Increased halide recognition strength by enhanced intercomponent preorganisation in triazolium containing [2]rotaxanes

Article abstract of DOI:10.1002/chem.201303122

Three triazolium-based [2]rotaxanes containing different sized axle and macrocycle components were synthesised in good yields (40-57 %) through chloride anion templation. The anion recognition properties of the interlocked receptor systems were investigat

Full text of DOI:10.1002/chem.201303122

FULL PAPER
DOI: 10.1002/chem.201303122
Increased Halide Recognition Strength by Enhanced Intercomponent
Preorganisation in Triazolium Containing [2]Rotaxanes
Nicholas G. White,[a] Paulo J. Costa,[b] Sꢀlvia Carvalho,[c] Vꢀtor Fꢁlix,[c] and
Paul D. Beer*[a]
Abstract: Three triazolium-based [2]ro-
taxanes containing different sized axle
and macrocycle components were syn-
thesised in good yields (40–57%)
through chloride anion templation. The
anion recognition properties of the in-
terlocked receptor systems were inves-
tigated using H NMR titration experi-
ments: all three rotaxanes display im-
pressive selectivities for halide anions
over the more basic oxoanion acetate.
The rotaxanes incorporating shorter,
more rigid axle components with aryl-
substituted triazolium groups display
substantially higher anion binding af-
finities than those with longer, bis-
alkyl-substituted heterocycles, which is
attributed to the increased intercompo-
nent preorganisation afforded by the
smaller axle component. Computation-
al DFT and molecular dynamics simu-
lations composed of unconstrained and
umbrella sampling simulations corrobo-
rate the experimental observations.
Keywords: anions
·
molecular
1
dynamics · preorganisation · rotax-
anes · triazolium
Introduction
The field of anion supramolecular chemistry has grown rap-
idly over the last few decades inspired by the fundamental
roles negatively charged species play in a range of biological
and environmental processes.[1,2] The recent discovery that
À
the readily-synthesised 1,2,3-triazole group is a potent C H
hydrogen bond donor for the recognition of anions in organ-
ic solvents has led to a variety of acyclic and macrocyclic re-
ceptors being prepared, with these systems showing efficient
anion complexation.[3–28] In contrast, in spite of the promise
of enhanced anion binding affinity through the addition of
a positive charge, the alkylated triazolium motif has not
been exploited in anion receptor design to the same
extent.[29–38]
Figure 1. Beerꢀs triazolium containing [2]rotaxane, 1·X.[32]
As part of our research programme concerned with the
construction of interlocked host molecules for anion recog-
nition applications,[39–43] we recently reported the first triazo-
lium axle containing rotaxane prepared by chloride and bro-
mide anion templation (Figure 1).[32] In an extension of this
preliminary work, we were interested to investigate whether
the strength and selectivity of anion recognition in such an
interlocked host could be altered significantly by subtle
modifications of the interlocked hostꢀs axle and macrocycle.
Although the concept of preorganisation[44] has been suc-
cessfully used to amplify recognition strength and selectivity
in acyclic[28,45–47] and macrocyclic host systems,[19,48,49] we are
unaware of any similar studies being undertaken with three
dimensional interlocked receptor molecules. Herein, we de-
scribe such an investigation by synthesising a range of tria-
zolium containing rotaxane receptor systems in which the
length, and hence the degree of preorganisation of the inter-
locked host structure is varied through the size and rigidity
of the triazolium axle component and the size of the iso-
phthalamide macrocycle (Figure 2).
[a] Dr. N. G. White, Prof. P. D. Beer
Chemistry Research Laboratory, Department of Chemistry,
University of Oxford, Mansfield Road, Oxford, OX1 3TA (UK)
E-mail: paul.beer@chem.ox.ac.uk
[b] Dr. P. J. Costa
Departamento de Quꢁmica, QOPNA
and Secc¼o Autꢂnoma de CiÞnciasda Safflde
Universidade de Aveiro, 3810-193 Aveiro (Portugal)
[c] Dr. S. Carvalho, Prof. V. FØlix
Departamento de Quꢁmica, CICECO and Seccżo Autꢂnoma de
CiÞncias da Safflde, Universidade de Aveiro
3810-193 Aveiro (Portugal)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201303122.
Chem. Eur. J. 2013, 19, 17751 – 17765
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
17751

The synthesis of the aryl-substituted triazolium axle com-
ponents required the preparation of new bulky stopper
groups 7 and 8, containing aryl azide, and aryl alkyne func-
tionality, respectively (Scheme 1). The azide 7 was prepared
À
Figure 2. Axle and macrocycle components used in this study (XÀ =BF4
,
ClÀ).
Scheme 1. Synthesis of 7 and 10. Conditions and reagents: a) NaNO2,
NaN3, 9:1 EtOH/conc. HCl(aq.), 08C to RT, 83%; b) trifluoromethanesul-
fonic anhydride, NEt3, CH2Cl2, 90%; c) TIPS-acetylene, CuI, Pd
NEt3, DMF, 758C, 61%.
ACHTUNGTRENNUNG(PPh3)4,
Results and Discussion
Synthesis of rotaxanes with varying length of axle compo-
nent: We initially investigated the effect of varying the
length and rigidity of the triazolium containing axle compo-
nent on the anion recognition capabilities of the resulting in-
terlocked rotaxane host systems. Our previously reported
[2]rotaxane 1+ contains an axle component 2+ with alkyl
substituents at both the C and N terminus of the triazolium
group, whereas in axle 3+ one alkyl and one aryl substituent
is present, and in 4+ the triazolium is bis-aryl-substituted.
This has two effects: firstly, the introduction of these aryl
groups will affect the rigidity of the axle, as well as the acidi-
ty of the triazolium proton; secondly having the bulky stop-
per groups directly attached to the triazolium motif reduces
the length of the axle. This reduces the potential translation-
al motion of the macrocycle along the axle, preorganising
the two components of the rotaxaneꢀs three-dimensional
recognition cavity to some extent.
in good yield by diazotisation of the corresponding amine
9,[50] using sodium nitrite and sodium azide in 9:1 ethanol/
conc. HCl(aq.). Triisopropylsilyl (TIPS) protected alkyne, 10
was prepared in two steps from the corresponding substitut-
ed phenol 11[51] by conversion to the aryl triflate 12, fol-
lowed by Sonogashira reaction[52] with TIPS-acetylene. De-
protection of the silyl group using tetrabutylammonium
(TBA) fluoride solution and isolation of the alkyne product
proved problematic with rapid decomposition being ob-
served. Therefore, the alkyne was prepared as the TIPS de-
rivative, and deprotected in situ in a subsequent reaction.
1
All new compounds were characterised by H and 13C NMR
spectroscopy and high resolution mass spectrometry (see the
Supporting Information for NMR spectra of new com-
pounds).
17752
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17751 – 17765

