Anion-induced shuttling of a naphthalimide triazolium rotaxane

Article abstract of DOI:10.1002/chem.201200317

The anion-templated synthesis of a rotaxane structure, incorporating the new naphthalimide triazolium motif, is described and the interlocked host shown to exhibit selective, uni-directional, anion-induced shuttling. Initial pseudorotaxane investigations demonstrate the ability of a naphthalimide triazolium threading component to form interpenetrated assemblies with counter-anion-dependent co-conformations. ¹H NMR studies reveal that the shuttling behaviour of the analogous rotaxane host system is controlled by selective anion binding and by the nature of the solvent conditions. Complete macrocycle translocation only occurs upon the recognition of the smaller halide anions (chloride and bromide). The rotaxane solid-state crystal structure in the presence of chloride is in agreement with the solution-phase co-conformation. The sensitivity of the axle naphthalimide absorbance band to the position of the macrocycle component within the interlocked structure enabled the molecular motion to be observed by UV/Vis spectroscopy, and the chloride-induced shuttling of the rotaxane was reversed upon silver hexafluorophosphate addition.

Full text of DOI:10.1002/chem.201200317

FULL PAPER
DOI: 10.1002/chem.201200317
Anion-Induced Shuttling of a Naphthalimide Triazolium Rotaxane
Graeme T. Spence,[a] Mateusz B. Pitak,[b] and Paul D. Beer*[a]
Abstract: The anion-templated synthe-
shuttling behaviour of the analogous
rotaxane host system is controlled by
selective anion binding and by the
nature of the solvent conditions. Com-
plete macrocycle translocation only
occurs upon the recognition of the
smaller halide anions (chloride and
bromide). The rotaxane solid-state
sis of a rotaxane structure, incorporat-
ing the new naphthalimide triazolium
motif, is described and the interlocked
host shown to exhibit selective, uni-di-
rectional, anion-induced shuttling. Ini-
tial pseudorotaxane investigations dem-
onstrate the ability of a naphthalimide
triazolium threading component to
form interpenetrated assemblies with
counter-anion-dependent co-conforma-
crystal structure in the presence of
chloride is in agreement with the solu-
tion-phase co-conformation. The sensi-
tivity of the axle naphthalimide absorb-
ance band to the position of the macro-
cycle component within the interlocked
structure enabled the molecular motion
to be observed by UV/Vis spectrosco-
py, and the chloride-induced shuttling
of the rotaxane was reversed upon
silver hexafluorophosphate addition.
Keywords: anions · molecular mo-
tion · molecular recognition · rotax-
anes · template synthesis
1
tions. H NMR studies reveal that the
Introduction
nalling method.[6,7] In these systems, co-conformational
changes were driven either by a change in the non-coordi-
nating counter anion,[6] or by the photo-induced creation of
a radical naphthalimide anion within the rotaxane struc-
ture.[7]
We report herein the synthesis of a rotaxane capable of
controlled, reversible shuttling induced by the addition of
coordinating anions (Figure 1). The interlocked structure
The ability of mechanically interlocked molecules to under-
go controlled, reversible molecular motion through changes
in the relative positions of their constituent parts is receiving
an ever increasing amount of interest due to the promise of
potential nano-technological applications as molecular
switches and machines.[1] However, despite the recent ad-
vances in anion supramolecular chemistry,[2] there are still
relatively few examples of such interlocked systems being
mediated by anions as an external stimulus;[3–6] and of these,
only a small number have been shown to display selectivity
between different coordinating anions.[4,5]
A key development in molecular machine-like devices is
the ability to detect molecular motion through optical or
electrochemical signalling. Of those systems induced by the
addition of coordinating anions, only Smithꢀs chloride-selec-
tive squaraine rotaxane display an associated signalling re-
sponse,[4] whereby an increase in fluorescence emission is
observed upon macrocycle translocation from the central
station to one of two triazole groups of the axle. In other,
more general types of anion-induced molecular motion, UV/
Vis spectroscopy has also been effectively employed as a sig-
Figure 1. Schematic representation of the anion-induced shuttling behav-
iour exhibited by a naphthalimide triazolium rotaxane.
contains the new naphthalimide triazolium motif within the
axle component, and was constructed by using an anion-
templation synthetic strategy.[8,9] Molecular motion is driven
by binding of an anion in the rotaxane-host cavity, as evi-
denced by 1H NMR spectroscopy. Importantly, selectivity
both in anion binding and the extent of uni-directional mac-
rocycle translocation was observed. In addition, the co-con-
formational changes exhibited by the rotaxane were found
to perturb the absorbance band of the naphthlamide triazo-
lium axle sufficiently for the shuttling process to be signalled
and monitored by changes in the rotaxane’s UV/Vis spec-
trum.
[a] G. T. Spence, Prof. P. D. Beer
Chemistry Research Laboratory, Department of Chemistry
University of Oxford, Mansfield Road, Oxford, OX1 3TA (UK)
Fax : (+44)1865-272690
E-mail: paul.beer@chem.ox.ac.uk
[b] Dr. M. B. Pitak
National Crystallography Service, School of Chemistry,
University of Southampton, Highfield Campus
Southampton, SO17 1BJ (UK)
Results and Discussion
Pseudorotaxane investigations: To probe the potential of
the naphthalimide triazolium motif to form interpenetrative
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201200317.
