Quantitative conformational study of redox-active [2]rotaxanes, part 1: Methodology and application to a model [2]rotaxane

Article abstract of DOI:10.1002/chem.200700898

This paper reports a novel methodology for the conformational analysis of [2]rotaxanes. It combines NMR spectroscopic (COSY, NOESY and the recently reported paramagnetic line-broadening and suppression technique) and electrochemical techniques to enable a quantitative analysis of the co-conformations of interlocked molecules and the conformations of their components. This methodology was used to study a model [2]rotaxane in solution. This [2]rotaxane consists of an axle that incorporates an electronpoor, doubly positively charged viologen that threads an electron-rich crown ether. It has been shown that the axle of the [2]rotaxane in its dicationic state adopts a folded conformation in solution and the crown ether is localised at the viologen moiety. Following a oneelectron reduction of viologen, the paramagnetic radical cation of the [2]rotaxane retains its folded conformation in solution. The data also demonstrate that in the radical cation the crown ether remains localised at the viologen, despite its reduced affinity for the singly reduced viologen. The combined quantitative NMR spectroscopic and electrochemical characterisation of the electromechanical function of the model [2]rotaxane in solution provides an important reference point for the study of switching in structurally related bistable [2]rotaxanes, which is the subject of the second part of this work.

Full text of DOI:10.1002/chem.200700898

FULL PAPER
DOI: 10.1002/chem.200700898
Quantitative Conformational Study of Redox-Active [2]Rotaxanes, Part 1:
Methodology and Application to a Model [2]Rotaxane
Silvano Altobello, Kirill Nikitin,* Jacek K. Stolarczyk,* Elena Lestini, and
Donald Fitzmaurice[a]
Abstract: This paper reports a novel
methodology for the conformational
analysis of [2]rotaxanes. It combines
NMR spectroscopic (COSY, NOESY
and the recently reported paramagnetic
line-broadening and suppression tech-
nique) and electrochemical techniques
to enable a quantitative analysis of the
co-conformations of interlocked mole-
cules and the conformations of their
components. This methodology was
used to study a model [2]rotaxane in
solution. This [2]rotaxane consists of
an axle that incorporates an electron-
poor, doubly positively charged violo-
gen that threads an electron-rich crown
ether. It has been shown that the axle
of the [2]rotaxane in its dicationic state
adopts a folded conformation in solu-
tion and the crown ether is localised at
the viologen moiety. Following a one-
electron reduction of viologen, the par-
amagnetic radical cation of the [2]ro-
taxane retains its folded conformation
in solution. The data also demonstrate
that in the radical cation the crown
ether remains localised at the viologen,
despite its reduced affinity for the
singly reduced viologen. The combined
quantitative NMR spectroscopic and
electrochemical characterisation of the
electromechanical function of the
model [2]rotaxane in solution provides
an important reference point for the
study of switching in structurally relat-
ed bistable [2]rotaxanes, which is the
subject of the second part of this work.
Keywords: cyclic voltammetry
·
molecular modeling · radical ions ·
rotaxanes
chemistry
·
supramolecular
Introduction
molecules may function as controllable molecular-level
switches or mechanical devices.
Interlocked molecules, such as rotaxanes and catenanes,
have been intensively studied owing to a large number of
potential applications in molecular electronic devices.[1] In
particular, these systems may exist in several diastereoiso-
meric structural forms,[2] which are usually referred to as
translational isomers or co-conformations.[3] Application of
an external stimulus, such as a change in pH,[4] a change of
electric potential[5] or photochemical excitation,[6] can induce
interconversion of the co-conformations of the rotaxanes or
catenanes by shuttling or circumrotational motions of the
components of the systems, respectively.[1,5,7] Hence, these
Electrochemical and photochemical switching of bistable
[2]rotaxanes has been known for more than a decade.[5,6,8] In
general, [2]rotaxanes comprise a dumbbell-shaped compo-
nent (hereafter referred to as an axle), which can include
one or more recognition sites (stations) and a ring compo-
nent. A representation of a simple viologen-based [2]ro-
taxane that contains only one electron-deficient, electro-
chemically active viologen station and one electron-rich
crown ether ring is shown in Figure 1. In such redox-active
molecules the position of the crown ether along the axle is
thought to be predominantly determined by the electro-
chemical states of the stations.[5] Therefore, a change in the
[a] S. Altobello, Dr. K. Nikitin, Dr. J. K. Stolarczyk, E. Lestini,
Prof. D. Fitzmaurice
School of Chemistry and Chemical Biology
University College Dublin
Belfield, Dublin 4 (Ireland)
Fax : (+353)1-716-2127
E-mail: kirill.nikitin@ucd.ie
jacek.stolarczyk@ucd.ie
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Figure 1. Schematic representation of a model [2]rotaxane.
Chem. Eur. J. 2008, 14, 1107 – 1116
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1107

electrochemical state of these stations may induce an inter-
conversion between the co-conformations of the molecule.
However, despite great progress in the studies of the inter-
conversion process in rotaxanes and catenanes,[1,3,5,7,9] rela-
tively little is known about the conformations of the individ-
ual components of the system in various oxidation states;
this particularly relates to paramagnetic states for which the
application of NMR spectroscopy is usually problematic.[3b]
The lack of full understanding of these interlocked molecu-
lar systems hinders the progress in developing their applica-
tions because their function often relies on the components
being in a particular conformation. It is clear that there is a
need for a general approach for the determination of the co-
conformations and conformations in various oxidation states
and to understand the relationship between conformation
and function.
For example, in a recent paper we described the self-as-
sembly and electrochemical switching of a heterosupramo-
lecular system that consisted of a tripodal bistable [2]ro-
taxane adsorbed at the surface of a titanium dioxide nano-
particle.[10] Our findings appeared to suggest that electron
transfer and shuttling were accompanied by significant con-
formational changes of the components, which included
folding of the axle. The study also raised important ques-
tions regarding the strength of the interactions between the
crown ether and viologen stations in different oxidation
states. On one hand, it showed that there was still a measur-
able interaction between them after the first, and even the
second, electron reduction of the viologen moiety. On the
other hand, we later found that a [2]pseudorotaxane that
consisted of a similar viologen moiety dissociated after the
first electron reduction, which indicated a very weak interac-
tion in the reduced state.[11] Such seemingly inconsistent re-
sults provide further arguments for a more rigorous ap-
proach to the conformational analysis of the [2]rotaxanes in
different oxidation states.
spatial constraints. This procedure enabled us to identify the
structure that most closely matches the experimental data.
The application of the NOE technique relies on the abili-
ty to resolve and assign the relevant resonances in the NMR
spectrum. This assignment proved to be very difficult for the
bistable [2]rotaxane system described earlier.[10] The spec-
trum of that system also contains a number of peaks that
are irrelevant for the conformational analysis of [2]ro-
taxanes. For these reasons and for clarity, herein the meth-
odology is applied to a less complicated model redox-active
[2]rotaxane (Figure 1). The [2]rotaxane consists of an axle
that incorporates a single electron-poor viologen station and
threads an electron-rich crown ether. It was envisaged that
such a design would retain the basic interaction between the
components of analogous bistable [2]rotaxanes and signifi-
cantly simplify the structure to permit unambiguous full as-
signment of all of the signals observed in the NMR spectra
in the aromatic and aliphatic regions. It is important that
such a simplification does not compromise the ability to
study the role of folding of the flexible chain in the axle in
detail and to study the interaction between the viologen sta-
tion and the crown ether.