Halide Recognition by Triazolium Containing Rotaxanes
FULL PAPER
The copper(I)-catalysed azide–alkyne cycloaddition
(CuAAC) reaction[53,54] was used to synthesise the triazole
containing axle components 13 and 14. Reaction of aryl
azide 7 and alkyl-substituted alkyne 15,[55] in dichlorome-
The new rotaxanes were synthesised by chloride anion
templated Grubbsꢀ-catalysed ring closing metathesis (RCM)
of macrocycle precursor 16 and triazolium axle components
3·Cl and 4·Cl. Purification by preparative thin layer chroma-
tography gave the [2]rotaxanes 17·Cl and 18·Cl in 51 and
57% yield, respectively (Scheme 3 and Scheme 4). Chloride
anion template removal by repeated washing with
NH4PF6(aq.) gave the hexafluorophosphate salts 17·PF6 and
18·PF6 for anion recognition studies. The new rotaxanes
were characterised by 1H, 13C, 19F, 31P NMR spectroscopy
thane with catalytic [CuACTHNUTRGENN(UG CH3CN)4]ACHTUTGNREN(NUGN PF6), tris(benzyltriazolyl-
methyl)amine (TBTA) and diisopropylethyl amine
(DIPEA) gave 13 in good yield (89%). Reaction of 7 and 8
proceeded under similar conditions, but with the addition of
TBA·F to remove the triisopropylsilyl group; it was found
that far greater yields were obtained (85%) when stoichio-
metric rather than catalytic copper(I) was used (<20%
yield using 10% catalyst loading), presumably due to com-
peting decomposition of the unprotected alkyne in the
slower catalytic reaction. Alkylation using trimethyloxonium
tetrafluoroborate in dry dichloromethane gave the corre-
sponding triazolium containing axle components as their
1
and high resolution ESI mass spectrometry, as well as H-1H
2D ROESY NMR spectroscopy (see the Supporting Infor-
mation).
The interlocked nature of both rotaxanes was also con-
firmed by single crystal X-ray structural analysis. Single crys-
tals of 17+ and 18+ as their chloride salts were obtained, as
well as of 17·BF4 and two solvates of 17·PF6 (see the Sup-
porting Information). In the solid state structures of both
17·Cl and 18·Cl, the halide anion binds inside the inter-
locked host cavity, forming hy-
À
BF4 salts in high yields (95 and 89%). Anion exchange
using chloride-loaded Amberliteꢄ resin gave 3·Cl and 4·Cl
(Scheme 2).
drogen bonds to the triazolium
group of the axle component,
and the amide protons of the
macrocycle.
The
mono-aryl-substituted
rotaxane does not feature sig-
nificant pyridinium···hydroqui-
none donor–acceptor interac-
tions
methyl···poly
or
ether
triazolium-
hydrogen
AHCTUNGTRENNUNG
bonding (Figure 3). In contrast,
the bis-aryl-substituted rotax-
ane (Figure 3) contains both
donor–acceptor
interactions
(both centroid···centroid distan-
ces=3.657(4) Š) and triazoli-
um-methyl···polyether hydrogen
bonding.
Anion binding studies: The
effect of these structural modi-
fications on the rotaxanesꢀ
anion guest recognition proper-
ties
was
investigated
by
1H NMR titration experiments
in polar 1:1 CDCl3/CD3OD sol-
vent mixture. Aliquots of a solu-
tion of anion as the TBA salt
were added to the hexafluoro-
phosphate salt of the rotaxane
host causing downfield shifts in
the rotaxaneꢀs cavity protons a,
2 and 3, as shown for 18·PF6 in
Figure 4. Interestingly, down-
field shifts in the axle phenyl
proton b are also observed, sug-
gesting the involvement of this
Scheme 2. Synthesis of 3·Cl and 4·Cl. Conditions and reagents: a) 7, [CuACTHNUGTRENNUG(CH3CN)4]AHCTUGNTERN(NUGN PF6), TBTA, DIPEA,
CH2Cl2, 89%; b) (Me3O)
86%; d) 7, TBA·F in THF, [Cu
f) AmberliteꢄClÀ anion exchange column, 4:1 acetone/H2O, >95%.
(BF4), CH2Cl2, 95%; c) AmberliteꢄClÀ anion exchange column, 4:1 acetone/H2O,
ACHTUNGTRENNUNG
ACHNUTGERTG(NUNN CH3CN)4]AHCTNUTREGG(NNNU PF6), TBTA, DIPEA, CH2Cl2, 85%; e) (Me3O)ACHTNGUTERN(NUGN BF4), CH2Cl2, 89%;
Chem. Eur. J. 2013, 19, 17751 – 17765
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17753

P. D. Beer et al.
determination of 1:1 stoichio-
metric association constants
(Table 1).
Both new rotaxane host sys-
tems display impressive selec-
tivity for the halide anions over
larger, more basic acetate, with
chloride and bromide being
bound more than an order of
magnitude more strongly than
the oxoanion. Importantly, the
rotaxanes containing aryl-sub-
stituted triazolium axles com-
plex all anions significantly
more strongly than the previ-
ously reported bis-alkyl-substi-
tuted rotaxane, with the bis-
aryl-substituted rotaxane 18·PF6
showing slightly higher anion
affinities than 17·PF6. Also of
note is the observed preference
for the smaller halide anions,
chloride and bromide over
larger iodide, which is signifi-
cantly increased for the new
aryl-substituted rotaxanes rela-
tive to the bis-alkyl-substituted
host system.
Scheme 3. Synthesis of new triazolium [2]rotaxane 17·PF6. Conditions and reagents: a) Grubbsꢀ II catalyst,
CH2Cl2, 51%; b) NH4PF6(aq.), CH2Cl2, >95%.
Two possible causes for the
increased halide anion binding
affinities of 17·PF6 and 18·PF6
proton in anion binding, as well as upfield movement of hy-
droquinone protons 4 and 5, consistent with increased
donor–acceptor interactions upon chloride complexation.
Similar peak movements were observed for rotaxane 17·PF6.
Addition of the larger halide anions, bromide and iodide
caused similar peak movements, but of a smaller magnitude,
whereas only very small shifts were observed upon acetate
over 1·PF6 can be postulated: the anion complexation
strength of the aryl-substituted triazolium motif may be
higher than the bis-alkyl-substituted due to the electron
withdrawing nature of the aryl substituents increasing this
protonꢀs acidity, as has been observed for acyclic triazole re-
ceptors.[58] Alternatively, the greater preorganisation provid-
ed by the shorter axle component, and the reduced opportu-
nities for movement of the macrocycle component away
from the triazolium motif could also explain the increased
addition. WINEQNMR2[56] analysis of the titration data, fol-
[57]
À
lowing the macrocycleꢀs interior C H proton allowed the
Scheme 4. Synthesis of new triazolium [2]rotaxane 18·PF6. Conditions and reagents: a) Grubbsꢀ II catalyst, CH2Cl2, 57%; b) NH4PF6(aq.), CH2Cl2, >95%.
17754
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17751 – 17765

