS can be
analyzed to capture the departure and arrival of the ring. Slow
scan rates (blue trace, Figure 5c) allow RS4+ to be fully formed. At
reaction across the three rotaxanes (RS4+, RM4+, and RL
)
increased with the linker’s length, the least-squares regression of
the Eyring plots produced activation parameters (ΔH⧧, ΔS⧧)
that were the same to within error (Table 2) and dominated by
an enthalpy barrier we attribute to sterics.
4+
fast scan rates, SRS is still present in the diffusion layer. This
metastable species can be identified from the peak at +0.49 V
(red trace, Figure 5c), which has a very similar potential to the
oxidation peak for the dumbbell (Figure 5a) indicating that the
TTF unit is still unencircled. Based on the voltage window
The negative activation entropies of approximately −4 cal K−1
mol−1, suggest there is a modest increase in the ordering of the
transition state structure in R5+⧧ relative to the initial structure of
R5+. The fact that this ordering is so small is consistent with an
early transition state. The structure proposed for the
scanned, a time window of 90 ms (10 V s−1) to 3.6 s (0.25 V s−1)
was available for resolving the return movement, SRS4+ → RS
,
4+
5+
of the ring from the oxyphenylene to the TTF station.
intermediate state RS (Scheme 2) is based on the idea that
Rates of Motion as a Function of Linker Length. The
overall sequence for switching forward and backward (Scheme 1)
was used as a basis to simulate and reproduce the shapes of the
CVs recorded at different scan rates (see Figures S13−S24).
Close correlations between the simulated and experimental data
enabled the kinetics parameters for each of the molecular
movements to be determined. The observed rate of escape
(k293obs) in the forward direction toward the oxyphenylene
stations has two equal probability pathways, to the left and to the
right. Consequently, the observed rate constants were halved (k)
in order to provide estimates of the microscopic activation
parameters (ΔG⧧, ΔH⧧, ΔS⧧) associated with overcoming just
one barrier at a time.
electrostatic repulsions from the monocationic TTF+ unit repel
the tetracationic CBPQT4+ ring. Consequently, the CBPQT4+
ring is expected to move toward the neutral half of the TTF+ unit
where it is forced up against the SMe and −S−CH2−CH2−O−
chains, which together act as steric blocking groups. In this
situation, with CBPQT4+ abutted against the steric barrier, it is
conceivable that the two chains might already be preorganized in
a favorable manner to help mitigate the entropic contribution to
the barrier, see the RLS in Scheme 2 for models of the
5+‡
intermediate, RS5+, and the transition state, RS
.
The fact that all three rotaxanes (RS4+, RM4+, and RL4+, Table 2)
show similar sets of activation parameters (ΔG⧧ ∼ +15 kcal
mol−1, ΔH⧧ ∼ +14 kcal mol−1, ΔS⧧ ∼ −4 cal K−1 mol−1)
indicates that the RLSs to escape have shared qualities.
Nevertheless, there is a measurable impact of the linker’s length
on the rate constants of escape indicating their involvement. This
effect was enhanced further with the extra long rotaxane (RXL4+),
The four different rotaxanes displayed similar E1/2 values
(Table 1), indicating that the linkers do not have an effect on
4+
their thermodynamics. CVs of three of the rotaxanes (RS4+, RM
,
and RL4+) could be readily simulated using the same electro-
6379
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
Table 2. Kinetics and Associated Activation Parameters for the Forward Movement (R5+ → SR5+) of the CBPQT4+ Ring From
ab c
, ,
TTF+ Over the SMe Barrier to the Oxyphenylene Station
k293obs [s−1
]
k293 [s−1
11.5 1.2
20
]
t1/2 [ms]
ΔG293⧧ [kcal mol−1
]
ΔH⧧ [kcal mol−1
]
ΔS⧧ [cal K−1 mol−1
]
short
23 1.2
60
35
21
<7
15.7 0.1
15.4 0.1
15.2 0.1
>14.5
14.5 1.0
13.8 0.5
14.5 0.5
−4.1 3.6
−5.5 1.6
−2.3 1.8
medium
long
40
2
2
65 3.5
32.5 3.5
>100
extra long
>200
a
b
obs
Conditions for the CV: 0.1 M TBAPF6, MeCN, 2.0 mM, glassy carbon working electrode, Pt counter electrode, 263−303 K, 10−0.2 V s−1. k293
is the observed rate of escape from the TTF+ station over either barrier; k293, t1/2, ΔG293⧧, ΔH⧧, and ΔS⧧ are the values for microscopic activation
parameters. Values for k293 and t1/2 are determined as follows: k293 = 0.5 k293 and t1/2 = ln(2)k293
c
obs
−1
.