Accordingly, the conformation of the model [2]rotaxane
(8·2PF6) in its parent dication state and single-electron re-
C
duced state (8 ·2PF6) was quantitatively characterised in so-
lution by using NOE and PASSY NMR spectroscopy tech-
niques. The results were supplemented by the electrochemi-
cal characterisation of 8·2PF6 and also by the molecular me-
chanics modelling of its two hypothetical diastereoisomers.
The reported findings support the notion that the electrome-
chanical function of the redox-active model [2]rotaxane de-
pends not only on the affinities of the components of the in-
terlocked molecule for each other in different oxidation
states, but also on the conformations of the components.
This paper describes the synthesis of 8·2PF6, the confor-
mational analysis of the dication state of 8·2PF6 in solution,
followed by the conformational analysis of the singly re-
In this paper we report a novel methodology to quantita-
tively study the diastereoisomerism and conformations of
rotaxanes in solution. To date, several NMR spectroscopy
techniques, particularly COSY and NOESY,[12] have proven
to be very useful in sensitive and non-invasive structure
characterisation of [2]rotaxanes and catenanes.[13] The NMR
spectroscopy techniques were used together with cyclic vol-
tammetry (CV),[1c] mass spectrometry[13e] or continuous-
wave EPR spectroscopy.[7g] 2D-IR spectroscopy was used to
observe the changes in the co-conformations on a picosec-
ond timescale.[7h] Although application of NMR spectrosco-
py to paramagnetic species (such as a singly reduced violo-
gen-based rotaxane) is limited, it has recently been demon-
strated that important structural information may be ob-
tained by a quantitative analysis of paramagnetic line-broad-
ening and suppression (PASSY) in [2]rotaxanes.[11] Herein,
we combine the results of NOESY data with PASSY, elec-
trochemical and reference crystallographic data. The com-
bined results are compared by using the concept of effective
distances with a number of hypothetical diastereoisomers
obtained by molecular mechanics modelling with different
C
duced [2]rotaxane 8 ·PF6 and a discussion of the electro-
chemical data. Details of the assignment of NMR spectra
are given in the Supporting Information.
Further investigations of switching in two-station [2]ro-
taxanes by using the methodology presented herein are re-
ported in detail in the second part of this work.[11b]
Results and Discussion
Synthesis of the model [2]rotaxane: Model [2]rotaxane
8·2PF6 (Scheme 1) has been designed so that the first bulky
stopper group (SA) is rigidly attached to the viologen moiety
(V) at one of the nitrogen atoms. The other nitrogen of V is
connected through a flexible polyether linker chain to the
second bulky stopper (SB). These stoppers ensure that the
molecular axle 9·2PF6 remains interlocked with crown ether
7.
The synthesis of the SA–V component of 5·PF6 is shown in
Scheme 2. Initially, tris-(4-tert-butylphenyl)methanol (2) was
1108
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1107 – 1116

Redox-Active [2]Rotaxanes
FULL PAPER
prepared from 1-bromo-4-tert-
butylbenzene (1) and diethyl
carbonate in a yield of 56%.
Alcohol 2 was converted into
the corresponding aniline (3)
by using a previously reported
procedure,[14a] which involved
an electrophilic aromatic sub-
stitution in acetic acid with an
excess of aniline hydrochloride
as the substrate. Chloride 5·Cl
was obtained by the nucleo-
philic reaction of 3 with 1-(2,4-
dinitrophenyl)-4,4’-bipyridini-
um chloride (4·Cl).[14b] Chlo-
ride 5·Cl was purified by
column chromatography and
converted into the correspond-
ing
hexafluorophosphate
(5·PF6).[14a]
Substrate 5·PF6 was alkylat-
ed with flexible stopper com-
ponent 6 in the presence of 7.
Applying our recently reported
high-concentration approach to
the synthesis of [2]rotaxanes at
room temperature and atmos-
pheric pressure,[15c] [2]rotaxane
8·2PF6 was obtained in a good
yield (56%).
Owing to the fact that our
attempts to crystallise 8·2PF6
were unsuccessful, a structural-
ly related [2]pseudorotaxane
(10·2PF6, Scheme 1) was pre-
pared and fully characterised
by X-ray crystallography.
[2]Rotaxane 8·2PF6 gave the
well-resolved 1H NMR spec-
trum shown in Figure 2. It was
possible, therefore, to make a
full and unambiguous assign-
Scheme 1. Chemical structures 8·2PF6 and related components. The numbering corresponds to the NMR spec-
trum shown in Figure 2.
ment of all of the proton resonances in this spectrum
(Table S1 in the Supporting Information). These assignments
were made by using standard COSY and 1D NOE data,
which are also given in the Supporting Information.
Structural NMRstudy of 8·2PF in solution: A quantitative
6
NOE study of 8·2PF6 was used to estimate the interatomic
distances in solution. The NOE measurements were ob-
tained from a double pulsed field gradient spin echo
(DPFGSE-NOE) experiment in deoxygenated acetonitrile
at 258C.
The NOE spectrum of 8·2PF6 was recorded when the
crown ether aromatic protons (H11) were irradiated at d=
6.14 (Figure S6 in the Supporting Information). All NOE
enhancements were positive, which demonstrated relatively
Scheme 2. Synthesis of 8·2PF6. Reagents and conditions: i) 1) BuLi; 2)
Et2CO3, 08C, THF, 12 h;[15] ii) PhNH3Cl, acetic acid, 1008C, 24 h, 93%;
iii) EtOH; reflux; 24 h; 73%; iv) 7, benzonitrile, 5 d, RT, 56%.
Chem. Eur. J. 2008, 14, 1107 – 1116
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1109

K. Nikitin, J. K. Stolarczyk et al.
this case, it may be expected that dref in the solid state is
slightly shorter than the distances between protons H11 and
H10 in the solution structures of both 8·2PF6 and 10·2PF6 be-
ꢀ
cause there is the possibility that rotation about the C O
bonds will move the adjacent methylene groups out of the
aromatic plane.
The value of 3.20 Š for dref was obtained according to
Equation (1):
ꢀ
ꢁꢀ =
1
nir
nref
6
X X
1
Nref
ꢀ6
dref
¼
d
ð1Þ
ij
i
j
in which nir (=8) is the number of irradiated H11 protons,
nref (=8) is the number of reference H10 protons. Hence,
N
ref =nrefnir is the number of interacting irradiated–reference
pairs in the system. The values of dij are the distances be-
tween protons H11 and H10, which were obtained from the
X-ray crystallographic data of 10·2PF6.
Figure 2. 1H NMR of 8·2PF6 (top) and its radical cation 8 ·PF6 (bottom).
Numbering is in accordance with that shown in Scheme 1.
C
Employing the value obtained above for dref, the effective
distances between sets of chemically equivalent protons H11
and Hx in 8·2PF6 were evaluated by using Equation (2).
Here, the effective distance (dx) denotes the average dis-
tance in terms of the NOE interaction (sixth power decay
with distance).
fast tumbling of the molecule under the conditions reported.