Halide Recognition by Triazolium Containing Rotaxanes
FULL PAPER
1
Figure 4. Partial H NMR spectra of 18·PF6 (bottom), and 18·PF6 and one
equivalent of TBA·Cl (top). Peak labelled x corresponds to residual
CHCl3 solvent signal, unlabelled peaks belong to bulky stopper groups
(1:1 CDCl3/CD3OD, 293 K, 500 MHz).
donor, it has the least electrostatic attraction for an anion,
as the positive charge is partially delocalised over the
phenyl rings. Further understanding of this electronic fea-
ture was obtained by DFT calculations (vide infra).
Table 1. Anion association constants, Ka [mÀ1], with rotaxane host sys-
tems.
[a]
[a]
[a]
1·PF6
17·PF6
18·PF6
ClÀ
BrÀ
IÀ
424(20)
423(14)
350(16)
42(3)
1114(109)
G
1287(50)
1455(76)
772(62)
126(10)
1322(21)
566(23)
92(15)
OAcÀ
Figure 3. Structures of 17·Cl and 18·Cl in the crystal. Solvent molecules,
hydrogen atoms, except those on triazolium and amide groups, are omit-
ted for clarity, as is end-to-end axle disorder, and twirling tert-butyl disor-
der in the structure of 17·Cl.
[a] Anions added as TBA salts. Association constants calculated using
WINEQNMR2[56] with estimated standard errors given in parentheses.
Solvent: 1:1 CDCl3/CD3OD, 293 K.
anion recognition strength. To investigate this further, the
anion recognition properties of the three axle components,
1
2·BF4, 3·BF4 and 4·BF4 were investigated using H NMR ti-
tration experiments in 9:1 CDCl3/CD3OD.[59]
As shown in Figure 5, addition of anions to the axle com-
ponents as their tetrafluoroborate salts results in significant
downfield shifts of triazolium resonance c, as well as a small
change in the environment of phenyl proton d. Quantitative
1:1 stoichiometric association constants were calculated
using WINEQNMR2,[56] monitoring the axleꢀs triazolium
proton. Surprisingly, all three axle components show similar
affinities for the three halide anions (Table 2). Although
anion association might be expected to be strongest for aryl-
substituted axle component 4, on the basis that the aryl sub-
À
stituents will increase the acidity of the triazolium C H
Figure 5. Partial 1H NMR spectra of 3·BF4 upon addition of TBA·Cl.
Peak labelled x corresponds to residual CHCl3 solvent signal, unlabelled
peaks belong to bulky stopper groups (9:1 CDCl3/CD3OD, 293 K,
500 MHz).
bond, this analysis overlooks the potential for charge deloc-
alisation afforded by the phenyl substituents. Thus, we sug-
gest that although 4+ has the most potent hydrogen bond
Chem. Eur. J. 2013, 19, 17751 – 17765
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17755

P. D. Beer et al.
Table 2. Anion association constants, Ka [mÀ1], with axle components.
[a]
[a]
[a]
2·BF4
3·BF4
4·BF4
ClÀ
BrÀ
IÀ
735(55)
891(60)
681(84)
701(38)
672(50)
584(54)
552(33)
688(20)
869(14)
[a] Anions added as TBA salts. Association constants calculated using
WINEQNMR2[56] with estimated standard errors given in parentheses
(9:1 CDCl3/CD3OD, 293 K).
Given that the increased anion binding affinity of 17·PF6
and 18·PF6 over 1·PF6 is not due to increased binding
strength of the axle component, and that the macrocycle is
identical for all three rotaxanes, it appears this is due to
a greater degree of inter-component preorganisation of the
interlocked receptor. Although in the solid state the macro-
cycle resides over the triazolium group of the axle compo-
nent, in solution the rotaxane is dynamic with the wheel
moving about the axle. While such molecular motion is in-
evitable to some extent, the smaller and more compact the
axle component, the fewer the degrees of freedom the mac-
rocycle has available to engage in such motion. In 17·PF6
and 18·PF6, the macrocycle is forced to remain reasonably
close to the triazolium group by the proximal bulky stopper
groups and this dramatically increases halide anion binding
affinity in comparison to the less preorganised and more
flexible bis-alkyl-substituted triazolium axle containing ro-
taxane 1·PF6 (this was also investigated further using molec-
ular dynamics calculations, vide infra).
To investigate whether this effect could be amplified still
further, and to study the possibility of increasing the selec-
tivity for the smaller halide anions, chloride and bromide
over larger iodide, a new rotaxane consisting of the bis-aryl-
substituted triazolium axle component 4+ and a smaller
sized macrocycle was targeted. It was hoped that a smaller
macrocycle may further minimise motion of this component,
effectively “trapping” it in an ideal co-conformation for
anion complexation.
Scheme 5. Synthesis of macrocycle precursor 19. Conditions and re-
agents: a) 2-(allyloxy)ethyl methanesulfonate, K2CO3, acetonitrile, reflux,
89%; b) LiAlH4, Et2O, RT to reflux, 81%; c) 5-nitroisophthaloyl dichlor-
ide, NEt3, CH2Cl2, 91%.
The anion recognition properties of 22·PF6 were investi-
1
gated by H NMR titration experiments following the proto-
col used for 17·PF6 and 18·PF6. Addition of halide anions
caused downfield shifts in resonances a and 2, corresponding
À
to the triazolium signal and interior C H macrocycle
proton, respectively (Figure 6). Notably, these shifts are
smaller in magnitude than in rotaxanes 17+ and 18+, which
contain the larger macrocycle.
As shown in Table 3, 22·PF6 retains the impressive selec-
tivity for halide anions over acetate, with addition of the ox-
oanion causing negligible movement of the cavity protons.
However, the anion affinity of 22·PF6 is significantly lower
than that of 17·PF6 and 18·PF6, which can be partially ex-
Synthesis of a rotaxane containing a smaller sized macrocy-
cle component: The new vinyl-appended macrocycle precur-
sor 19 was prepared in high overall yield in three steps start-
ing from 4-cyanophenol (Scheme 5). In brief, this nitrile was
treated with 2-(allyloxy)ethyl methanesulfonate in refluxing
acetonitrile to afford 20. Reduction using LiAlH4 in diethyl
ether gave vinyl-appended amine 21, which was condensed
with 5-nitroisophthaloyl dichloride in dichloromethane to
give the target isophthalamide precursor 19.
Grubbsꢀ-catalysed RCM reaction of macrocycle precursor
19 and triazolium axle 4·Cl gave the novel rotaxane 22·Cl.
Purification by preparative thin layer chromatography, fol-
lowed by recrystallisation gave the novel rotaxane in 40%
yield (Scheme 6). The chloride anion template was ex-
changed for hexafluorophosphate by repeated washing with
1
NH4PF6(aq.) and the rotaxane was characterised by H, 13C,
19F, 31P NMR spectroscopy, high resolution ESI mass spec-
trometry, 1H-1H 2D ROESY NMR spectroscopy (see the
Supporting Information) and X-ray single crystallography.
Scheme 6. Synthesis of rotaxane 22·PF6. Conditions and reagents: a) 19,
Grubbsꢀ II catalyst, CH2Cl2, 40%; b) NH4PF6(aq.), CH2Cl2, 79%.
17756
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17751 – 17765