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

and interlocked assemblies capable of anion-induced, con-
trolled molecular motion, pseudorotaxane studies were car-
ried out initially with thread 1·BF4 (prepared as detailed in
the Supporting Information) and macrocycle 2[10] (Figure 2).
The naphthalimide motif is intrinsically electron-deficient
and further polarized in 1·BF4 by the direct attachment of
(Scheme 1 and Figure 3a and b). Pseudorotaxane formation
was indicated by a significant upfield shift of macrocycle hy-
droquinone protons g and h (Dd=À0.51 ppm), due to the
highly favourable p-donor–p-acceptor interactions with the
electron-deficient naphthalimide thread. In addition, hydro-
gen-bonding interactions between the imide carbonyl groups
of 1·BF4 and the macrocycle amide groups resulted in down-
field shifts in macrocycle protons c and d. These comple-
mentary donor–acceptor and hydrogen-bonding interactions
indicate that macrocycle 2 resides directly over the naphtha-
limide group of thread 1·BF4 in the pseudorotaxane assem-
bly, as shown in Scheme 1.
a positively charged triazolium group, which can bind anions
[9]
À
through C H hydrogen bonding. Macrocycle 2 contains an
À
À
isophthalamide anion binding cleft, containing C H and N
H hydrogen-bond donors, and electron-rich hydroquinone
groups.
Subsequent addition of one equivalent of tetrabutylam-
monium (TBA) chloride to 1·2·BF4 resulted in a significant
change in pseudorotaxane co-conformation, with the macro-
cycle moving over to the triazolium component of the
thread (Scheme 1). Cooperative hydrogen bonding to the
halide anion by the macrocycle isophthalamide cleft and
thread triazolium group is indicated by downfield shifts in
triazolium proton 8 and macrocycle protons c and d (Fig-
ure 3c). The hydroquinone protons shift downfield from
their position in 1·2·BF4 (Dd=0.16 ppm), due to the reduced
shielding effect of the triazolium group in this co-conforma-
tion, compared with the larger naphthalimide group, and are
split in the presence of the halide anion.
The respective interpenetrated co-conformational nature
of assemblies 1·2·BF4 and 1·2·Cl was confirmed by 2D
1H NMR ROESY spectroscopy. A large number of cou-
plings between protons of the macrocycle and the thread
are observed for both pseudorotaxane co-conformations,
with the spectra and couplings shown in the Supporting In-
formation (Figure S4–S5).
To quantify the stabilities of these two assemblies,
1H NMR pseudorotaxane titrations were carried out in
CDCl3. Both 1·BF4 and 1·Cl (prepared by anion exchange
detailed in Scheme S1 in the Supporting Information), were
titrated into solutions of macrocycle 2 and the signals corre-
sponding to the hydroquinone protons g and h monitored
throughout each titration (see Supporting Information, Fig-
Figure 2. Naphthalimide thread 1·X and isophthalamide macrocycle 2
used in pseudorotaxane investigations.
These components were designed to undergo interpene-
trative assembly due to strong p-donor–p-acceptor interac-
tions between the electron-deficient naphthalimide thread
and electron-rich hydroquinone groups of the macrocycle,
with cooperative hydrogen bonding to an anion also possible
through both the triazolium thread and macrocycle iso-
phthalamide binding motifs. Thus pseudorotaxane co-con-
formational changes between thread 1·BF4 and macrocycle 2
may be controlled by the presence, or absence, of a coordi-
nating anion.
The addition of one equivalent of 1·BF4 to macrocycle 2
ure S8).
Stability
constants
were
obtained
by
in CDCl3 was monitored by 1H NMR spectroscopy
WinEQNMR2[11] analysis, which gave values of 829(30) mÀ1
Scheme 1. Formation of pseudorotaxane 1·2·BF4 and anion-induced change in co-conformation.
&
2
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Anion-Induced Shuttling of a Rotaxane
FULL PAPER
Figure 4. UV/Vis spectra in CHCl3 at 293 K of thread 1·BF4 (1.0ꢂ10À3 m)
upon addition of 0.0, 1.0, 2.0, 3.0, 4.0 and 5.0 equivalents of macrocycle 2.
sible for the pseudorotaxane assembly and result in signifi-
cant perturbations in the UV/Vis spectra.
Figure 3. Partial 1H NMR spectra (500 MHz) in CDCl3 at 293 K of
a) macrocycle 2, b) 2 plus 1.0 equivalent of thread 1·BF4 and c) 1·2·BF4
plus 1.0 equivalent of TBACl.
To investigate whether changes in the pseudorotaxane co-
conformation could also be observed by UV/Vis spectrosco-
py, the effect of chloride addition to solutions of 1·BF4 and 2
was studied. However, no changes in absorbance were ob-
served (see the Supporting Information, Figure S15).