The intensities of the observed NOE enhancements were in-
terpreted on the basis of the assumption that the contribu-
tions of molecular motion and spin diffusion were equal for
all positions in the molecule.[12]
The calculations of interatomic distances in 8·2PF6 in so-
lution were done by using crystallographic data for [2]pseu-
dorotaxane 10·2PF6 as a reference. Specifically, the distances
between protons H11 and H10 in the crystal structure of
10·2PF6 (Figure 3) were used to obtain the reference dis-
ꢀ
ꢁꢀ =
1
6
Nref
Nx NOEref
NOEx
ð2Þ
dx ¼ dref
Â
NOEx is the cross-peak volume integral for the set of equiv-
alent protons (Hx) of interest, which was extracted from the
NOE spectrum of 8·2PF6. These volumes are given in
Table S2 in the Supporting Information. Nx is the corre-
sponding number of interacting proton–proton pairs, which
is equal to nirnx in which nx is the number of Hx protons in
the analysed set. NOEref is the cross-peak volume integral
for protons H10.[18]
Analysis of the effective proton–proton distances estimat-
ed from the NOE spectrum (Table 1) is particularly useful
for V protons H6 to H9. For instance, the effective d6 and d9
distances are almost identical (4.35 and 4.31 Š, respective-
ly), which is also the case for distances d7 and d8 (3.86 and
3.85 Š, respectively). These results show that 7 is localised
at V. Furthermore, these results agree well with the distan-
ces observed in the crystal structure of 10·2PF6 (Figure 3).
Interestingly, the estimated distance from protons H11 to
protons H16 in 7 (d16 =4.20 Š) is shorter than the distance to
the H9 protons (d9 =4.31 Š). This finding may be explained
in terms of the tilt of the N-benzyl aromatic ring with re-
spect to the plane of V. The tilt arises from the sp3 hybridi-
sation of the N-benzyl methylene carbon and causes the H16
protons to be lifted above the plane of V and approach the
crown ether aromatic protons. For similar reasons, the effec-
tive distance (d5 =4.41 Š) of the conjugated phenylene
group adjacent to the other side of V is virtually equal to
distance d6. This is probably also owing to the fact that the
Figure 3. The crystal structure of 10·2PF6.
tance (dref). This value for dref corresponds to an imaginary
experiment in which the NOE signal of H10 is measured
when the H11 protons of 10·2PF6 are irradiated. Whereas
10·2PF6 is, at first glance, much simpler than 8·2PF6, the vi-
ologen moieties in both structures are almost identical and
they incorporate the same crown ether (7). Therefore, it is
expected that the average distances between protons H11
and H10 of 7 would be similar is both cases.
Generally, using solid-state data to elucidate the structure
in solution can only be considered an approximation owing
to the crystal lattice packing and solvation effects.[16,17] In
1110
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1107 – 1116

Redox-Active [2]Rotaxanes
FULL PAPER
Rotaxane 82+ (i.e., without counterions) was minimised in
vacuum by using molecular mechanics routines (MacroMo-
del 8.5 with AMBER force field) with two different con-
straints between two central quaternary carbon atoms of the
stoppers, 19 and 33 Š. These minimisations produced the
folded and stretched conformations shown in Figure 4.[19]
The effective distances were then calculated by using Equa-
tion (1), which was applied to the distances in the model
structures, for both of the minimised structures by assuming
that the aromatic H11 protons of 7 were irradiated. The re-
sults are presented in Table 1.
For certain proton sets whose signal could not be detected
in the NOE spectrum, the minimum distances (dmin) were
estimated according to Equation (2), which allowed compar-
isons between all of the conformations to be made. The
noise level in the spectrum was substituted for NOEx in
these cases as an upper limit for the NOE signal. Therefore,
the dmin values obtained (shown in italics) indicate that the
Hx protons in question do not, on average, approach the H11
Table 1. Effective distances [Š] between protons H11 and Hx in [2]ro-
taxane 8·2PF6 for experimental and model conformations.
Hx
No. Hx protons
Conformations
Experimental[a]
Folded
Stretched
H1
H2
H3
H25
H4
H5
H6
H7
H8
27
6
9.55Æ0.60
>8.80[b]
9.22
8.99
8.37
5.30
6.037.67
4.07
4.76
4.28
3.57
4.75
3.36
5.85
4.62
5.30
8.10
8.79
11.76
12.85
10.1318.72
8.92
12.22
10.74
9.24
6
6.44Æ0.33
2–6[c]
24.20
2
2
2
2
2
2
8
2
2
2
2
2
2
2
2
2
5.58Æ0.20
4.41Æ0.09
4.35Æ0.08
3.86Æ0.05
3.85Æ0.05
4.31Æ0.08
3.20Æ0.03
>7.33[b]
5.36
4.49
3.90
3.84
4.48
3.33
5.99
6.17
8.25
10.67
12.23
14.68
16.02
H9
H10
H15
H16
H17
H18
H19
H20
H21
H22
H23
H24
H26
4.20Æ0.07
4.63Æ0.12
>7.33[b]
>7.33[b]
>7.33[b]
>7.33[b]
>7.33[b]
protons closer than dmin
.
>7.33[b]
21.12
23.32
2–6[c]
9–27[c]
>7.54[b]
7.22
25.04
Comparison of the calculated distances (Table 1) shows
that the folded conformation is a good approximation of the
average structure of 8·2PF6 in solution. The results for the
stopper (SB) protons (H25, H26) are of particular relevance
owing to the large differences between the two conforma-
tions for distances d25 and d26.
These results allow us to conclude that SB in 8·2PF6 folds
back towards the centre of the molecule. The signals that
correspond to protons H25 and H26 were observed in the
NOE spectrum and a relatively good agreement was found
between the experimental and calculated values of distances
d25 and d26 in the folded conformation. The presence of a
signal is itself an important observation because in the case
7.59Æ0.835.89
[a] Distances were calculated by using dref =3.20 Š, based on X-ray data
for the crystal structure of 10·2PF6. [b] No corresponding signal was ob-
served in the NOE spectrum, the minimum distances (shown in italics)
were estimated on the basis of the noise level in the spectrum and dref
(see above). [c] Ambiguity in the number of interacting protons corre-
sponds to rotation of the C3 symmetrical stopper SB.
p-phenylene group is tilted with respect to the plane of the
viologen.
Whereas the above analysis allowed localisation of the
crown ether at V, the application of this approach to the
remote flexible parts of 8·2PF6
was problematic. As a result of
the sixth power decay of the
NOE interactions with dis-
tance, these interactions are
very sensitive to changes in the
conformation of the flexible
ether chain of the axle and the
stoppers. Therefore, to assist
the experimental NOE data,
two hypothetical conforma-
tions of 8·2PF6 were modelled.
They were termed folded and
stretched on account of the
conformation of the flexible
chain of the axle. This ap-
proach allowed a comparative
analysis of the effective distan-
ces dx obtained from experi-
mental NOE spectrum and ef-
fective distances calculated for
the two model structures to be
conducted.
Figure 4. Model of the folded (top) and stretched (bottom) conformations of 8·2PF6.
Chem. Eur. J. 2008, 14, 1107 – 1116
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1111

K. Nikitin, J. K. Stolarczyk et al.
of the stretched conformation protons H25 and H26 are far
beyond the limit of the NOE interactions (24.20 and
25.04 Š, respectively).
Structure of the radical cation of 8·2PF6 in solution: Many
of the applications envisaged for redox-active [2]rotaxanes
are based on assumptions about the relationship between
the oxidation/reduction states of these molecules and their
conformations in solution or at a surface.[1,7c,9] In this con-
text, and having established the detailed conformation of
8·2PF6 in solution, we have also investigated the detailed
conformation in the singly reduced state when the inter-
The experimental values of d25 and d26 are in fact slightly
greater than the corresponding values for the folded confor-
mation, which suggests that, on average, rotaxane 8·2PF6
adopts a more open conformation than that of the folded
one in solution. It should be noted, however, that another
effect may also contribute to the observed discrepancies.
The calculated distances between the SB protons and the H11
protons depend strongly on the internal conformation of SB.