Halide Recognition by Triazolium Containing Rotaxanes
FULL PAPER
ening of these favourable “secondary” interactions between
the axle and macrocycle.
Molecular modelling
DFT calculations; interpretation of the anion recognition
properties of the axle components: The experimental results
indicate that, among the three axles, 4+ appears to have the
least electrostatic attraction for anions. In order to elucidate
this feature, the surface-electrostatic potential maximum
À
(VS,max) at the 1,2,3-triazolium C H hydrogen bond donor
site was estimated by means of DFT calculations. This pa-
rameter is known to correlate very well with hydrogen
bonding capacity[60,61] and acidity.[62] Thus, the isolated struc-
tures of 2+, 3+ and 4+ were optimised at the B3LYP/6-
31G** and the VS,max values were calculated at the same
level of theory (as described in the Supporting Information).
Furthermore, since the synthesised [2]rotaxanes depend on
chloride anion templation, we will focus the subsequent
analysis on the comparison between the experimentally de-
termined axle chloride association constants and the com-
puted VS,max values gathered together in Table 4.
Figure 6. Partial 1H NMR spectra of 22·PF6 upon addition of TBA·Cl.
Peak labelled x corresponds to residual CHCl3 solvent signal, unlabelled
peaks belong to bulky stopper groups (1:1 CDCl3/CD3OD, 293 K,
500 MHz).
Table 3. Association constants, Ka [mÀ1], for association of rotaxane
22·PF6 with anions.
Table 4. Surface electrostatic potential maximum at the triazole C H hy-
À
drogen bond donor, VS,max [kcalmolÀ1] for axles 2+, 3+ and 4+ together
with the corresponding association constants, Ka [mÀ1], for chloride.
ClÀ
BrÀ
IÀ
OAcÀ
N.B.[b]
2+
3+
4+
98.17
552(33)
[a]
22·PF6
538(55)
612(49)
402(12)
VS,max
Ka
107.29
735(55)
101.40
701(38)
[a] Anions added as TBA salts. Association constants calculated using
WINEQNMR2[56] with estimated standard errors given in parentheses
(1:1 CDCl3/CD3OD, 293 K). [b] Shifts too small to calculate association
constant (<0.05 ppm after 10 equiv of anion).
The VS,max values follow the trend of the association con-
stants. Indeed, the axle 2+ with the triazolium motif separat-
ed from the aryl substituents by aliphatic linkages displays
the most positive VS,max (107.29 kcalmolÀ1) and therefore it
is electrostatically more favourable for anion binding, lead-
ing to the largest Ka. On the other hand, 4+ with the triazoli-
um central unit directly connected to two aryl stoppers has
the lowest Ka and the smallest VS,max (98.17 kcalmolÀ1). The
VS,max value for 3+ (101.40 kcalmolÀ1), containing only one
aryl bulky stopper directly linked to the triazolium unit
À
through a C N bond, is between the VS,max values for the
other two axle components. Thus, on the basis of purely
À
electrostatic arguments the 1,2,3-triazole C H proton is
Figure 7. Structure of 22·Cl in the crystal. Solvent molecules, hydrogen
atoms, except those on triazolium and amide groups, and end-to-end dis-
order of axle component are omitted for clarity.
more acidic in 2+ and hence this axle is the most potent hy-
drogen bond donor showing the importance of the phenyl
rings in delocalising positive charge away from the heteroar-
omatic ring.
plained by analysis of the X-ray single crystal structure of
22·Cl (Figure 7). It would appear from this structure that the
macrocyclic component is too small to allow the chloride
anion to be encapsulated within the interlocked host cavity
and concomitantly facilitate favourable donor–acceptor in-
teractions between the phenolic rings of the macrocycle and
the axle componentꢀs triazolium motif. It is suggested that
the observed decreased halide affinity of 22·PF6 arises be-
cause complexation of the halide anion necessitates a weak-
Unconstrained MD simulations and umbrella sampling simu-
lations: In order to evaluate the dynamic behaviour at the
molecular level of the [2]rotaxane complexes in solution, un-
constrained MD simulations were initially performed on
[2]rotaxanes 17·Cl, 18·Cl, 22·Cl, and 1·Cl under periodic con-
ditions in explicit CHCl3/CH3OH (1:1) solvent mixture using
the Amber12[63] GPU accelerated code.[64,65] The macrocyle
and axle components of the interlocked structures were de-
Chem. Eur. J. 2013, 19, 17751 – 17765
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17757