In summary, the naphthalimide triazolium threading
motifs 1·BF4 and 1·Cl were observed to form pseudorotax-
ane assemblies with macrocycle 2 of different co-conforma-
tions in the presence and absence of a coordinating halide
anion. Pseudorotaxane formation was signalled by signifi-
cant changes in the absorbance band of the naphthalimide
triazolium thread, whilst changes in the anion-dependent co-
conformation of the interpenetrative assembly were not de-
tectable by UV/Vis spectroscopy.
for 1·2·BF4 and 804(20) mÀ1 for 1·2·Cl. Thus, the two pseu-
dorotaxane assemblies exhibit similar thermodynamic stabil-
ities compared with the respective threading components,
1·BF4 and 1·Cl. However, any comparison is complicated by
the fact that strictly the latter value is an apparent 1:1 stabil-
ity constant for a ternary complex. The pseudorotaxane sta-
bility constant for 1·2·BF4, with the non-coordinating tetra-
fluoroborate anion, reflects the strength of intercomponent
naphthalimide–hydroquinone donor–acceptor and imide–
amide hydrogen-bonding interactions present in this assem-
bly. In contrast, the apparent 1:1 stability constant for 1·2·Cl
represents only the intercomponent triazolium–hydroqui-
none donor–acceptor and macrocycle–halide hydrogen-
bonding contributions, and does not take into account the
electrostatic and hydrogen-bonding interactions between the
triazolium thread and the chloride anion, also present in
1·Cl. However, the triazolium–halide interaction is highly fa-
vourable in this non-competitive solvent mixture, both in
1·Cl and 1·2·Cl, and thus can be said to drive the anion-in-
duced co-conformational change exhibited by 1·2·BF4 upon
the addition of TBACl in CDCl3 (Scheme 1 and Figure 3).
Attempts were made to study both the formation and co-
conformations of the pseudorotaxane assemblies with UV/
Vis spectroscopy in CHCl3. These titration experiments
Rotaxane synthesis: The construction of an interlocked ro-
taxane structure, incorporating the naphthalimide triazolium
motif, was undertaken with the expectation that this system
would be capable of undergoing anion-induced molecular
motion.
Asymmetric axle component 3·X (Figure 5) was prepared
as detailed in the Supporting Information (Scheme S2), and
a “clipping” method was employed for rotaxane formation
by using Grubbsꢀ-catalysed ring-closing metathesis (RCM)
of bis vinyl-appended macrocycle precursor 4.[12] Anion tem-
plation required the presence of a coordinating anion, and
reaction of 3·Cl with 1.5 equivalents of 4 in dry CH2Cl2 with
Grubbsꢀ second-generation RCM catalyst afforded the
target rotaxane 5·PF6 (Scheme 2), isolated in 47% yield, fol-
lowing anion exchange to the non-coordinating hexafluoro-
phosphate salt and purification by preparative silica gel
thin-layer chromatography.
1
were undertaken at the same concentration as the H NMR
experiments (1ꢂ10À3 m) to avoid the need for large excesses
of either component, both of which are UV active. Upon ad-
dition of macrocycle 2 to thread 1·BF4, the appearance and
increase in macrocycle absorbance (290 nm, see the Sup-
porting Information, Figure S14) was accompanied by a de-
crease in absorption and a slight bathochromic shift in the
naphthalimide triazolium band (Figure 4). The donor–ac-
ceptor interactions between the macrocycle hydroquinone
groups and the naphthalimide triazolium thread are respon-
1
The H NMR spectrum of rotaxane 5·PF6, along with the
spectra of axle 3·BF4 and macrocycle precursor 4, is shown
in Figure 6. Upon rotaxane formation, the signals associated
with the hydroquinone environments f and g move signifi-
cantly upfield (Dd=À0.92 ppm) due to the expected donor–
acceptor interactions. This shift is consistent with the pro-
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

P. D. Beer et al.
Figure 5. Axle component 3·X and macrocycle precursor 4 used in rotax-
ane formation.
Figure 6. 1H NMR spectra (300 MHz) in CDCl3 at 293 K of a) axle 3·BF4,
b) rotaxane 5·PF6 and c) macrocycle precursor 4.
posed co-conformation of 5·PF6, based on that observed for
pseudorotaxane 1·2·BF4. In addition, macrocycle protons
b and c shifted downfield, presumably due to hydrogen-
bonding interactions with the axle imide group, along with
axle triazolium proton 14.
Further characterisation of rotaxane 5·PF6 was provided
by high-resolution mass spectrometry and 2D 1H NMR
ROESY spectroscopy, with the proposed co-conformation
supported by the observed intercomponent ROESY cou-
plings. The spectrum and couplings are shown in the Sup-
porting Information (Figure S6).
Importantly, rotaxane formation was also attempted in
the absence of the coordinating halide counter anion by
using axle 3·BF4, however no rotaxane product was ob-
served in the reaction mixture. This may be rationalised
when taking into account that the formation of the analo-
gous pseudorotaxane assembly, 1·2·BF4, is driven by the
donor–acceptor interactions between the macrocycle hydro-
quinone and axle naphthalimide groups. The pre-organisa-
tion of the hydroquinone groups necessary for this associa-
tion to occur does not exist in macrocycle precursor 4 and is
only created in the RCM reaction. Therefore, the synthesis
of rotaxane 5·PF6 requires the presence of a chloride anion,
which successfully templates the association of the two com-
ponents prior to macrocycle formation. The templation
occurs by virtue of halide anion binding interactions with
the triazolium and isophthalamide motifs of the axle and
macrocycle precursor.
Rotaxane NMR studies—shuttling: 1H NMR spectroscopy
was used to investigate the chloride-anion-induced co-con-
formational behaviour of rotaxane 5·PF6 (Scheme 3).
The 1H NMR spectrum of rotaxane 5·PF6 in CDCl3 is
shown in Figure 7a, along with the spectrum observed after
addition of one equivalent of TBACl (Figure 7b). Downfield
shifts in protons 14, b, c, f and g were observed, as well as
a significant upfield shift in the signal corresponding to axle
protons 16. The downfield shifts in protons 14, b and c indi-
cate cooperative hydrogen bonding to the chloride anion by
the macrocycle isophthalamide and axle triazolium binding
Scheme 2. Synthesis of naphthalimide triazolium rotaxane 5·PF6.