Table 3. Kinetics and Associated Activation Parameters for the Return Movement (SR4+ → R4+) of the CBPQT4+ Ring from the
a
Oxyphenylene Station Back Over the SMe Barrier to the TTF Station
k293 [s−1
]
t1/2 [s]
ΔG293⧧ [kcal mol−1
]
ΔH⧧ [kcal mol−1
]
ΔS⧧ [cal K−1 mol−1
]
short
0.50 0.08
0.10 0.03
0.04 0.01
<0.01
1.4
17.6 0.1
18.5 0.2
19.0 0.2
>19.8
6.9 0.9
−37
−20
−6.6 2.5
3
2
medium
long
6.9
12
17
1
1
17.3
>69
extra long
a
Conditions for the CV: 0.1 M TBAPF6, MeCN, 2.0 mM, glassy carbon working electrode, Pt counter electrode, 293−333 K, 10−0.2 V s−1.
which takes <7 ms, confirming the veracity of this trend. The
most likely explanation for the increased rates for the forward
reaction is the increasing involvement of the longer glycolic
linkers in stabilizing the CBPQT4+ ring41 when it passes over the
sterically-based transition state (RL5+⧧, Scheme 3).
molecules tend to be dominated by escape from the initial
station. Removal of the SMe groups might shift the RLS to be
one that depends on the ring’s movement along the linker.
Furthermore, use of a secondary station that is redox-active
would allow the time of the ring’s arrival to be measured.
Return Switching and Intermediates on Pathway
Between Stations. An Eyring analysis of the return switching
(Figure 6b, 293−333 K), for which the departure and arrival
times are measured, shows that longer linkers slow the rate
(Table 3) of motion considerably and that the nature of the RLS
also changes. The movement of the CBPQT4+ ring from the
oxyphenylene station back to the neutral TTF unit in the
rotaxane with a short linker (t1/2 = 1.4 s, ΔG⧧ = +17.6 kcal mol−1)
has a small enthalpy of activation, ΔH⧧ = +6.9 0.9 kcal mol−1,
After passing over the steric barrier, the CBPQT4+ ring moves
to the oxyphenylene station at speeds that cannot be
distinguished using this current series of rotaxanes. This idea is
consistent with Leigh’s study23 that RLSs in interlocked
while the activation entropy is large and negative, ΔS⧧ = −37
3
cal K−1 mol−1. The magnitude of this entropic factor is consistent
with the need for a significant increase in the ordering of both the
SMe thioether and the glycol linker to allow passage of the
CBPQT4+ ring (SRS4+⧧, Scheme 2).
By contrast, the long rotaxane shows a barrier (t1/2 = 17.3 s,
ΔG⧧ = +19 kcal mol−1) that is now enthalpic in nature (ΔH⧧ =
+17 1 kcal mol−1) with a much-reduced entropic contribution
(ΔS⧧ = −6.6 2.5 cal K−1 mol−1). The large activation enthalpy
supports the idea that the longer rotaxane needs to break
additional stabilizing interactions between the CBPQT4+ ring
and glycol chain before moving over the steric barrier. The fact
that glycol chains can bind to CBPQT4+ is confirmed in the
literature40,41,60 and is typically observed as CH···O hydrogen-
bond interactions between the α-CH hydrogens atoms in the
bipyridinium units of CBPQT4+ and the oxygen atoms present in
the glycol chains. If the interactions are in fact present as CH···O
hydrogen bonds, they are likely to be highly ordered (Scheme 3,
SRL4+⧧). Therefore, breaking these contacts should be an
entropically favorable process, as the glycols would then be
free to move. The configurational entropy is associated with
increased freedom of the glycol chains on going from the small to
medium (13 cal K−1 mol−1) and small to long rotaxanes (30 cal
K−1 mol−1), corresponding to extensions by ethylene and
diethylene glycol, respectively. These values are consistent with
the loss in entropy (∼6 cal K−1 mol−1, Me2CO) observed for the
Figure 6. Eying plots for the (a) forward and (b) reverse movement of
4+
CBPQT4+ along rotaxanes RS4+, RM4+, and RL as determined using
kinetic parameters from the CV simulations.