Any changes to its conformation (including the orientation
of the terminal tert-butyl groups) induce changes in the
NOE interaction, and hence, in the calculated results. There-
fore, the discrepancy in the d26 values (8.12 vs. 5.89 Š) is still
consistent with the folding of SB of the axle. For protons H25
the analysis is more complicated because the signals of the
H3 and H25 protons overlap. It is clear, nonetheless, that the
experimentally determined distance for the combined set of
H3 and H25 protons (6.64 Š) corresponds well to the values
calculated separately for both sets in the folded conforma-
tion (8.37 and 5.30 Š, respectively), but not for the stretched
conformation (9.24 and 24.20 Š).
No signal in the NOE spectrum was observed for the
polyether flexible chain and the aromatic protons of the
axle (H18–H22), which is in agreement with the distances cal-
culated for both model conformations and the estimated de-
tection limit. Whereas these interactions may not be used to
discriminate between the folded and stretched conforma-
tions, the lack of corresponding signals in the spectrum pro-
vides an additional argument for the predominant localisa-
tion of 7 at the viologen station and against its shuttling
along the axle.
C
locked molecule accepts an electron to form 8 ·PF6.
The application of NMR spectroscopy to paramagnetic
species is, generally speaking, problematic.[22–25] However,
1H NMR spectroscopy has recently been used to structurally
characterise the radical cation states of a series of viologens
and [2]rotaxanes.[11,26] It was demonstrated that an analysis
of spin relaxation paramagnetic enhancement can be used
to reveal the position of the crown ether in [2]rotaxanes and
to elucidate the conformation of the axle.[11]
Following a single-electron reduction of 8·2PF6 by zinc in
acetonitrile, the 1H NMR spectrum of the corresponding
radical cation was recorded at 258C (Figure 2). It was found
C
that in the NMR spectrum of 8 ·PF6, all of the peaks belong-
ing to the viologen moiety (H6, H7, H8, H9) and to adjacent
positions (H15, H16, H17, H5, H4, H3) were no longer ob-
served. The corresponding paramagnetic suppression zone
(PSZ), surmised from these data, is represented in Figure 5
as a blue ellipsoid. However, the characteristic resonance of
the aromatic protons of 7 at d=6.14 (H11) is observed in the
C
NMR spectrum of 8 ·PF6 and is not shifted.
The folded conformation exhibits a tilt of 66.88 of the N-
benzyl aromatic ring with respect to the plane of viologen,
which leads to the proximity of protons H16 to the irradiated
H11 protons. The effective distance calculated for this con-
formation is comparable with the distance from the H9 viol-
ogen protons to protons H11. This result agrees well with the
experimentally determined distance d16, as discussed
above.[20]
The analysis of the results obtained for the H1 protons of
stopper SA indicates that folded stopper SB provides a steric
barrier to the movement of 7 along the axle of the rotaxane.
This folding effectively locks the crown ether at the violo-
gen. This conclusion is supported by very reasonable agree-
ment between the experimental d1 value (9.55 Š) and the
distance d1 calculated for the folded conformation (9.22 Š),
whereas the experimental d1 value is much shorter than that
calculated for the stretched conformation (12.22 Š). In the
stretched conformation the crown ether shifts towards the
N-benzyl group because there is no steric barrier provided
by stopper SB.
Figure 5. A possible representation of paramagnetic suppression zone in
C
the [2]rotaxane 8 ·PF6.
C
To determine the detailed conformation of 8 ·PF6 in solu-
tion, the relaxation times (T1) and relaxation rates (R1) for
all of the protons were measured as described in detail else-
where.[11,23] Then, the values for electron–proton paramag-
A question arises about the reasons for the folding of the
flexible axle chain and stopper SB. It is possible that folding
is driven by CH–p interactions between the tert-butyl
groups of SB and the aromatic rings of the crown ether.[21]
netic relaxation enhancement (PRE, Re1) were calculated on
[11]
C
the basis of T1 and R1 for 8·2PF6 and 8 ·PF6 (Table 2).
The values of Re1 do not provide, however, convenient
C
means of analysing the conformation of 8 ·PF6. They also do
1112
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1107 – 1116

Redox-Active [2]Rotaxanes
FULL PAPER
ꢀ
ꢁ
1
=
6
Table 2. Relaxation times, rates and estimated effective distances in
[2]rotaxane 8·2PF6.
PREref
PREx
dex ¼ deref
ð4Þ
[f]
Hx Dication Dication Radical Radical PRE dxe
dx
T1 [s]
R1 [sꢀ1
]
cation cation R1
A
An analysis of the dex values, which were estimated from
the PRE data, revealed that the ether chain of SB is folded
in the reduced state of the [2]rotaxane (8 ·PF6). Specifically,
the distances between the centre of the PSZ and protons
H25 and H26 (de =11.11 and 10.97 Š, respectively) are in
good agreement with the corresponding distances calculated
for the folded conformation. It is concluded that the confor-
T1 [s]
A
]
H1 1.8
H2 1.35
H3 1.2
H4 0.94
H5 0.94
H6 0.87
H7 0.87
H8 0.87
H9 0.87
H10 0.432.3 0.0080
H11 1.1 0.930.0080
H12 0.432.3 0.0080
H13 0.432.3 0.0080
H14 0.432.3 0.080
H15 0.36
H16 1.1
H17 1.2
H18 0.58
H19 0.65
H20 0.65
H21 0.58
H22 1.2
H23 1.2
H24 1.2
H25 1.2
H26 1.9
0.56
0.74
0.81
1.1
0.36
2
12.1Æ1.312.1
C
[a]
[b]
[b]
[b]
[b]
[b]
[b]
[b]
[b]
[c]
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
<5.6[d,e]
<5.6[d,e]
<5.6[d,e]
<5.6[d]
<5.6[d]
<5.6[d]
<5.6[d]
<5.6[d]
Æ0.7
10.5
9.1
7.2
4.9
3.4
2.2
2.8
4.7
5.3
4.2
4.4
6.2
7.0
6.3
6.4
8.2
11.0
[a]
[c]
[c]
[c]
[c]
[c]
[c]
[c]
[a]
[a]
1.1
[a]
1.15
1.15
1.15
1.15
[a]
C
mation of SB in 8 ·PF6 in acetonitrile is similar to that of the
[a]
[a]
parent dication form (8·2PF6), that is, folded.
The distances between protons H10 to H13 in 7 and the
centre of the PSZ are relatively short (de ꢁ6.2 Š) and gener-
ally in agreement with the range of values expected for the
model folded conformation and the crystal structure of
10·2PF6.[27]
125
125
61.23
125
124 6.2
Æ0.7
61.23
61.23
10
Æ0.7
125
Æ0.7
12.5
9.4
Æ1.0
[a]
[a]
[a]
[b]
[b]
[b]
[c]
2.8
–
–
–
–
–
–
–
–
<5.6[d]
<5.6[d]
<5.6[d,e]
9.3Æ1.0
[c]
[c]
0.93–
0.81
1.7
1.5
1.5
C
It is, therefore, concluded that in 8 ·PF6 the flexible poly-
ether chain attached to SB remains in the folded conforma-
tion and that 7 remains localised at V. The latter is an im-
portant observation because there are three viable possibili-
ties for the state of 7 after the reduction step: 1) it remains
at V, 2) it travels without restraint along the axle or 3) it
moves to the flexible polyether chain of the axle. The viabil-
ity of the second option is based on a weakening of the elec-
trostatic interaction between positively charged V and elec-
tron-rich 7 and a previously observed decomplexation of a
[2]pseudorotaxane that contains a similar viologen moiety
after single-electron reduction.[11] The third option is based
on a possible affinity of 7 for polyether chain units.[28] The
results presented herein indicate that the interactions of 7
with V remain relatively strong in the singly reduced state
and underline the crucial role of the folding of the polyether
chain of the axle, which locks 7 into position.