&
4
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Anion-Induced Shuttling of a Rotaxane
FULL PAPER
Scheme 3. Anion-induced molecular shuttling exhibited by rotaxane 5·PF6.
motifs. The creation of a unique anion-binding cavity within
the rotaxane requires macrocycle translocation (Scheme 3),
and the different extents of shielding associated with the
two rotaxane co-conformations are reflected in the shifts ob-
served for protons f, g and 16. The macrocycle hydroqui-
none protons experience a greater shielding effect when
they reside over the naphthalimide group of the axle in the
starting co-conformation (5·PF6), compared with over the
triazolium group in 5·Cl. This reduction in shielding results
in a downfield shift of macrocycle hydroquinone protons f
and g after the addition of TBACl (Figure 7b), and is similar
to that observed between pseudorotaxane co-conformations
1·2·BF4 and 1·2·Cl (Figure 3). Moreover, when the macrocy-
cle resides over the triazolium group, axle CH2 protons 16
become partially shielded by the hydroquinone groups. This
shielding effect is not present in 5·PF6, thus an upfield shift
in the signal corresponding to protons 16 is observed upon
halide anion addition.[13]
Crystals suitable for single-crystal X-ray diffraction struc-
tural determination of rotaxane 5·Cl were analysed by using
synchrotron radiation at Diamond Light Source beamline
I19.[14] The co-conformation of the interlocked structure in
the solid state is in agreement with that proposed for the so-
lution phase, namely, with the macrocycle hydroquinone
groups residing over the axle triazolium group (Figure 8).
Importantly, the chloride anion is bound within the rotax-
ane-host cavity by hydrogen-bonding interactions to protons
b
À
c
À
À
À
b, c, 14 and 16 (C H ···Cl distance 3.348(9) ꢃ, N H ···Cl
c’
À
14
À
À
À
3.218(8) ꢃ, N H ···Cl 3.297(8) ꢃ, C H ···Cl 3.309(6) ꢃ
16
À
À
and C H ···Cl 3.391(6) ꢃ). Further examination of the
crystalline network reveals that
the naphthalimide group of the
axle component is involved in
p-donor–p-acceptor
interac-
tions with hydroquinone
a
group of an adjacent rotaxane
molecule (centroid-to-centroid
distance 3.471(4) ꢃ, ring dihe-
dral angle 8.3(3)8; see the Sup-
porting
Information,
Fig-
ure S19). This illustrates the po-
tential for this interaction, with
the analogous intercomponent
donor–acceptor interactions re-
sponsible for the proposed solu-
tion-phase co-conformation of
5·PF6.
On account of the change in
rotaxane co-conformation being
driven by the hydrogen-bonding
interactions between the iso-
phthalamide and triazolium
groups of 5·PF6 and the coordi-
nated chloride anion, increasing
the competitive nature of the
solvent through protic-solvent
addition should weaken these
interactions and eventually re-
verse the molecular-motion
Figure 7. Partial 1H NMR spectra (500 MHz) at 293 K of a) rotaxane 5·PF6 in CDCl3, b) rotaxane 5·PF6 plus
1.0 equivalent of TBACl in CDCl3 and c) rotaxane 5·PF6 +1.0 equivalent of TBACl in 1:1 CDCl3/CD3OD.
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

P. D. Beer et al.
more competitive 4:1 CDCl3/
CD3OD solvent mixture was
used.[15]
Downfield shifts for the axle
triazolium and macrocycle iso-
À
phthalamide C H protons (14
and b) and macrocycle amide
À
N H protons (c) of the rotax-
ane were observed upon addi-
tion of TBACl to 5·PF6 in these
solvent conditions (Figure 10).
It was not possible to follow the
triazolium and amide protons
(14 and c) throughout the titra-
tion due to deuterium ex-
change. Additional shifts oc-
Figure 8. Crystal structure of rotaxane 5·Cl. Protons and lower-occupancy positions of disorder are omitted for
clarity.
shuttling process. Upon addition of increasing amounts of
1
CD3OD to a CDCl3 solution of 5·Cl, H NMR shifts were
observed to indicate that macrocycle translocation back to
the initial co-conformation was indeed occurring; the spec-
trum for 1:1 CDCl3/CD3OD is shown in Figure 7c. The loss
of the halide anion from the rotaxane binding cavity is re-
flected in an upfield shift in isophthalamide proton b (the
signals corresponding to triazolium proton 14 and amide
protons c were not observed after CD3OD addition due to
deuterium exchange). The return to the starting co-confor-
mation, and the associated changes in the shielding effects,
are shown by an upfield shift in hydroquinone protons f and
g and a downfield shift in axle protons 16—the reverse of
the shifts observed upon chloride anion addition to 5·PF6 in
CDCl3 (Figure 7b).
The change in rotaxane co-conformation can be moni-
tored by following the relative positions of the macrocycle
hydroquinone and alkene proton signals throughout the
halide and CD3OD NMR titration experiments (Figure 9).