6380
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
a
4+
Scheme 2. Proposed Mechanism for the Switching Cycle in the Short Rotaxane RS
a
The RLS is indicated for both the forward and return directions of motion.65
binding of CBPQT4+ to monopyrrolo-TTF when a triethylene
glycol unit was wrapped around the bipyridinium.41
The structure of the switched rotaxane SRL (Scheme 4)
be broken in order for the CBPQT4+ to move, and they all
contribute to the large enthalpic barrier. The unfolding of the
structure to allow movement of the CBPQT4+ ring would also be
entropically favorable. The same folded structure must be
present in the extra long rotaxane to account for the fact that its
switching is even slower. In the shortest rotaxane SRS4+ (Scheme
2), it has fewer CH···O hydrogen bonds and is less competent at
forming the alongside charge-transfer interactions; consequently,
4+
would likely involve a folded conformation where the neutral
TTF unit is contacting one side of the CBPQT4+ ring to form a
weak charge-transfer interaction, which would also impact the
ΔH⧧ and ΔS⧧. Evidence for these alongside effects is growing in
the literature.61−64 The movement of the CBPQT4+ ring from
the oxyphenylene back to the neutral TTF unit is too fast to
it has less bonds to break and has less entropy offset during its
4+
4+
isolate SRL for analysis, consequently, evidence for the
switching back to RS
.
alongside interaction was achieved using the switched form of
the blocked long rotaxane SEt−SRL4+ (Figure 3). Starting from
SEt−RL4+, oxidation followed by reduction, purification, and ion
exchange (see Supporting Information) afforded a sample for
analysis. A UV−vis−NIR spectrum of SEt−SRL4+ (MeCN, 280
K, Figure 7) was measured, and a weak charge-transfer
absorption was located at 625 nm64 that would be expected
from a folded conformation (Scheme 4). We believe this
structure will also be present in the metastable intermediate
Impacts of Linker Length on Time-Resolved Move-
ments Forward and Backward. In summary (Figure 8), the
escape from the primary station gets faster with longer linkers,
while the return movement from the secondary to the primary
station slows down. Prior to initiating the return movement, the
switched rotaxane with the long linker, SRL4+, has the possibility
to fold up into a more compact and thermodynamically stable
structure by making noncovalent interactions (Scheme 3). These
contacts contribute to the high enthalpy (ΔH⧧) barrier of
activation and the freedom gained upon unfolding leads to a
smaller entropic (ΔS⧧) barrier in the reverse switching of the
SRL prior to movement of the CBPQT4+ ring back to TTF
4+
station.
4+
4+
The presence of a folded structure in SRL requires the
long rotaxane, RL4+. Conversely the short rotaxane, RS has a
charge-transfer interactions between the TTF group and the
CBPQT4+ ring to be broken before it is free to move (Scheme 3).
Furthermore, it is known from the binding study (vide supra) that
there is a stable interaction between the CBPQT4+ ring and the
oxyphenylene station, which may be governed by C−H···π
interactions between the encircled oxyphenylene’s protons and
high entropy (ΔS⧧) factor because the increased ordering in the
substituent (the SMe thioether and the short glycol linker) that is
required for CBPQT4+ to move back to the neutral TTF
(Scheme 2) is not offset as it is with RL4+. Across the series, the
linker’s length has a minor effect on the forward escape and a
large impact on the return movement. For the short rotaxane, the
forward and return movements shunt the ring between stations.