P. D. Beer et al.
scribed with the general Amber force field (GAFF) parame-
ters[66] and restricted electrostatic potential (RESP) charges,
while for chloride a À1 net charge was used with van der
Waals parameters taken from ref. [67]. The remaining force
field parameters together with the simulation details are
given in the Supporting Information.
Three replicate simulations were performed for each
system with data collection periods of 50 ns each, which
were preceded by equilibration stages as detailed in the
Supporting Information. The corresponding trajectories
were clustered and the subsequent analysis is based on the
most populated cluster using a representative co-conforma-
tion, as exemplified for a replicate simulation of system
18·Cl (Figure 8). Representative co-conformations for the
affinity of the interlocked host 18·PF6 towards chloride
(Table 1). Rotaxane 17·PF6 has a similar association con-
stant for this anion, though slightly lower. Therefore, for this
system, the representative frames of the three simulations
exhibit equivalent binding scenarios (Figure S22 in the Sup-
porting Information), but with some structural differences.
The triazolium···hydroquinone and triazolium-methyl···po-
lyether interactions are not observed, as the triazolium
moiety tends to stay outside the macrocycle plane. This was
also observed in the X-ray structure reported in Figure 3.
Nevertheless, the dynamic behaviour given by the variations
À
À
in N H···Cl and C Htriazolium···Cl distances for 50 ns (Fig-
ure S26 in the Supporting Information) shows that both in-
teractions are sporadically interrupted in replicates 2 and 3,
whereas in replicate 1 the chloride leaves the rotaxane host
cavity at least twice.
Rotaxane 22·Cl behaves more like 17·Cl than 18·Cl, de-
spite containing the same axle component as the latter. Ap-
parently, the smaller macrocycle size in 22·Cl forces the tria-
zolium motif to be preferentially positioned outside of the
macrocycle where the anion recognition occurs (Figure S24
in the Supporting Information) for the three replicates. As
monitored before for 17·Cl, the anion binding to the inter-
locked receptor 22+ is interrupted throughout the course of
the MD simulations, in particular in replicate 2 (Figure S28
in the Supporting Information). Again, this molecular dy-
namical behaviour is consistent with the experimental bind-
ing data.
Concerning the simulations with 1·Cl, the MD simulations
show different structural findings among replicates. Fig-
ure S29 in the Supporting Information (left view) shows that
in the first replicate the chloride rotaxane association is not
preserved after the last 20 ns of the simulation time, with 1+
adopting preferentially the co-conformation depicted in Fig-
ure S25 in the Supporting Information (left view). In repli-
cates 2 and 3 the anion binding occurs either through
Figure 8. Representative frame of the MD simulation of 18·Cl carried out
À
À
in CHCl3/CH3OH in solution (replicate 1) with C Htriazolium···Cl and N
H···Cl hydrogen bonds drawn as dark blue and purple and dashed lines,
respectively. Solvent molecules and hydrogen atoms apart from those on
triazolium and amide groups were omitted for clarity.
À
a single C H···Cl interaction with the axle triazolium motif
À
or through the C H···Cl interaction operating together with
À
two N H···Cl hydrogen bonds formed with the macrocycle,
individual replicates of all systems are presented in Figur-
es S22–S25 in the Supporting Information. The evolution of
respectively (Figure S25 in the Supporting Information,
centre and right views). Furthermore, as reported for 17·Cl
and 22·Cl, in replicate 2 the binding between the chloride
and the interlocked receptor is also interrupted for a signifi-
cant period of the simulation time, the last 10 ns as shown in
Figure S29 in the Supporting Information (middle plot),
whereas in replicate 3, the binding scenario composed of
three hydrogen bonds is maintained on an intermittent basis
(right plot). This more labile behaviour reported for the
three replicates is in agreement with the lower association
constant compared with 17·Cl and 18·Cl.
À
À
the N H···Cl and C Htriazolium···Cl distances throughout the
course of the MD simulations for three replicates are also
shown in Figures S26–S29 in the Supporting Information for
17·Cl, 18·Cl, 22·Cl, and 1·Cl, respectively.
There are remarkable similarities between the representa-
tive frame of three MD replicates of 18·Cl (Figure S23 in
the Supporting Information) and the X-ray structure pre-
sented in Figure 3, featuring the same type of triazolium···-
hydroquinone and triazolium-methyl···polyether interactions.
The halide anion is concomitantly recognised by both the
The experimental results and the unconstrained MD sim-
ulations detailed above indicate that the chloride anion
binding strength is due to the degree of intercomponent pre-
organisation of the interlocked receptor, that is, the degree
of movement of the axle component inside the macrocycle.
Additional computational data supporting this hypothesis
were obtained from unconstrained MD simulations also per-
À
À
triazolium C H donor and the two N H amide binding sites
of the macrocycle and is relatively embedded in the rotax-
À
ane cavity. All three replicates yield similar results with C
Htriazolium···Cl interactions intermittently maintained (Fig-
ure S27 in the Supporting Information) for 50 ns of simula-
tion time, which is in agreement with the observed greater
17758
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17751 – 17765

Halide Recognition by Triazolium Containing Rotaxanes
FULL PAPER
Figure 9. Superposition of the representative frames of the three replicate simulations of 17·PF6, 18·PF6, 22·PF6, and 1·PF6 illustrating the conformational
variability among systems. The macrocycle (5 or 6) was fitted and used as reference. The carbon atoms of the axles in replicate 1, 2 and 3 are drawn in
red, blue and grey, respectively. In all replicates the macrocyle is in yellow.
formed on the interlocked hosts 17·PF6, 18·PF6, 22·PF6, and
1·PF6. The representative co-conformations for each repli-
cate and each system are presented in Figure 9 after fitting
the corresponding macrocycle.
the order 1·PF6 >17·PF6 >18·PF6 mirroring the increasing of
the axle length 2+ >3+ >4+. Furthermore, in agreement
with the preorganisation hypothesis, the former order is re-
verse of the experimental chloride binding affinity order
1·PF6 <17·PF6 <18·PF6.
Further insights into the rotaxane preorganisation assem-
bly were obtained by the evaluation of the free energy
change derived from the translational motion of the axle
inside the macrocycle. Umbrella sampling simulations[68]
were performed on interlocked receptor systems 17·PF6,
18·PF6, 22·PF6, and 1·PF6, followed by the extraction of the
corresponding potential of mean force (PMF). The compu-
tational details of these constrained MD simulations are
also given in the Supporting Information.
It is easily observed that the co-conformational variability
of 22·PF6 composed of the smaller macrocycle (5) and small-
er axle (4+) is the lowest since the three replicates yield
very similar representative co-conformations that are almost
superimposable. Indeed, through the course of the three
MD simulation replicates the methyl group oscillates out-
side the macrocycle, pointing towards the oxygen atoms
from the ether loop. In contrast, in some replicates of the
systems 17·PF6, 18·PF6 and 1·PF6 having the larger macrocy-
cle 6, a complete circumrotation of the axle inside the mac-
rocycle is observed as illustrated in Figure S30 in the Sup-
porting Information for replicate 2 of 18·PF6.
The PMF profiles along the reaction coordinate defined
as the distance between the centre of mass of the both hy-
droquinone rings of the macrocycle (COMhyd) and the qua-
ternary carbon atom of the bulky stopper (Cstopper) are
Comparing the systems 17·PF6, 18·PF6 and 1·PF6, the con-
formational variability among replicates varies according to
Chem. Eur. J. 2013, 19, 17751 – 17765
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17759