The gradual changes in the relative positions of these peaks
Figure 9. Changes in the chemical shift of alkene protons k (ꢂ) and hy-
indicate that the molecular motion exhibited by rotaxane
5·PF6 is a dynamic process, which is fast on the NMR time-
scale. The shuttling behaviour is reversible and, importantly,
can be controlled by dual stimulus, namely, halide anion ad-
dition followed by an increase in the competitive nature of
the solvent.
*
droquinone protons f and g ( ) in rotaxane 5·PF6 in CDCl3 at 293 K
upon a) addition of TBACl, and b) subsequent addition of CD3OD.
curred for the signals corresponding to protons 16, f, g and
k, due to the same shielding effects discussed above and
seen in Figure 7.
The anion-induced co-conformational change exhibited by
naphthalimide triazolium rotaxane was further supported by
During the titration experiments for the other anions
(bromide, iodide, acetate and dihydrogen phosphate), down-
field shifts in isophthalamide and amide anion binding pro-
tons b and c were observed as expected, however, triazolium
proton 14 exhibited anion-dependent differential behaviour.
Before deuterium exchange, this proton signal shifted down-
field upon addition of chloride and dihydrogen phosphate,
and slightly upfield with iodide addition, whereas no signifi-
cant shifts were observed upon bromide and acetate addi-
tion (Figure S10 in the Supporting Information). These con-
trasting observations are thought to be the consequence of
the combination of downfield and upfield perturbations, re-
1
2D H NMR ROESY spectroscopy, with the proposed solu-
tion-phase co-conformation of 5·Cl confirmed by the ob-
served intercomponent couplings (see the Supporting Infor-
mation, Figure S7).
Rotaxane NMR studies—anion binding: The anion-binding
properties of rotaxane 5·PF6 with a range of TBA anion
salts, that is, chloride, bromide, iodide, dihydrogen phos-
phate and acetate, were investigated by 1H NMR titration
experiments. Binding was too strong for accurate associa-
tion-constant values to be determined in CDCl3, so the
&
6
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Anion-Induced Shuttling of a Rotaxane
FULL PAPER
ments (see the Supporting Information, Figure S11). These
protons shifted significantly upfield upon addition of chlo-
ride and bromide (Dd=À0.19 and À0.16 ppm, respectively,
after 10.0 equivalents), assigned to increased shielding from
hydroquinone groups upon macrocycle translocation. In
contrast, very slight downfield shifts of 0.02–0.05 ppm oc-
curred with the larger anions, indicating that any alteration
in the relative position of the macrocycle upon addition of
iodide and the oxoanions is not sufficient to perturb these
protons.
Some change in the co-conformation of the rotaxane-host
5·PF6 is necessary for all of the anions to interact with both
axle and macrocycle C-H and amide N-H hydrogen bond
1
donor anion binding motifs, however these H NMR shifts
indicate that complete macrocycle translocation from the
naphthalimide group of the axle to the triazolium group (as
seen in the solid state, Figure 8) only occurs with the strong-
ly bound chloride and bromide anions. The larger anions are
not able to bind within the rotaxane-host cavity, resulting in
weaker external binding and a reduction in the extent of
macrocycle translocation, presumably to retain some of the
favourable naphthalimide–hydroquinone donor–acceptor in-
teractions. This potential selectivity is also supported by the
behaviour of hydroquinone and alkene protons (f, g and k)
during the titrations (see the Supporting Information, Fig-
ure S12 and Table S1). Considerably smaller perturbations
were observed with iodide, dihydrogen phosphate and ace-
tate than with chloride and bromide, again indicating that
the extent of co-conformational change is much less with
the weaker-bound anions.
Figure 10. Partial 1H NMR spectra (500 MHz) in 4:1 CDCl3/CD3OD at
293 K of a) rotaxane 5·PF6, b) rotaxane 5·PF6 plus 1.0 equivalent of
TBACl and c) rotaxane 5·PF6 plus 5.0 equivalents of TBACl.
sulting from hydrogen-bonding interactions with the anion
and increased shielding from the hydroquinone groups upon
macrocycle translocation. The relative magnitudes of these
effects are likely to be dependent on the nature of the
anion, relating to the basicity of the anion and any differen-
ces in the exact co-conformation of the rotaxane-host struc-
ture.
By using WinEQNMR2,[11] 1:1 stoichiometric association
constants were determined from the titration data by moni-
toring macrocycle isophthalamide proton b of the rotaxane
host (Table 1). This proton is directly involved in anion
Rotaxane UV/Vis studies: Having quantified the selectivity
in the anion-induced shuttling of rotaxane 5·PF6 with
1H NMR titration experiments, we then investigated wheth-
er the change in the interlocked co-conformation could be
detected by UV/Vis spectroscopy. The magnitude of any
such response to different anions and the reversibility of the
molecular motion were studied in the less-competitive sol-
vent CHCl3, to ensure saturation in binding, and the associ-
ated co-conformational changes, but after the addition of
only several equivalents of anion.
Table 1. Association constants, K [MÀ1], of rotaxane 5·PF6 with anions in
4:1 CDCl3/CD3OD at 293 K. Obtained by monitoring proton b. Errors
are given in brackets.