0.080
0.22
0.22
0.22
0.22
0.29
0.29
0.22
12.5
4.5
4.5
4.5
4.5
3.5
3.5
4.5
4
11
311.5
311.5
311.6
4
3
3
4
Æ1.4 11.1
Æ1.4 13.6
1.7
Æ1.5 14.6
0.81
0.81
0.81
0.81
0.530.22
11.1Æ1.2 12.1
11.7Æ1.311.4
11.7Æ1.310.7
11.1Æ1.2
Æ1.2
9.7
9.9
4.5
11.0
[a] T1 is below 2.5”10ꢀ3 s. [b] Expected R1 is above 4.0”102 sꢀ1. [c] Very
high PRE value. [d] Based on R1 >4”102 sꢀ1 and by using Equation (4).
[e] Discrepancy is caused by a high PRE owing to spin-diffusion within
the aromatic system [f] Distances between the centre of the PSZ and Hx,
in folded diastereomer (Figure 4a) by using Equation (1).
not allow a straightforward comparison with the structure of
8·2PF6, which was expressed in terms of effective distances,
to be made. Therefore, the concept of effective distances
was extended for PRE data. In this case, the effective dis-
tances were calculated between all Hx proton groups and
the centre of the PSZ (see Figure 5) arbitrarily located at
the 4-position of the viologen pyridine ring adjacent to stop-
per SA. This asymmetric location was chosen to account for
the possible conjugation of SA with the viologen aromatic
system and the respective delocalisation of the unpaired
electron.
The effective distance from the centre of the PSZ to the
H1 protons for the folded conformation of 8·2PF6 was calcu-
lated by using Equation (1) and was used as deref (Figure 5).
The choice of the H1 protons was based on the assumption
that their position and respective distance do not change
during reduction. The dex values between the centre of the
PSZ and proton groups Hx were then calculated by using
Equation (4), which assumes a sixth power decay of Re1 with
respect to the distance away from the centre of the PSZ.
The results are given in Table 2. It is important to stress,
however, that owing to delocalisation of the unpaired elec-
tron and strong spin diffusion in the vicinity of the viologen
moiety, these results have to be treated carefully.
Electrochemical study of the supramolecular function: As
the reduction potential of 8·2PF6 is the measure of how
readily viologen accepts an electron, the value of the reduc-
tion potential is sensitive to the presence and the location of
7 in this [2]rotaxane. It has previously been demonstrated
that the complexation of 7 with V shifts the reduction po-
tential (E) of that moiety to more negative values.[5,10,14a]
Therefore, the supramolecular electromechanical function of
8·2PF6 depends on the location of 7 and the ability of this
[2]rotaxane to adopt various conformations in different oxi-
dation states.
As shown above, the detailed conformational study of
8·2PF6 in its two oxidation states (V2+ and V+) reveals that
the position of 7 and the conformation of SB remain very
similar upon reduction. Because it is feasible to electro-
chemically study 8·2PF6 in various oxidation states, a com-
parative cyclic voltammetry study of 8·2PF6 and 9·2PF6 was
undertaken.
CV measurements were carried out at a scan rate of
100 mVsꢀ1 by using a 0.1 moldmꢀ3 solution of tetrabutylam-
monium perchlorate as an electrolyte and non-aqueous Ag/
AgNO3 as the reference electrode. The voltammograms re-
Chem. Eur. J. 2008, 14, 1107 – 1116
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1113

K. Nikitin, J. K. Stolarczyk et al.
corded are shown in Figure 6 and the respective peak poten-
tials (Epeak) are given in Table 3.
Conclusion
The methodology presented herein allowed us to character-
ise diastereoisomerism and conformations of the model ro-
taxane 8·2PF6 in solution. It was demonstrated that 7 is lo-
calised at V in 8·2PF6 and the flexible stopper group is
folded back to the centre of the molecule. In the singly re-
C
duced state of 8 ·PF6, the position of 7 does not significantly
change and the interaction with the viologen moiety is not
completely inhibited. Additionally, the folding of SB, which
is possibly driven by CH–p interactions, provides further
C
stabilisation of 7 at V in 8 ·PF6.
It can be concluded that the technique presented is a
useful tool to study structure and conformations of inter-
locked systems in both diamagnetic and paramagnetic states.
The conformational factors can significantly influence ex-
pected electrochemical and electromechanical function of
interlocked molecules.
The reported methodology provides a level of insight that
is arguably unnecessary to fully analyse the simple model
[2]rotaxane 8·2PF6. It was, nevertheless, presented in its en-
tirety to demonstrate the capabilities of the methodology,
but also to provide an important reference point for the
study of switchable [2]rotaxanes in solution, and possibly, at
the surface. In this context, the structural studies of more
complex bistable [2]rotaxanes by using the above approach
are reported in Part 2 of this paper.
Figure 6. Cyclic voltammograms of 8·2PF6 (3.0·10ꢀ3 moldmꢀ3) and 9·2PF6
(3.0·10ꢀ3 moldmꢀ3) were recorded at 100 mVsꢀ1 and 258C in acetonitrile
containing TBAP (0.10 moldmꢀ3).
Table 3. Cyclic voltammetry data[a] for [2]rotaxane 8·2PF6 and the corre-
sponding axle 9·2PF6.[b]
Experimental Section
Epeak 9·2PF6 [mV]
Epeak 8·2PF6 [mV]
DE[mV]
NMR spectra were recorded by using a Varian Inova 300 or a Varian
Inova 500 spectrometer in acetonitrile at 258C. For the PASSY measure-
ments, solutions (10ꢀ2 moldmꢀ3) were deoxygenated in the NMR tube
(Wilmad 535 PP-8 equipped with a rubber septum Z124338) by two suc-
cessive freeze–pump–thaw cycles. Zinc filings (10 mg) were then added
and the solutions deoxygenated by a further two successive freeze–
pump–thaw cycles. Finally, the above samples were sonicated for 5 min
under nitrogen and NMR spectra of the resulting deeply coloured solu-
tions of the radical cation were recorded.
V2+!V+
ꢀ668Æ10
ꢀ998Æ10
ꢀ767Æ10
ꢀ99Æ14
C
0
V+ !V
ꢀ1107Æ10
ꢀ109Æ14
C
[a] Values correspond to first derivative zero crossing points. [b] At a
concentration of 3.0”10ꢀ3 moldmꢀ3 in acetonitrile that contains the sup-
porting electrolyte nBu4NClO4 (0.010 moldmꢀ3), scan rate 100 mVsꢀ1
258C.
,
T1 relaxation times were measured by using the inversion recovery
method on a Varian Inova 500 spectrometer at 258C.[23,24] The samples
were irradiated by using the following sequence, D1 =8 s and P1 =1808,
D2 in the range 10ꢀ5 to 2 s and P2 =908.