Whereas, the extra long rotaxane’s movements reflect a shove
1
the p-xylyl rings of CBPQT4+ as was shown in the H NMR
spectra (Figure 2). All of these noncovalent interactions have to
6381
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
a
4+
Scheme 3. Proposed Mechanism for the Switching Cycle in the Long Rotaxane RL
a
The RLS is indicated for both the forward and return directions of motion.65
Scheme 4. Equilibrium Between the Folded and Unfolded
a
4+
Structure of SEt−SRL
Figure 7. UV−vis−NIR absorption spectrum of SEt−SRL4+ recorded in
MeCN (0.2 mM) at 280 K.
a
The structure shows the possibility of forming CH···O and charge-
transfer interactions in the folded alongside conformation (right) that
brings the oxyphenylene-localized CBPQT4+ into contact with the
neutral TTF unit.
over the steric barrier to access the secondary station and
generate a folded coconformation that, after reduction to the
neutral TTF unit (TTF0), must unfurl at some point during
passage of the CBPQT4+ ring back to the starting point.
6382
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
Figure 8. Plots showing how the rate constants of motion change as a function of linker length for (a) escape forward and (b) return movements.
CONCLUSION
AUTHOR INFORMATION
■
■
Corresponding Authors
joj@sdu.dk
aflood@indiana.edu
In this study of the complete cycle of motion in redox-driven
TTF-CBPQT4+ rotaxanes, we have shown that the linkers that
exist between two stations play a critical role in both the kinetics
and the mechanism of motion. In these rotaxanes, the RLS for the
forward movement of the CBPQT4+ ring involves electrostatic
preparation of an intermediate that helps organize passage
through an early transition state defined by a steric speed bump
(SMe). Subsequently, translation of the ring to the oxyphenylene
station occurs but was not measured in these experiments. In the
return movement from the oxyphenylene to the charge-neutral
TTF, it appears that the RLS is determined by the stability and
conformation of the mestastable rotaxane. The rotaxane with the
longer glycol linkers is more flexible and thereby allows folding
into a compact stable structure in which the neutral TTF unit is
located alongside the CBPQT4+ ring. Therefore, additional
noncovalent interactions have to be broken concomitant with
attainment of greater conformational freedom during the
movement of the CBPQT4+ ring back to the neutral TTF unit.
In contrast, the rotaxane with the shorter linker makes fewer
noncovalent interactions, and the shorter glycol linker has to
dramatically increase its ordering to allow CBPQT4+ to pass back
to the neutral TTF station. This study shows how the linkers’
lengths can impact switching rates in ways that differ from and
which complement the insights from Leigh’s study; there, escape
was fast enough to allow random walks to impact rates of
movements, and here, the linkers participated in the thermally
activated escape. Overall, these insights contribute design
guidance for the generation of machines and motors from the
switching progenitors.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Support is acknowledged from the Villum Foundation (to J.O.J.),
the Danish Natural Science Research Council (FNU, Project 11-
106744 to J.O.J.), and the NSF (CHE 0844441 to A.H.F.).
REFERENCES
■
(1) Kinbara, K.; Aida, T. Chem. Rev. 2005, 105, 1377.
(2) Coskun, A.; Banaszak, M.; Astumian, R. D.; Stoddart, J. F.;
Grzybowski, B. A. Chem. Soc. Rev. 2012, 41, 19.
(3) Kay, E. R.; Leigh, D. A.; Zerbetto, F. Angew. Chem., Int. Ed. 2007,
46, 72.
(4) Benson, C. R.; Share, A. I.; Flood, A. H. In Bioinspiration and
Biomimicry in Chemistry; John Wiley & Sons, Inc.: New York, 2012.
(5) Fahrenbach, A. C.; Bruns, C. J.; Cao, D.; Stoddart, J. F. Acc. Chem.
Res. 2012, 45, 1581.
(6) Jeppesen, J. O.; Perkins, J.; Becher, J.; Stoddart, J. F. Angew. Chem.,
Int. Ed. 2001, 40, 1216.
(7) Keaveney, C. M.; Leigh, D. A. Angew. Chem., Int. Ed. 2004, 43, 1222.
(8) Schmidt, T.; Vogtle, F. Sci. Spectra 1999, 58.
̈
(9) Stoddart, J. F. Chem. Soc. Rev. 2009, 38, 1802.