À
Anion
ClÀ
527 (47)
BrÀ
673 (7)
IÀ
H2PO4
83 (6)
OAcÀ
59 (1)
K [MÀ1
]
173 (17)
Naphthalimide triazolium axle 3·BF4 did not display any
perturbations in its UV/Vis spectrum upon addition of
anions in CHCl3 (see the Supporting Information, Fig-
ure S16). However, the addition of TBACl to rotaxane 5·PF6
yielded a slight hypsochromic shift of the naphthalimide ab-
sorbance band, along with an enhancement in absorption
(Figure 11a). These gradual changes were observed upon
addition of up to two equivalents of chloride, after which
the UV/Vis spectra remained unchanged. Hence, the rela-
tive position of the macrocycle within the rotaxane-host
structure does indeed affect the naphthalimide absorbance
band, presumably due to changes in the donor–acceptor in-
teractions present in the two co-conformations. It is thought
that macrocycle translocation is signalled effectively, in con-
trast to the pseudorotaxane assemblies, due to the inter-
locked nature of rotaxane 5·PF6.
binding and could be accurately monitored throughout all ti-
tration experiments.
Table 1 shows that the smaller halide anions, that is, chlo-
ride and bromide, are bound considerably more strongly
than the larger iodide and more basic oxoanions, dihydrogen
phosphate and acetate. This is due to the three-dimensional
anion-binding cavity created within the interlocked structure
upon co-conformational change, which is of complementary
size and shape for the smaller halide guest species, with bro-
mide bound slightly more strongly than chloride.
The possibility of a different binding mode, external to
the rotaxane cavity, for the larger anions (iodide, dihydrogen
phosphate and acetate) is supported by differential shifts ob-
served for axle CH2 protons 16 during the titration experi-
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

P. D. Beer et al.
Conclusion
By exploiting the naphthalimide triazolium motif, a novel
rotaxane host system was prepared by using chloride-anion
templation and demonstrated to exhibit selective, uni-direc-
tional, anion-induced shuttling as evidenced by 1H NMR
and UV/Vis spectroscopy.
Initial 1H NMR pseudorotaxane investigations between
tetrafluoroborate and chloride thread salts and an isophtha-
lamide-containing macrocycle revealed the ability of the
naphthalimide triazolium motif to form interpenetrated as-
semblies with counter-anion-dependent co-conformations.
The uni-directional shuttling behaviour of the analogous
rotaxane was driven by anion binding and controlled by the
dual stimulus of anion addition and by increasing the com-
petitive nature of the solvent. 1H NMR titration experiments
revealed selectivity in rotaxane-host binding, and the extent
of macrocycle translocation was observed to be anion de-
pendent. Complete macrocycle translocation only occurred
upon the recognition of the strongly bound smaller halide
anions, namely, chloride and bromide. The three-dimension-
al anion-binding cavity of the rotaxane, which results upon
co-conformational change, is responsible for this selectivity.
The solid-state crystal structure of the rotaxane in the pres-
ence of chloride supports the proposed solution-phase co-
conformational behaviour.
Figure 11. UV/Vis spectra of a) rotaxane 5·PF6 upon addition of 0.0, 0.5,
1.0, 2.0 and 5.0 equivalents of TBACl; b) i) rotaxane 5·PF6, ii) after addi-
tion of 5.0 equivalents of TBACl and iii) after subsequent addition of
AgPF6 and filtration. All spectra were recorded in CHCl3 at 293 K, c=
1.0ꢂ10À5 m.
Changes in the co-conformation of the rotaxane upon
anion addition were detected by UV/Vis spectroscopy due
to the sensitivity of the axle naphthalimide absorbance band
to the position of the macrocycle component within the in-
terlocked structure. The chloride-induced shuttling behav-
iour of the rotaxane was demonstrated to be reversible
upon addition of silver hexafluorophosphate.
Similar perturbations were also observed upon addition of
bromide, iodide and dihydrogen phosphate to rotaxane
5·PF6 (see the Supporting Information, Figure S17).[16] With
chloride and bromide, increases in absorbance at 340 nm of
15 and 16% were observed after saturation, whilst dihydro-
gen phosphate addition resulted in a considerably smaller
change of 6%. This indicates that the reduced extent of
macrocycle translocation with the larger oxoanion is reflect-
ed directly in the UV/Vis response to dihydrogen phosphate.
However, complete macrocycle translocation was not ob-
Experimental Section
General considerations: Commercially available solvents and chemicals
were used without further purification, unless otherwise stated. Dry sol-
vents were obtained by purging with nitrogen and then passing through
an MBraun MPSP-800 column. Water was de-ionised and micro-filtered
1
served for iodide either in the H NMR studies, but this was
not evident in the UV/Vis investigations (14% increase in
absorbance).
ꢄ
with a Milli-Q Millipore machine. All tetrabutylammonium salts and
To investigate whether the reversibility of this shuttling
behaviour could be demonstrated, a large excess of AgPF6
was added to a solution of rotaxane 5·PF6 in the presence of
five equivalents of TBACl. Importantly, the naphthalimide
absorbance band shifted back to its initial position (Fig-
ure 11b), indicating a return to the original co-conforma-
tion.
In summary, perturbations of the naphthalimide axle ab-
sorbance band—due to co-conformational-dependent p-
donor–p-acceptor interactions with the macrocycle hydro-
quinone groups—enabled the reversible, anion-induced
shuttling exhibited by rotaxane-host-structure 5·PF6 to be
signalled and monitored by UV/Vis spectroscopy.[17]
Grubbsꢀ second-generation catalyst were stored in a vacuum desiccator
prior to use.
NMR spectra were recorded on Varian Mercury 300, Varian Unity Plus
500 and Bruker AVII 500 (with 13C Cryoprobe) spectrometers. Mass
spectrometry was carried out on a Bruker micrOTOF spectrometer.