The negative shifts found for the two reduction and two
oxidation peaks Epeak in 8·2PF6 mark a significant intercom-
ponent interaction within 8·2PF6 in its various oxidation
states. More specifically, it can be concluded that in both ox-
idation states (V2+ and V+) 7 is localised on the viologen
moiety. Interestingly, a negative shift in the first oxidation
peak V0!V+ (Figure 6) shows that even in the fully re-
duced state of V0, it is possible that 7 remains localised at
the viologen moiety. These findings are in full agreement
with the structural interpretation of the NOE, PASSY and
the T1 relaxometry measurements as well as with earlier re-
ported findings for similar tripodal [2]rotaxanes on surfa-
ces.[10]
Mass spectra were recorded by using a Finnigan Mat INCOS 50 quadru-
pole mass spectrometer. CVs were recorded by using a Solartron SI 1287
potentiostat controlled by a Labview program running on a Macintosh
Power PC at a scan rate of 100 mVsꢀ1. All CVs were recorded under the
following conditions: The working electrode was an isolated platinum
wire. The counter electrode was an isolated platinum foil. The reference
electrode was a non-aqueous Ag/Ag+ electrode with a fill solution that
consisted of AgNO3 (0.010 moldmꢀ3) in the electrolyte solution. The
electrolyte solution consisted of tetrabutylammonium perchlorate
(0.10 moldmꢀ3) in dry solvent. All compounds were separately dissolved
in the electrolyte solutions, which were bubbled with argon gas for
20 min prior to measurements being recorded.
Crystal data were collected by using a Bruker SMART APEX CCD area
detector diffractometer.
A full sphere of the reciprocal space was
scanned by phi-omega scans. Pseudo-empirical absorption correction
based on redundant reflections was performed by the program SADAB-
S.[29a] The structure was solved by direct methods by using SHELXS-97
and refined by full-matrix least-squares on F2 for all data by using
1114
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1107 – 1116

Redox-Active [2]Rotaxanes
FULL PAPER
SHELXL-97.[29b,c] Hydrogen atoms were added at calculated positions
and refined by using a riding model. Their isotropic temperature factors
were fixed to 1.2 times (1.5 times for methyl groups) the equivalent iso-
tropic displacement parameters of the carbon atom to which the hydro-
gen atom is attached. Anisotropic temperature factors were used for all
non-hydrogen atoms.
[2]Pseudorotaxane 10·2PF6: 1,1’-diethyl-4,4’-bipyridinium bis(hexafluoro-
phosphate) (0.005 g, 0.010 mmol) and 7 (0.0054 g, 0.010 mmol) were dis-
solved in acetone (0.5 mL) and the solution was concentrated by using
vapour diffusion conditions to give 10·2PF6 as a red solid (100%). M.p.
1218C (reversible decomposition into V and 7); 1H NMR (CD3CN,
300 MHz): d=9.35 (d, J=7.0 Hz, 2H), 8.59 (d, J=7.0 Hz, 2H), 6.54 (s,
8H), 5.01 (q, J=7.5 Hz, 4H), 3.82–3.84 (m, 8H), 3.71–3.75 (m, 8H), 3.72
(s, 16H), 1.83ppm (t, J=7.5 Hz, 6H); elemental analysis calcd (%) for
C42H58F12N2O10P2: C 48.47, H 5.62, F 45.21, N 2.29; found: C 48.48, H
5.71, N 2.28.
Melting points were estimated by using a Gallenkamp melting point
device and were not corrected. Reagents and solvents were purchased
from suppliers and used as received. All reactions were performed under
nitrogen by using glassware that was flame dried. Chromatographic sepa-
rations were performed by using silica (Merck, 40–63micron) and the
specified solvent system.
Compound 10·2PF6 was crystallised from acetone to obtain triclinic crys-
tals of 10·2PF6·2C3H6O suitable for X-ray diffractometry measurements.
CCDC-648423contains the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Compounds 2,[15] 4·Cl,[14] 6[15] and 7[30] were prepared by following previ-
ously reported procedures.
Amine 3: Alcohol
2 (0.23g, 0.54 mmol) and aniline hydrochloride
(0.14 g, 1.08 mmol) were heated in acetic acid (10 mL) at 1108C in a
sealed tube for 3h. The solvent was evaporated under vacuum, after
which methanol (15 mL) and HCl (2 moldmꢀ3, 1 mL) were added. The
slurry was stirred under reflux for 20 h and the solvent was evaporated
under reduced pressure. The residue was extracted with chloroform and
washed with aqueous NaHCO3. The extract was dried over MgSO4 and
then was concentrated under reduced pressure. The residue was purified
by column chromatography (ethyl acetate/cyclohexane) to give 3 as a
pale solid (0.25 g, 93%): M.p. 286 8C (lit. m.p. 2888C[15a]); 1H NMR
(CDCl3, 300 MHz): d=7.22 (d, J=8.8 Hz, 6H), 7.08 (d, J=8.6 Hz, 6H),
6.95 (d, J=8.6 Hz, 2H), 6.56 (d, J=8.6 Hz, 2H), 1.29 ppm (s, 27H); MS
(ES): m/z: 504.2 [M+H]+.
Acknowledgements
This research was funded by Science Foundation Ireland and EU Marie
Curie Host Development grants. The authors thank Dr. Helge Müller-
Bunz of the Chemical Services Unit (crystallography) at University Col-
lege Dublin for technical support.
[1] a) A. H. Flood, R. J. A. Ramirez, W. Deng, R. P. Muller, W. A. God-
dard, J. F. Stoddart, Aust. J. Chem. 2004, 57, 301–322; b) P. M.
Mendes, A. H. Flood, J. F. Stoddart, Appl. Phys. A: Mater. Sci. Pro-
cess. 2005, 80, 1197–1209; c) J. W. Choi, A. H. Flood, D. W. Steuer-
man, S. Nygaard, A. B. Braunschweig, N. N. P. Moonen, B. W. Laurs-
en, Y. Luo, E. DeIonno, A. J. Peters, J. O. Jeppesen, K. Yu, J. F.
Stoddart, J. R. Heath, Chem. Eur. J. 2006, 12, 261–279.
[2] Following IUPAC nomenclature, the term diastereoisomer is used
here to indicate any particular spatial arrangement of the atoms,
diastereoisomers that can be interconverted by rotation about single
bonds are referred to as conformations and translational isomers are
referred to as co-conformations.
[3] a) M. C. T. Fyfe, P. T. Glink, S. Menzer, J. F. Stoddart, A. J. P. White,
D. J. Williams, Angew. Chem. 1997, 109, 2158–2160; Angew. Chem.
Int. Ed. Engl. 1997, 36, 2068–2070; b) J. F. Stoddart, S. A. Vignon,
Collect. Czech. Chem. Commun. 2005, 70, 1493–1576.
[4] a) J. D. Badjic, V. Balzani, A. Credi, S. Silvi, J. F. Stoddart, Science
2004, 303, 1845–1849; b) P. R. Ashton, R. Ballardini, V. Balzani, I.
Baxter, A. Credi, M. C. T. Fyfe, M. T. Gandolfi, M. Gomez-Lopez,
M. V. Martinez-Diaz, A. Piersanti, N. Spencer, J. F. Stoddart, M.
Venturi, A. J. P. White, D. J. Williams, J. Am. Chem. Soc. 1998, 120,
11932–11942.
[5] a) P. R. Ashton, R. Ballardini, V. Balzani, A Credi, K. R. Dress, E.
Ishow, C. J. Kleverlaan, O. Kocian, J. A. Preece, N. Spencer, J. F.
Stoddart, M. Venturi, S. Wenger, Chem. Eur. J. 2000, 6, 3558–3574;
b) V. Balzani, M. Clemente-León, A. Credi, B. Ferrer, M. Venturi,
A. H. Flood, J. F. Stoddart, Proc. Natl. Acad. Sci. USA 2006, 103,
1178–1183; c) C. P. Collier, E. W. Wong, M. Belohradsky, F. M.