(10) Andersen, S. S.; Jensen, M.; Sorensen, A.; Miyazaki, E.; Takimiya,
K.; Laursen, B. W.; Flood, A. H.; Jeppesen, J. O. Chem. Commun. 2012,
48, 5157.
(11) Choi, J. W.; Flood, A. H.; Steuerman, D. W.; Nygaard, S.;
Braunschweig, A. B.; Moonen, N. N. P.; Laursen, B. W.; Luo, Y.;
DeIonno, E.; Peters, A. J.; Jeppesen, J. O.; Xu, K.; Stoddart, J. F.; Heath,
J. R. Chem.Eur. J. 2006, 12, 261.
(12) Hamacek, J.; Borkovec, M.; Piguet, C. Dalton Trans. 2006, 1473.
ASSOCIATED CONTENT
■
́
(13) Moonen, N. N. P.; Flood, A.; Fernandez, J.; Stoddart, J. F. In
Molecular Machines; Kelly, T. R., Ed.; Springer: Berlin Heidelberg, 2005;
Vol. 262.
(14) Share, A. I.; Parimal, K.; Flood, A. H. J. Am. Chem. Soc. 2010, 132,
1665.
(15) Astumian, R. D. Annu. Rev. Biophys. 2011, 40, 289.
S
* Supporting Information
Syntheses, chemical switching, binding studies, variable scan rate,
and variable-temperature CV data, and their analysis. This
material is available free of charge via the Internet at http://pubs.
acs.org.
(16) Astumian, R. D. J. Phys. Condens. Mater. 2005, 17, S3753.
6383
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384

Journal of the American Chemical Society
Article
(17) Serreli, V.; Chin-Fa, L.; Kay, E. R.; Leigh, D. A. Nature 2007, 445,
523.
(18) Flood, A. H.; Peters, A. J.; Vignon, S. A.; Steuerman, D. W.; Tseng,
H.-R.; Kang, S.; Heath, J. R.; Stoddart, J. F. Chem.Eur. J. 2004, 10,
6558.
(46) Anelli, P. L.; Ashton, P. R.; Ballardini, R.; Balzani, V.; Delgado, M.;
Gandolfi, M. T.; Goodnow, T. T.; Kaifer, A. E.; Philp, D.; Pietrazkiwicz,
M.; Prodi, L.; Reddington, M. V.; Slawin, A. M. Z.; Spencer, N.;
Stoddart, J. F.; Vicent, C.; Williams, D. J. J. Am. Chem. Soc. 1992, 114,
193.
(47) Material and methods are available in Supporting Information
(48) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877.
(49) Jeppesen, J. O.; Nygaard, S.; Vignon, Scott A.; Stoddart, J. F. Eur. J.
Org. Chem. 2005, 2005, 196.
(19) Koumura, N.; Geertsema, E. M.; Gelder, M. B. v.; Meetsma, A.;
Feringa, B. L. J. Am. Chem. Soc. 2002, 124, 5037.
(20) Fletcher, S. P.; Dumur, F.; Pollard, M. M.; Feringa, B. L. Science
2005, 310, 80.
(21) Leigh, D. A.; Wong, J. K. Y.; Dehez, F.; Zerbetto, F. Nature 2003,
424, 174.
(50) A comparison of the 1H NMR spectra of the dumbbell DS and the
oxidized dumbbell DS2+ shows that the SMe groups attached to the TTF
unit experience a downfield shift (Δδ = 0.35 ppm) in their resonances
upon formation of TTF2+.
́
(22) Hernandez, J. V.; Kay, E. R.; Leigh, D. A. Science 2004, 306, 1532.
(23) Baggerman, J.; Haraszkiewicz, N.; Wiering, P. G.; Fioravanti, G.;
Marcaccio, M.; Paolucci, F.; Kay, E. R.; Leigh, D. A.; Brouwer, A. M.
Chem.Eur. J. 2013, 19, 5566.
(51) Tseng, H.-R.; Vignon, S. A.; Stoddart, J. F. Angew. Chem., Int. Ed.
2003, 42, 1491.
(52) Tseng, H.-R.; Vignon, S. A.; Celestre, P. C.; Perkins, J.; Jeppesen,
J. O.; Fabio, A. D.; Ballardini, R.; Gandolfi, M. T.; Venturi, M.; Balzani,
V.; Stoddart, J. F. Chem.Eur. J. 2004, 10, 155.