Melting points were recorded on a Gallenkamp capillary melting point
apparatus and are uncorrected.
Literature procedures were used in the preparations of 2[10] and 4.[12] The
syntheses of 1·BF4, 1·Cl, 3·BF4 and 3·Cl are given in the Supporting Infor-
mation.
Rotaxane 5·PF6: Axle 3·Cl (40 mg, 0.029 mmol) and macrocycle precur-
sor 4 (28 mg, 0.044 mmol) were dissolved in dry CH2Cl2 (10 mL) and
stirred for 30 min. Grubbsꢀ second-generation catalyst (10% by weight,
2.8 mg) was then added and the reaction mixture was left to stir at room
temperature for 15 h under N2. The solvent was removed in vacuo and
macrocyclic byproducts removed by preparative silica gel thin-layer chro-
matography (6:4 CH2Cl2/CH3CN). The crude product and unreacted 3·Cl
were dissolved in CH2Cl2 (20 mL) and washed with 0.1m NH4PF6 (aq)
&
8
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Anion-Induced Shuttling of a Rotaxane
FULL PAPER
(10ꢂ10 mL) and water (4ꢂ10 mL). The organic fraction was dried over
MgSO4 and in vacuo, and the residue purified by preparative silica gel
thin-layer chromatography (4% CH3OH/CH2Cl2) to yield 5·PF6 as
a yellow solid (29 mg, 0.014 mmol, 47%). M.p.: 179–1818C; 1H NMR
(500 MHz, CDCl3): d=9.30 (s, 1H, H14), 9.26 (apps, 1H, Hb), 9.06 (d,
4J=1.3 Hz, 1H, Ha), 8.52 (d, 3J=7.3 Hz, 1H, H11), 8.28 (d, 3J=8.5 Hz,
Huang, W.-C. Hung, C.-C. Lai, Y.-H. Liu, S.-M. Peng, S.-H. Chiu,
Angew. Chem. 2007, 119, 6749–6753; Angew. Chem. Int. Ed. 2007,
46, 6629–6633.
[4] J. J. Gassensmith, S. Matthys, J.-J. Lee, A. Wojcik, P. V. Kamat, B. D.
Smith, Chem. Eur. J. 2010, 16, 2916–2921.
[5] H. Zheng, W. Zhou, J. Lv, X. Yin, Y. Li, H. Liu, Y. Li, Chem. Eur. J.
2009, 15, 13253–13262.
[6] B. W. Laursen, S. Nygaard, J. O. Jeppesen, J. F. Stoddart, Org. Lett.
2004, 6, 4167–4170.
[7] A. M. Brouwer, C. Frochot, F. G. Gatti, D. A. Leigh, L. Mottier, F.
Paolucci, S. Roffia, G. W. H. Wurpel, Science 2001, 291, 2124–2128.
[8] M. D. Lankshear, P. D. Beer, Acc. Chem. Res. 2007, 40, 657–668.
[9] K. M. Mullen, J. Mercurio, C. J. Serpell, P. D. Beer, Angew. Chem.
2009, 121, 4875–4878; Angew. Chem. Int. Ed. 2009, 48, 4781–4784.
[10] M. R. Sambrook, P. D. Beer, J. A. Wisner, R. L. Paul, A. R. Cowley,
F. Szemes, M. G. B. Drew, J. Am. Chem. Soc. 2005, 127, 2292–2302.
[11] M. J. Hynes, J. Chem. Soc. Dalton Trans. 1993, 311–312.
[12] M. R. Sambrook, P. D. Beer, M. D. Lankshear, R. F. Ludlow, J. A.
Wisner, Org. Biomol. Chem. 2006, 4, 1529–1538.
3
3
1H, H13), 8.18 (d, J=7.9 Hz, 1H, H9), 7.99 (d, J=7.9 Hz, 1H, H10), 7.93
(dd, 3J=8.5 Hz, 3J=7.3 Hz, 1H, H12), 7.58 (brs, 2H, c), 7.22–7.27 (m,
12H, H2 and H20), 7.05–7.12 (m, 16H, H3, H4, H18 and H19), 6.98 (d, 3J=
8.8 Hz, 2H, H17), 6.61 (d, 3J=8.8 Hz, 2H, H5), 5.92–5.98 (m, 2H, Hk),
5.80–5.91 (m, 8H, Hf and Hg), 5.47 (s, 2H, H16), 4.58 (s, 3H, H15), 4.21–
4.27 (m, 3J=7.0 Hz, 2H, H8), 3.95 (brs, 4H, Hj), 3.86–3.89 (m, 2H, H6),
3.59–3.85 (m, 16H, Hd, He, Hh and Hi), 2.16–2.23 ( m, 2H,H7), 1.32 (s,
27H, H1 or H21), 1.31 ppm (s, 27H, H1 or H21); 19F NMR (282.5 MHz,
CDCl3): d=À73.4 ppm (d, 1J=712 Hz, PF6); 13C NMR (125 MHz,
CDCl3): d=165.8, 164.9, 162.6, 156.3, 154.7, 152.0, 151.6, 149.2, 148.6,
148.3, 144.1, 143.7, 142.4, 140.4, 140.2, 136.0, 133.9, 132.9, 132.8, 132.3,
131.3, 130.8, 130.7, 130.6, 129.6, 129.5, 129.4, 128.8, 127.5, 126.7, 125.6,
125.2, 124.2, 124.1, 122.2, 114.2, 113.9, 113.1, 112.9, 70.7, 69.5, 67.1, 66.9,
65.4, 63.1, 63.0, 58.3, 40.9, 39.2, 38.8, 34.3, 34.3, 31.4, 31.4, 28.0 ppm;
HRMS (ESI): m/z: calcd for C125H138N7O14: 1962.0329 [MÀPF6]+; found:
1962.0347 .