Raymo, J. F. Stoddart, P. J. Kuekes, R. S. Williams, J. R. Heath, Sci-
ence 1999, 285, 391–394; d) C. P. Collier, G. Mattersteig, E. W.
Wong, Y. Luo, K. Beverly, J. Sampaio, F. M. Raymo, J. F. Stoddart,
J. R. Heath, Science 2000, 289, 1172–1175.
Monocation 5·PF6: Amine 3 (0.1 g, 0.198 mmol) and the salt 4·Cl (0.11 g,
0.3mmol) were heated in dry ethanol (12 mL) at 80 8C in a sealed tube
for 25 h. The solvent was evaporated under reduced pressure. The resi-
due was dissolved in chloroform (60 mL) and was washed several times
with water to remove excess reagent. The organic layer was dried over
MgSO4 and then was concentrated. The residue was purified by column
chromatography (MeOH then MeOH/MeNO2/KPF6 (aq) 8.5:1:0.5). The
respective fraction was concentrated under vacuum, the residue was dis-
solved in MeNO2 and the organic phase was washed with water to
remove excess of KPF6. The solvent was evaporated under vacuum to
give 5·PF6 as a yellow solid (0.114 g, 73%): Decomp 230 8C; 1H NMR
(CD3OD, 300 MHz): d=9.03(d, J=7 Hz, 2H), 8.86 (d, J=6 Hz, 2H),
8.48 (d, J=7 Hz, 2H), 7.84 (d, J=6 Hz, 2H), 7.60 (d, J=2.2 Hz, 4H),
7.30 (d, J=8.6 Hz, 6H), 7.12 (d, J=8.6 Hz, 6H), 1.32 ppm (s, 27H); MS
(ES): m/z: 643.2 [MꢀPF6]+; elemental analysis calcd (%) for
C47H51N2PF6: C 71.56, H 6.52, N 3.55; found: C 71.00, H 6.53, N 3.52.
[2]Rotaxane 8·2PF6 and compound 9·2PF6: Monocation 5·PF6 (36 mg,
0.045 mmol) and crown ether 7 (50 mg, 0.09 mmol) were added to 6
(50 mg, 0.065 mmol) in benzonitrile (0.2 mL) at room temperature. After
5 d the residue was purified by column chromatography (SiO2; MeOH
then MeOH/MeNO2/KPF6 (aq) 9.25:0.5:0.25 v/v) to give 9·2PF6 as a
yellow solid (26 mg, 37%) and 8·2PF6 as a red glassy solid (54 mg, 56%).
8·2PF6: 1H NMR (CD3CN, 500 MHz): d=9.16 (d, J=7.2 Hz, 2H), 8.87
(d, J=6.9 Hz, 2H), 8.00 (d, J=7.2 Hz, 2H), 7.86 (d, J=6.9 Hz, 2H), 7.78
(d, J=9 Hz, 2H), 7.73(d, J=9 Hz, 2H), 7.66 (d, J=8.7 Hz, 2H), 7.41 (d,
J=8.7 Hz, 6H), 7.30 (d, J=9.4 Hz, 12H), 7.14 (d, J=8.7 Hz, 8H), 7.10
(d, J=9.0 Hz, 2H), 6.78 (d, J=9.0 Hz, 2H) 6.14 (s, 8H), 5.86 (s, 2H),
4.20 (septuplet, J=4.7 Hz, 2H), 4.07 (t, J=4.7 Hz, 2H), 3.86 (t, J=
4.7 Hz, 2H), 3.82 (t, J=4.7 Hz, 2H), 3.64–3.52 (m, 24H), 3.47–3.40 (m,
8H), 1.33 (s, 27H), 1.29 ppm (s, 27H); MS (ES) m/z: 931.1 [Mꢀ2PF6]2+
;
elemental analysis calcd (%) for C123H148F12N2O13P2: C 68.63, H 6.93, N
[6] a) G. Bottari, D. A. Leigh, E. M. Perez, J. Am. Chem. Soc. 2003,
125, 13360–13361; b) A. P. de Silva, N. D. McClenaghan, Chem. Eur.
J. 2004, 10, 574–586; c) H. Azehara, W. Mizutani, Y. Suzuki, T.
Ishida, Y. Nagawa, H. Tokumoto, K. Hiratani, Langmuir 2003, 19,
2115–2123; d) J. Berna, D. A. Leigh, M. Lubomska, S. M. Mendoza,
E. M. Perez, P. Rudolf, G. Teobaldi, F. Zerbetto, Nat. Mater. 2005, 4,
704–710; e) H. Murakami, A. Kawabuchi, K. Kotoo, M. Kunitake,
N. Nakashima, J. Am. Chem. Soc. 1997, 119, 7605–7606; f) C. A.
Stanier, S. J. Alderman, T. D. W. Claridge, H. L. Anderson, Angew.
1.30; found: C 68.21, H 6.95, N 1.50.
9·2PF6: 1H NMR (CD3CN, 500 MHz): d=9.11 (d, J=6.9 Hz, 2H), 8.98
(d, J=6.6 Hz, 2H), 8.53(d, J=6.7 Hz, 2H), 8.44 (d, J=6.7 Hz, 2H),
7.66–7.72 (m, 4H), 7.5 (d, J=8.9 Hz, 2H), 7.41 (d, J=8.5 Hz, 6H), 7.32
(d, J=8.6 Hz, 6H), 7.25 (d, J=8.5 Hz, 6H), 7.18–7.22 (m, 8H), 7.07 (d,
J=8.8 Hz, 2H), 6.83(d, J=8.9 Hz, 2H), 5.78 (s, 2H), 4.18 (t, J=4.5 Hz,
2H), 4.10 (t, J=4.5 Hz, 2H), 3.85–3.90 (m, 4H), 1.33 (s, 27H), 1.29 ppm
(s, 27H); MS (ES): m/z: 1324.8 [Mꢀ2PF6]+.
Chem. Eur. J. 2008, 14, 1107 – 1116
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1115

K. Nikitin, J. K. Stolarczyk et al.
Chem. 2002, 114, 1847–1850; Angew. Chem. Int. Ed. 2002, 41, 1769–
1772.
the summed cross-peak volume integral NOEsum can be extracted
from the spectrum. NOEref is then calculated as a contribution of
the H10–H11 interactions to the NOEsum. Aromatic crown ether pro-
tons H11 are much closer to protons H10 than to protons H12, H13 and
H14 in all conformations of the crown ether. Consequently the con-
tribution of protons H10 dominates in the total signal. It was as-
sumed, therefore, that the contribution of H10 could be calculated
from the solid state structure of the auxiliary pseudorotaxane
10·2PF6 by using Equation (3):
[7] a) M. C. Jimenez, C. Dietrich-Buchecker, J.-P. Sauvage, Angew.
Chem. 2000, 112, 3422–3425; Angew. Chem. Int. Ed. 2000, 39, 3284–
3287; b) V. Balzani, M. Gomez-Lopez, J. F. Stoddart, Acc. Chem.
Res. 1998, 31, 405–414; c) T. Yamamoto, H.-R. Tseng, J. F. Stoddart,
V. Balzani, A. Credi, F. Marchioni, M. Venturi, Collect. Czech.
Chem. Commun. 2003, 68, 1488–1514; d) W. Q. Deng, R. P. Muller,
W. A. Goddard, J. Am. Chem. Soc. 2004, 126, 13562–13563; e) D. A.
Leigh, A. Troisi, F. Zerbetto, Angew. Chem. 2000, 112, 358–361;
Angew. Chem. Int. Ed. 2000, 39, 350–353; f) J. V. Hernandez, E. R.