(24) Avellini, T.; Li, H.; Coskun, A.; Barin, G.; Trabolsi, A.; Basuray, A.
N.; Dey, S. K.; Credi, A.; Silvi, S.; Stoddart, J. F.; Venturi, M. Angew.
Chem., Int. Ed. 2012, 51, 1611.
(53) Detailed description of the synthesis can be found in Supporting
Information
(25) Nygaard, S.; Laursen, B. W.; Flood, A. H.; Hansen, C. N.;
Jeppesen, J. O.; Stoddart, J. F. Chem. Commun. 2006, 2, 144.
(26) Panman, M. R.; Bodis, P.; Shaw, D. J.; Bakker, B. H.; Newton, A.
C.; Kay, E. R.; Brouwer, A. M.; Buma, W. J.; Leigh, D. A.; Woutersen, S.
Science 2010, 328, 1255.
(27) Trabolsi, A.; Fahrenbach, A. C.; Dey, S. K.; Share, A. I.; Friedman,
D. C.; Basu, S.; Gasa, T. B.; Khashab, N. M.; Saha, S.; Aprahamian, I.;
Khatib, H. A.; Flood, A. H.; Stoddart, J. F. Chem. Commun. 2010, 46,
871.
(28) Zhang, K. D.; Zhao, X.; Wang, G. T.; Liu, Y.; Zhang, Y.; Lu, H. J.;
Jiang, X. K.; Li, Z. T. Angew. Chem., Int. Ed. 2011, 50, 9866.
(29) Durola, F.; Sauvage, J.-P. Angew. Chem., Int. Ed. 2007, 46, 3537.
(30) Spruell, J. M.; Paxton, W. F.; Olsen, J.-C.; Benitez, D.; Tkatchouk,
E.; Stern, C. L.; Trabolsi, A.; Friedman, D. C.; G, W. A., III; Stoddart, J.
F. J. Am. Chem. Soc. 2009, 131, 11571.
(54) Nygaard, S.; Hansen, S. W.; Huffman, J. C.; Jensen, F.; Flood, A.
H.; Jeppesen, J. O. J. Am. Chem. Soc. 2007, 129, 7354.
(55) Fahrenbach, A. C.; Barnes, J. C.; Li, H.; Benítez, D.; Basuray, A.
N.; Fang, L.; Sue, C.-H.; Barin, G.; Dey, S. K.; Goddard, W. A.; Stoddart,
J. F. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 20416.
(56) Nijhuis, C. A.; Ravoo, B. J.; Huskens, J.; Reinhoudt, D. N. Coord.
Chem. Rev. 2007, 251, 1761.
(57) Flood, A. H.; Kaifer, A. E. In Supramolecular Chemistry: From
Molecules to Nanomaterial; John Wiley and Sons, Inc.: New York, 2012;
p 451.
(58) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals
and Applications, 2nd ed.; John Wiley and Sons, Inc.: New York, 2001.
(59) Nygaard, S.; Laursen, B. W.; Hansen, T. S.; Bond, A. D.; Flood, A.
H.; Jeppesen, J. O. Angew. Chem., Int. Ed. 2007, 46, 6093.
(60) Wang, C.; Cao, D.; Fahrenbach, A. C.; Grunder, S.; Dey, S. K.;
Sarjeant, A.; Stoddart, J. F. Chem. Commun. 2012, 48, 9245.
(31) Mobian, P.; Kern, J.-M.; Sauvage, J.-P. Angew. Chem., Int. Ed. 2004,
43, 2392.
(32) Jeppesen, J. O.; Vignon, S. A.; Stoddart, J. F. Chem.Eur. J. 2003,
9, 4611.
̌
́
(61) Menzer, S.; White, A. J. P.; Williams, D. J.; Belohradsky, M.;
Hamers, C.; Raymo, F. M.; Shipway, A. N.; Stoddart, J. F.
Macromolecules 1998, 31, 295.
(33) Baroncini, M.; Silvi, S.; Venturi, M.; Credi, A. Chem.Eur. J.
2010, 16, 11580.