P. D. Beer et al.
Figure 10. PMF profile for the translational movement of the axle along the macrocycle component of the rotaxane systems 17·PF6, 18·PF6, 22·PF6, and
1·PF6. The points along the PMF corresponding to significant rotaxane co-conformations are identified together with their relative energies.
shown in Figure 10 for all four rotaxane systems. Represen-
tative frames taken from the constrained MD simulation
windows and relative to important points throughout the
PMF of each system are represented along the Figures S35–
S38 in the Supporting Information. Overall, the PMFs for
the three interlocked receptors display an asymmetric pro-
file around the corresponding global minimum, denoted as
a, reflecting the asymmetric structure around the triazolium
motif of 2+, 3+ and 4+.
The systems 18·PF6 and 22·PF6, having the same axle
component and macrocycles with different cavity sizes show
similar free energy profiles with a minima at 5.36 and
5.56 Š, respectively, in the curves of these two rotaxanes. In
both cases, the minimum corresponds to a co-conformation
in which the triazolium ring is partially sandwiched between
2.85 and 3.04 kcalmolÀ1 for 18·PF6 and 22·PF6, respectively.
These points are indicated as b in the corresponding PMF
profiles and are associated with co-conformations in which
À
the hydroquinone rings are over the C N bond connecting
the triazolium to the bulky stopper (Figure S36 in the Sup-
porting Information, central view, for 18·PF6 and Figure S37
in the Supporting Information, right view, for 22·PF6). In the
case of 18·PF6 a second local minimum c was found at a re-
action coordinate distance of 9.62 Š, which is 1.19 kcalmolÀ1
higher on the PMF relative to the global minimum. This cor-
responds to a co-conformation in which the triazolium motif
is located outside the macrocyclic cavity adopting a spatial
disposition approximately perpendicular to the stopper
À
phenyl ring linked by the C N bond (see right view in Fig-
ure S36 in the Supporting Information). Overall, the energy
barriers described are low indicating that the smaller axle 4+
is able to move inside the macrocycle along the
COMhyd···Cstopper PMF coordinate for a certain range of dis-
tances without a considerable energetic cost using approxi-
mately an harmonic potential. This range of distances is
about 4 to 10 Š for 18·PF6 and 4 to 9 Š for 22·PF6. Outside
of these ranges, the energy rises abruptly due to unfavoura-
ble steric interactions caused by the proximity of the bulky
stopper groups of 4+ (with a rigid structure) to the macrocy-
cle. In addition, this energy increase is larger for 22·PF6
À
the hydroquinone rings but slightly deviated towards the C
C bond connecting the triazolium moiety to the stopper
(Figure S36 in the Supporting Information for 18·PF6 and
Figure S37 in the Supporting Information for 22·PF6). It is
also evident that a phenyl ring of a stopper group of the
smaller axle 4+ is almost perpendicular to the triazolium
ring and as the COMhyd···Cstopper distance increases, the
energy increases. With the translational movement of the
axle, the hydroquinone rings of macrocycles 5 and 6 force
the rotation of that phenyl rings yielding energy barriers of
17760
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17751 – 17765

Halide Recognition by Triazolium Containing Rotaxanes
FULL PAPER
given the smaller size of the macrocycle 5 in comparison
with 6 in 18·PF6.
quiring the weakening of favourable donor–acceptor inter-
actions between the axleꢀs triazolium group and electron-
rich rings in the macrocycle.
The effect of varying the length of the axle component on
the energetics of its translational motion through macrocycle
6 is conveniently assessed through the comparison of the
PMFs of 1·PF6 and 18·PF6 with longer and smaller axles, re-
spectively, followed by the analysis of the PMF of 17·PF6.
The PMF of 1·PF6 (Figure 10) displays a global minimum at
9.35 Š (a) and a 2.81 kcalmolÀ1 barrier at 12.90 Š (b) with
the corresponding co-conformations depicted in Figure S38
in the Supporting Information. The minimum (left view) is
associated with a co-conformation in which the two hydro-
quinone rings stack the triazolium motif whereas the barrier
shows a co-conformation in which the hydroquinone rings
À
The VS,max values at the 1,2,3-triazolium C H hydrogen
bond donor site, estimated by DFT, for the three axle com-
ponents and molecular dynamics simulations of the three ro-
taxane receptors and their chloride complexes corroborate
the experimental observations. Free-energy profiles associat-
ed with the translational motion of the axle component
along the macrocycle were also estimated for these three in-
terlocked systems from umbrella sampling simulations,
which are also in agreement with experimental findings.
This work demonstrates that for complex three-dimen-
sional hosts, such as rotaxanes, guest complexation strength
and selectivity depends not only on having appropriately
preorganised recognition elements within each component
of the interlocked architecture, but also having these compo-