Kay, D. A. Leigh, Science 2004, 306, 1532–1537; g) A. Godt, G.
Jeschke, Magn. Reson. Chem. 2005, 43, S110-S118; h) O. F. A.
Larsen, P. Bodis, W. J. Buma, J. S. Hannam, D. A. Leigh, S. Wouters-
en, Proc. Natl. Acad. Sci. USA 2005, 102, 13378–13382.
[8] a) R. Bissel, E. Cordoba, A. E. Kaifer, J. F. Stoddart, Nature 1994,
369, 133–137; b) A. Livoreil, C. Dietrich-Buchecker, J.-P. Sauvage,
J. Am. Chem. Soc. 1994, 116, 9399–9400.
[9] H.-R. Tseng, S. A. Vignon, P. S. Celestre, J. Perkins, J. O. Jeppesen,
A. Di Fabio, R. Ballardini, M. T. Gandolfi, M. Venturi, V. Balzani,
J. F. Stoddart, Chem. Eur. J. 2004, 10, 155–172.
nir nref
PP
1
ꢀ =
6
dij
i
j
ꢀ
ꢁ
NOEref
=
NOEsum
(3)
j13
nir
nref
n12
n14
P
P
P
P
P
1
ꢀ =
6
1
1
ꢀ =
6
1
ꢀ =
6
ꢀ =
6
dij
þ
dij
þ
dij
þ
dij
i
j
j
j
j
in which n12, n13 and n14 are the numbers of H12, H13 and H14 protons,
respectively. By using the X-ray data for 10·2PF6 it was found that
NOEref =0.96NOEsum, that is, indeed NOEref is a dominant contribu-
tion to NOEsum. It can be concluded that using X-ray data here does
not affect the accuracy of the calculations.
[19] Remarkably, the minimisation without any distance constraints re-
sulted is a strongly folded conformation of the axle.
[10] B. Long, K. Nikitin, D. Fitzmaurice, J. Am. Chem. Soc. 2003, 125,
15490–15498.
[11] a) K. Nikitin, D. Fitzmaurice, J. Am. Chem. Soc. 2005, 127, 8067–
8076; b) K. Nikitin, E. Lestini, J. K. Stolarczyk, H. Müller-Bunz, D.
Fitzmaurice, Chem. Eur. J. 2008, 14, 1117–1128.
[12] D. Neuhaus, M. P. Williamson, The Nuclear Overhauser Effect in
Structural and Conformational Analysis, 2nd ed., Wiley-VCH, New
York, 2000.
[20] The absence of the signal in the NOE spectrum that corresponds to
protons H15 can be attributed to relatively short relaxation time T1
of H15 (0.36 s, see Table 2) compared to T1 of irradiated protons H11
(1.1 s).
[21] M. Nishio, M. Hirota, Tetrahedron 1989, 45, 7201–7245.
[22] NMR of Paramagnetic Molecules (Eds.: G. N. La Mar, Jr., W. Hor-
rocks, R. H. Holm), Academic Press, New York, 1973.
[23] R. L. Vold, J. S. Waugh, M. P. Klein, D. E. Phelps, J. Chem. Phys.
1968, 48, 3831–3832.
[24] G. I. Likhtenshtein, I. Adin, A. Novoselsky, A. Shames, I. Vaisbuch,
R. Glaser, Biophys. J. 1999, 77, 443–453.
[25] R. Sharp, L. Lohr, J. Miller, Prog. Nucl. Magn. Reson. Spectrosc.
2001, 38, 115–158.
[26] W. S. Jeon, E. Kim, Y. H. Ko, I. H. Hwang, J. W. Lee, S.-Y. Kim, H.-
J. Kim, K. Kim, Angew. Chem. 2005, 117, 89–93; Angew. Chem. Int.
Ed. 2005, 44, 87–91.
[27] The effective distances between a specified carbon atom in the
centre of the PSZ and protons H10–H13 calculated for the crystal
structure of 10·2PF6 are within the range of 4.03to 7.48 Š.
[28] a) H. Kitano, T. Hirabayashi, M. Gemmei-Ide, M. Kyogoku, Macro-
mol. Chem. Phys. 2004, 205, 1651–1659; b) Y. Kikuchi, Y. Aoyama,
Supramol. Chem. 1996, 7, 147–152.
[29] a) G. M. Sheldrick, SADABS, Bruker AXS, Madison, WI 53711,
2000; b) G. M. Sheldrick, SHELXS-97, Program for the Solution of
Crystal Structures, University of Gçttingen, Gçttingen (Germany),
1997; c) G. M. Sheldrick, SHELXL-97-2, Program for the Refine-
ment of Crystal Structures, University of Gçttingen, Gçttingen
(Germany), 1997.
[13] a) P. R. Ashton, R. Ballardini, V. Balzani, I. Baxter, A. Credi,
M. C. T. Fyfe, M. T. Gandolfi, M. Gomez-Lopez, M. V. Martinez-
Diaz, A. Piersanti, N. Spencer, J. F. Stoddart, M. Venturi, A. J. P.
White, D. J. Williams, J. Am. Chem. Soc. 1998, 120, 11932–11942;
b) A. C. Benniston, A. Harriman, V. Lynch, J. Am. Chem. Soc. 1995,
117, 5275–5291; c) C. A. Hunter, J. Am. Chem. Soc. 1992, 114,
5303–5311; d) P. R. Ashton, J. A. Bravo, F. M. Raymo, J. F. Stoddart,
A. J. P. White, D. J. Williams, Eur. J. Org. Chem. 1999, 899–908;
e) A. B. Braunschweig, C. M. Ronconi, J.-Y. Han, F. Aricó, S. J. Can-
trill, J. F. Stoddart, S. I. Khan, A. J. P. White, D. J. Williams, Eur. J.
Org. Chem. 2006, 1857–1866.
[14] a) B. Long, K. Nikitin, D. Fitzmaurice, J. Am. Chem. Soc. 2003, 125,
5152–5160; b) M. Hanke, C. Jutz, Synthesis 1980, 31–32.
[15] a) H. W. Gibson, S.-H. Lee, P. T. Engen, P. Lecavalier, J. Sze, Y. X.
Shen, M. Bheda, J. Org. Chem. 1993, 58, 3748–3756; b) P. R.
Ashton, R. Ballardini, V. Balzani, M. Belohradsky, M. T. Gandolfi,
D. Philp, L. Prodi, F. M. Raymo, M. V. Reddington, N. Spencer, J. F.
Stoddart, M. Venturi, D. J. Williams, J. Am. Chem. Soc. 1996, 118,
4931–4951; c) K. Nikitin, B. Long, D. Fitzmaurice, Chem. Commun.
2003, 282–283.
[16] G. Hilmersson, P. I. Arvidsson, Ö. Davidsson, M. Håkansson, J. Am.
Chem. Soc. 1998, 120, 8143–8149.
[17] P. M. Iovine, G. Veglia, G. Furst, M. J. Therien, J. Am. Chem. Soc.
2001, 123, 5668–5679.
[30] a) R. C. Helgeson, T. L. Tarnowski, J. M. Timko, D. J. Cram, J. Am.
Chem. Soc. 1977, 99, 6411; b) R. C. Helgeson, J. M. Timko, D. J.
Cram, J. Am. Chem. Soc. 1974, 96, 7380.
[18] The resonances assigned to the aliphatic crown ether protons of
8·2PF6 (H10, H12, H13 and H14) overlap, and as a consequence, only
Received: June 13, 2007
Published online: November 14, 2007
1116
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 1107 – 1116