(62) Jeppesen, J. O.; Nielsen, K. A.; Perkins, J.; Vignon, S. A.; Fabio, A.
D.; Ballardini, R.; Gandolfi, M. T.; Venturi, M.; Balzani, V.; Becher, J.;
Stoddart, J. F. Chem.Eur. J. 2003, 9, 2982.
(34) Trabolsi, A.; Khashab, N.; Fahrenbach, A. C.; Friedman, D. C.;
Colvin, M. T.; Cotí, K. K.; Benítez, D.; Tkatchouk, E.; Olsen, J.-C.;
Belowich, M. E.; Carmielli, R.; Khatib, H. A.; Goddard, W. A.;
Wasielewski, M. R.; Stoddart, J. F. Nat. Chem. 2010, 2, 42.
(35) Hirose, K.; Shiba, Y.; Ishibashi, K.; Doi, Y.; Tobe, Y. Chem.Eur.
J. 2008, 14, 3427.
(63) Saha, S.; Flood, A. H.; Stoddart, J. F.; Impellizzeri, S.; Silvi, S.;
Venturi, M.; Credi, A. J. Am. Chem. Soc. 2007, 129, 12159.
(64) Hansen, S. W.; Stein, P. C.; Sørensen, A.; Share, A. I.; Witlicki, E.
H.; Kongsted, J.; Flood, A. H.; Jeppesen, J. O. J. Am. Chem. Soc. 2012,
134, 3857.
(36) Sugino, H.; Kawai, H.; Umehara, T.; Fujiwara, K.; Suzuki, T.
(65) For simplicity, a positive charge has been placed on one of the
dithiole rings and a radical on the other dithiole ring in the radical cation
states of the oxidized TTF unit throughout the scheme. However, in
general, it is believed that the TTF radical cation is fully conjugated, and
consequently both 1,3-dithiole rings are equivalent; therefore, one
cannot distinguish the “radical half” or “cation half”.
Chem.Eur. J. 2012, 18, 13722.
(37) Deutman, A. B. C.; Monnereau, C.; Elemans, J. A. A. W.; Ercolani,
G.; Nolte, R. J. M.; Rowan, A. E. Science 2008, 322, 1668.
(38) Li, H.; Zhu, Z.; Fahrenbach, A. C.; Savoie, B. M.; Ke, C.; Barnes, J.
C.; Lei, J.; Zhao, Y.-L.; Lilley, L. M.; Marks, T. J.; Ratner, M. A.; Stoddart,
J. F. J. Am. Chem. Soc. 2012, 135, 456.
(39) Panman, M. R.; Bodis, P.; Shaw, D. J.; Bakker, B. H.; Newton, A.
C.; Kay, E. R.; Leigh, D. A.; Buma, W. J.; Brouwer, A. M.; Woutersen, S.
Phys. Chem. Chem. Phys. 2012, 14, 1865.
(40) Coskun, A.; Spruell, J. M.; Barin, G.; Dichtel, W. R.; Flood, A. H.;
Botros, Y. Y.; Stoddart, J. F. Chem. Soc. Rev. 2012, 41, 4827.
(41) Nygaard, S.; Hansen, C. N.; Jeppesen, J. O. J. Org. Chem. 2007, 72,
1617.
(42) Raymo, F. M.; Houk, K. N.; Stoddart, J. F. J. Am. Chem. Soc. 1998,
120, 9318.
(43) Asakawa, M.; Ashton, P. R.; Balzani, V.; Credi, A.; Mattersteig, G.;
Matthews, O. A.; Montalti, M.; Spencer, N.; Stoddart, J. F.; Venturi, M.
Chem.Eur. J. 1997, 3, 1992.
(44) Lau, J.; Blanchard, P.; Riou, A.; Jubault, M.; Cava, M. P.; Becher, J.
J. Org. Chem. 1997, 62, 4936.
(45) Gibson, H. W.; Lee, S. H.; Engen, P. T.; Lecavalier, P.; Sze, J.;
Shen, Y. X.; Bheda, M. J. Org. Chem. 1993, 58, 3748.
6384
dx.doi.org/10.1021/ja5013596 | J. Am. Chem. Soc. 2014, 136, 6373−6384