nents effectively preorganised with respect to each other.
À
are almost over the C C bond of the axle attaching the
bulky stopper to the triazolium unit from the N terminus
(right view). More importantly, besides the barrier being
slightly lower than the one found for 18·PF6 the abrupt
energy increase for long COMhyd···Cstopper distances, only
occurs at values above 15 Š, meaning that the axle 2+ in
1·PF6 has an increased mobility when compared with 1+ in
18·PF6.
Experimental Section
The PMF shape of 17·PF6 indicates that the translation
motion of 3+ inside this interlocked structure is easier than
in 18·PF6. Indeed the energy barrier between the global
minimum a and local maximum b for 17·PF6 (corresponding
structures shown in Figure S35 in the Supporting Informa-
tion), is only 1.41 kcalmolÀ1 whereas in 18·PF6 this free
energy difference is 2.85 kcalmolÀ1. This result is expected
considering the intermediate length of axle 3+, between 1+
and 2+ (Figure 2). Nevertheless, the PMF of 17·PF6 also
presents a considerable energy rise for COMhyd···Cstopper dis-
tances outside the range between 4 and 12 Š. Thus, in the
analysis of the free energy profiles of 17·PF6, 18·PF6 and
1·PF6 rotaxanes, the axle in the latter one appears with
a greater mobility in agreement with the intercomponent
preorganisation hypothesis suggested by the experimental
binding data and unconstrained molecular dynamics simula-
tions.
Bulky stopper groups 9[50] 11,[51] and 15,[55] macrocycle precursor 16,[69]
and TBTA[70] were prepared as previously described. “Dry” solvents
were purged with nitrogen and passed through an MBraun MPSP-800
column. Water was de-ionised and microfiltered using a Milli-Q Millipore
machine. Tetrabutylammonium salts were stored in a vacuum dessicator.
Chromatography was performed on silica gel (particle size: 40–63 mm).
Details of instrumentation used, 1H and 13C NMR spectra of new com-
pounds, and NMR titration protocols are provided in the Supporting In-
formation.
Single crystal X-ray diffraction data were collected using graphite mono-
chromated CuKa radiation (l=1.54184 Š) or synchrotron radiation (l=
0.68889 Š) on Beamline I19 of Diamond Lightsource.[71] Cell parameters
and intensity data (including interframe scaling) were processed using
CrysAlis Pro.[72] The structures were solved by charge-flipping methods
using SUPERFLIP[73] and refined using full-matrix least-squares on F2
within the CRYSTALS suite.[74] Full details are provided in the Support-
ing Information.
Azide-stopper 7: Aniline stopper 9, (1.01 g, 2.00 mmol) was suspended in
9:1 ethanol/conc. HCl(aq.) (500 mL), and the mixture was gently warmed,
causing all material to dissolve. This was cooled to 08C and sodium ni-
trite (0.276 g, 4.00 mmol) was added, followed by sodium azide (0.260 g,
4.00 mmol). The reaction was stirred under a nitrogen atmosphere for 1 h
at 08C, before being allowed to warm to room temperature, overnight.
The reaction was diluted with water (200 mL), and extracted with diethyl
ether (2”200 mL). The combined organic fractions were washed with sa-
turated aqueous sodium carbonate (2”150 mL) and brine (150 mL),
dried (magnesium sulfate) and taken to dryness under reduced pressure.
The resulting white powder was taken up in 1:1 petrol/dichloromethane
and filtered through a short pad of silica, and washed through with fur-
ther 1:1 petrol/dichloromethane. Evaporation of the solvent gave pure 7
as a white powder. Yield: 0.876 g (83%).
Conclusion
We have investigated the effect of intercomponent preorga-
nisation on the anion binding affinities of three new triazoli-
um-containing rotaxane host architectures. Increasing preor-
ganisation by decreasing the length and increasing the rigidi-
ty of the triazolium-containing axle component, and thus re-
ducing the ability of the macrocycle to move about this axle
results in significantly increased halide anion binding affini-
ty. Control experiments demonstrate that this stronger anion
recognition is not due to the nature of the axle component
itself, as bis-alkyl-substituted, alkyl-aryl-substituted and bis-
aryl-substituted triazolium containing axles all displayed in-
distinguishable halide binding strengths. A rotaxane contain-
ing a smaller macrocycle component exhibited diminished
anion affinity, which was attributed to guest recognition re-
3
3
1H NMR (CDCl3): d=7.24 (d, J=8.6 Hz, 6H Ar-H), 7.17 (d, J=8.8 Hz,
2H, Ar-H), 7.07 (d, 3J=8.6 Hz, 6H, Ar-H), 6.90 (d, 3J=8.8 Hz, 2H, Ar-
H), 1.30 ppm (s, 54H, tBu-H); 13C NMR (CDCl3): d=148.7, 144.5, 143.8,
137.4, 132.7, 130.8, 124.3, 117.9, 63.4, 34.5, 31.5 ppm; HRMS (EI/FI):
529.3448, calcd for [C37H43N3]: 529.3457.
Stopper triflate 12: The aryl alcohol 11, (1.51 g, 3.00 mmol) was suspend-
ed in dry dichloromethane (30 mL). Triethylamine (0.83 mL, 0.61 g,
6.0 mmol) was added, causing all material to dissolve, followed by tri-
fluoromethanesulfonic anhydride (0.61 mL, 1.0 g, 3.6 mmol) in dry di-
chloromethane (5 mL). The reaction was stirred at room temperature
Chem. Eur. J. 2013, 19, 17751 – 17765
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17761