[13] Only minimal shifts in the axle naphthalimide protons of 5·PF6 were
observed upon chloride addition, indicating no significant interac-
tions between these protons and the halide anion.
[14] Crystal data for 5·Cl: formula: C93H103N4O4·C32H35N3O10·Cl, M=
1997.87, monoclinic, P21/c (no. 14), a=11.519(6), b=14.600(7), c=
73.86(4) ꢃ,
b=90.161(7)8,
V=12422(11) ꢃ3,
, T=100 K, colourless plate (0.10ꢂ
Z=4,
1cald =
1.068 gcmÀ3
,
m=0.090 mmÀ1
0.04ꢂ0.01 mm3), Crystal Logic diffractometer and Rigaku Saturn
724+ detector (l=0.68890 ꢃ, Diamond Light Source, I-19 beam-
line); 19476 independent measured reflections (Rint =0.1266), F2 re-
Acknowledgements
G.T.S. thanks the EPSRC for a studentship (EP/F011504) and Nicholas
G. White for help with crystallography. M.B.P. thanks the Diamond Light
Source (U.K.) for synchrotron beam time on I-19 and the (EPSRC) Na-
tional Crystallography Service for data collection.
2ACTHNUTRGNEUNG
finement, final R1 (F2 >2s(F2))=0.1006, wR (F2>2s(F2))=0.2518,
GoF=0.907, data completeness to q=24.218=88.7%. CCDC-
862707 contains the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_-
request/cif.
[1] A. Coskun, M. Banaszak, R. D. Astumian, J. F. Stoddart, B. A. Grzy-
bowski, Chem. Soc. Rev. 2012, 41, 19–30; V. Balzani, A. Credi, M.
Venturi, Molecular Devices and Machines: Concepts and Perspec-
tives for the Nanoworld, Wiley, Weinheim, 2008; E. R. Kay, D. A.
Leigh, F. Zerbetto, Angew. Chem. 2007, 119, 72–196; Angew. Chem.
Int. Ed. 2007, 46, 72–191; V. Balzani, A. Credi, S. Silvi, M. Venturi,
Chem. Soc. Rev. 2006, 35, 1135–1149; Structure and Bonding,
Vol 99: Molecular Machines and Motors (Ed.: J. P. Sauvage), Spring-
er, Berlin Heidelberg, 2001.
[2] P. A. Gale, T. Gunnlaugsson, Chem. Soc. Rev. 2010, 39, 3595–3596.
[3] E. V. Dzyuba, L. Kaufmann, N. L. Lçw, A. K. Meyer, H. D. F. Win-
kler, K. Rissanen, C. A. Schalley, Org. Lett. 2011, 13, 4838–4841;
C. J. Serpell, R. Chall, A. L. Thompson, P. D. Beer, Dalton Trans.
2011, 40, 12052–12055; N. H. Evans, C. J. Serpell, P. D. Beer, Chem.
Eur. J. 2011, 17, 7734–7738; T.-C. Lin, C.-C. Lai, S.-H. Chiu, Org.
Lett. 2009, 11, 613–616; M. J. Barrell, D. A. Leigh, P. J. Lusby,
A. M. Z. Slawin, Angew. Chem. 2008, 120, 8156–8159; Angew.
Chem. Int. Ed. 2008, 47, 8036–8039; C.-F. Lin, C.-C. Lai, Y.-H. Liu,
S.-M. Peng, S.-H. Chiu, Chem. Eur. J. 2007, 13, 4350–4355; Y.-L.
[15] Anion binding did not occur in the more competitive 1:1 CDCl3/
CD3OD solvent mixture, which was used for the triazolium rotaxane
in ref. [9], presumably due to the extra energetic cost of macrocycle
translocation in 5·PF6.
[16] Due to the insolubility of TBAOAc in CHCl3, the response upon
acetate addition was could not be investigated.
[17] We also intended to investigate the fluorescence of this system.
However, rotaxane 5·PF6 was barely fluorescent, displayed excita-
tion-dependent emission and produced very minor, irreproducible
responses to anions. It is thought that this is due to quenching ef-
fects from the macrocycle, as well as potentially complicating twist-
ed internal charge transfer (TICT) behaviour (J. R. Lakowicz, Prin-
ciples of Fluorescence Spectroscopy, 3rd ed., Springer, New York,
2006).
Received: January 30, 2012
Published online: && &&, 0000
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

P. D. Beer et al.
Molecular Motion
G. T. Spence, M. B. Pitak,
P. D. Beer* .................... &&&& — &&&&
Anion-Induced Shuttling of a Naphtha-
limide Triazolium Rotaxane
Selective shuttling: The anion-tem-
plated synthesis of a rotaxane struc-
ture, incorporating the novel naphtha-
limide triazolium motif, is described
and the interlocked host shown to
exhibit selective, uni-directional,
anion-induced shuttling behaviour (see
1
scheme), as evidenced by H NMR and
UV spectroscopy.
&
10
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!