P. D. Beer et al.
under a nitrogen atmosphere for 1 h. It was then diluted with diethyl
ether (100 mL) and washed sequentially with HCl(aq.) (1.0m, 100 mL), sa-
turated NaHCO3(aq.) (100 mL) and brine (100 mL), and dried (magnesium
sulfate). Column chromatography (1:1 petrol/dichloromethane) gave 12
as a white crystalline solid. Yield: 1.71 g (90%).
1H NMR (CDCl3): d=7.21–7.29 (m, 8H, Ar-H), 7.09–7.16 (m, 2H, Ar-
H), 7.01–7.08 (m, 6H, Ar-H), 1.30 ppm (s, 27H, tBu-H); 19F NMR
(CDCl3): d=73.0 ppm (s); 13C NMR (CDCl3): d=149.0, 148.1, 147.6,
143.3, 133.1, 130.8, 124.5, 120.0, 63.6, 34.5, 31.5 ppm; HR-MS (EI/FI):
636.2885; calcd for [C38H43O3F3S]: 636.2885.
being cooled to room temperature, it was repeatedly passed down an
Amberliteꢄ ClÀ anion exchange column. The acetone was removed under
reduced pressure, and the aqueous phase was extracted with dichlorome-
thane (3”20 mL), the combined organic fractions were dried (magnesi-
um sulfate) and taken to dryness to give 3·Cl as a white powder. Yield:
0.039 g (86%).
1H NMR (CDCl3): d=10.59 (s, 1H, trz+-H), 7.88 (d, 3J=8.8 Hz, 2H, Ar-
H adjacent to trz+), 7.50 (d, 3J=8.8 Hz, 2H, Ar-H), 7.22–7.28 (m, 12H,
Ar-H), 7.16 (d, 3J=8.9 Hz, 2H, Ar-H), 7.04–7.08 (m, 12H, Ar-H), 6.93
3
(d, J=8.9 Hz, 2H, Ar-H), 5.70 (s, 2H, Ph-O-CH2-trz+), 4.51 (s, 3H, trz+
-CH3), 1.31 (s, 27H, tBu-H), 1.29 ppm (s, 27H, tBu-H); LRESI-MS
Stopper TIPS-alkyne 10: The aryl triflate 12, (1.27 g, 2.00 mmol), copper
iodide (0.076 g, 0.40 mmol) and palladium(0) tetrakis(triphenylphos-
phine) (0.232 g, 0.200 mmol) were placed in a microwave vial. Dry N,Nꢀ-
dimethylformamide (20 mL), triethylamine (1.67 mL, 1.21 g, 12.0 mmol)
and triisopropylsilyl acetylene (0.54 mL, 0.44 g, 2.4 mmol) were added;
the suspension was deoxygenated by bubbling nitrogen through it, and
the vial was sealed. It was heated to 758C, overnight, cooled to room
temperature and filtered through cotton wool, and the residue was
washed thoroughly with toluene. The combined filtrates were taken to
dryness under reduced pressure. The resulting brown oil was taken up in
toluene (50 mL), and washed with HCl(aq.) (1.0m, 2”50 mL) and brine
(50 mL), and dried (magnesium sulfate). Purification by column chroma-
tography (silica, 24:1 petrol/dichloromethane) gave 10 as a white powder.
Yield: 0.815 g (61%).
(pos.): 1086.71, calcd for [C78H92N3O]+: 1086.72.
Triazole axle 14: The protected alkyne 10, (0.368 g, 0.550 mmol) and the
aryl azide 7, (0.265 g, 0.500 mmol) were dissolved in dry dichloromethane
(25 mL). DIPEA (0.17 mL, 0.13 g, 1.0 mmol), TBTA (0.265 g,
0.500 mmol) and [CuI
ACHTUNGRTENNUG(CH3CN)4]AHCTUNGTRNEN(UGN PF6) (0.186 g, 0.500 mmol) were added,
and the yellow solution was deoxygenated by bubbling nitrogen gas
through it. TBA·F (1.0m in tetrahydrofuran, 1.5 mL) was added and the
solution was stirred at room temperature under a nitrogen atmosphere
for 2 days. It was washed with aqueous NH4OH/EDTA solution (50 mL)
and the aqueous phase was extracted with dichloromethane (3”50 mL).
The combined organic layers were dried (magnesium sulfate) and puri-
fied by column chromatography (2:3 dichloromethane/petrol) to give 14
as a white powder. Yield: 0.444 g (85%).
1H NMR (CDCl3): d=7.33–7.38 (m, 2H, Ar-H adjacent to alkyne), 7.21–
7.25 (m, 6H, Ar-H), 7.13–7.17 (m, 2H, Ar-H), 7.06–7.11 (m, 6H, Ar-H),
1.30 (s, 27H, tBu-H), 1.11 (s, 18H, Si-CH-CH3), 1.10 ppm (s, 3H, Si-CH-
CH3); 13C NMR (CDCl3): d=148.6, 148.1, 143.7, 131.2, 131.1, 130.8,
124.4, 120.8, 107.4, 90.3, 63.9, 34.5, 31.5, 18.8, 11.5 ppm; MS (probe EI/
FI): 668.462, calcd for [C48H64Si]: 668.478.
1H NMR (CDCl3): d=8.10 (s, 1H, trz-H), 7.76 (d, 3J=8.4 Hz, 2H, Ar-H
adjacent to trz), 7.64 (d, 3J=8.6 Hz, 2H, Ar-H adjacent to trz), 7.39 (d,
3J=8.4 Hz, 2H, Ar-H), 7.21–7.32 (m, 14H, Ar-H), 7.10–7.14 (m, 12H,
Ar-H), 1.31 ppm (s, 54H, tBu-H); 13C NMR (CDCl3): d=148.9, 148.6,
148.3, 148.0, 143.8, 143.5, 134.8, 132.6, 131.9, 130.9, 130.8, 127.7, 124.9,
124.5, 124.4, 124.3, 119.4, 117.5, 63.8, 63.7, 34.5, 34.5, 31.5, 31.5 ppm;
HRESI-MS (pos.): 1064.6781, calcd for [C76H87N3·Na]+: 1064.6792.
Triazole axle 13: The azide 7, (0.159 g, 0.300 mmol) and alkyne 15,
(0.179 g, 0.330 mmol) were dissolved in dichloromethane (25 mL).
DIPEA (0.17 mL, 0.13 g, 1.0 mmol), TBTA (0.016 g, 0.030 mmol) and
Triazolium axle 4·BF4: The triazole axle 14, (0.261 g, 0.250 mmol) was
dissolved in dry dichloromethane (30 mL). Trimethyloxonium tetrafluor-
oborate (0.044 g, 0.30 mmol) was added and the colourless solution was
stirred at room temperature under a nitrogen atmosphere, overnight. It
was quenched by the addition of methanol (1 mL), taken to dryness
under reduced pressure and purified by column chromatography (2%
methanol in chloroform) to give 4·BF4 as a white powder. Yield: 0.242 g
(89%).
[CuI
ACHTUNGTRENNUNG(CH3CN)4]ACHTUNGTRENNUNG(PF6) (0.011 g, 0.030 mmol) were added and the clear
yellow solution was stirred under a nitrogen atmosphere for 4 days. It
was taken to dryness under pressure and purified by column chromatog-
raphy (silica, gradient: 3:2 dichloromethane/petrol; dichloromethane;
3% methanol in dichloromethane) to give 13 as a white powder. Yield:
0.286 g (89%).
1H NMR (CDCl3): d=7.99 (s, 1H, trz-H), 7.59 (d, 3J=8.5 Hz, 2H, Ar-H
adjacent to trz), 7.37 (d, 3J=8.5 Hz, 2H, Ar-H), 7.21–7.27 (m, 12H, Ar-
H), 7.05–7.11 (m, 14H, Ar-H), 6.87 (d, 3J=8.1 Hz, 2H, 2 ” H e), 5.26 (s,
2H, O-CH2-trz), 1.30 ppm (s, 54H, tBu-H); 13C NMR (CDCl3): d=156.2,
148.9, 148.8, 148.5, 145.2, 144.2, 143.4, 140.4, 134.7, 132.6, 132.5, 130.9,
130.8, 124.5, 124.2, 121.0, 119.6, 113.4, 63.7, 63.2, 62.2, 34.5, 34.4, 31.5,
31.5 ppm; HRESI-MS (pos.): 1094.6909, calcd for [C77H89N3O·Na]+:
1094.6898.
1H NMR (CDCl3): d=8.74 (s, 1H, trz-H+), 7.78–7.81 (m, 2H, Ar-H adja-
cent to trz+), 7.60–7.63 (m, 2H, Ar-H adjacent to trz+), 7.47–7.52 (m, 4H,
Ar-H), 7.25–7.29 (m, 12H, Ar-H), 7.06–7.11 (m, 12H, Ar-H), 4.37 (s, 3H,
trz+-CH3), 1.31 ppm (s, 54H, tBu-H); 13C NMR (CDCl3): d=152.6, 152.5,
149.1, 149.0, 143.0, 142.9, 133.2, 132.7, 132.5, 130.8, 130.7, 128.6, 126.2,
124.7, 124.6, 121.0, 118.9, 64.0, 63.9, 39.1, 34.5, 31.5 ppm; 19F NMR
(CDCl3): d=À153.9 ppm (m); HRESI-MS (pos.): 1056.7147, calcd for
[C77H90N3]+: 1056.7129.
Triazolium axle 3·BF4: The neutral axle component 13, (0.215 g,
0.200 mmol) was dissolved in dry dichloromethane (40 mL). Trimethy-
loxonium tetrafluoroborate (0.032 g, 0.22 mmol) was added, causing the
clear colourless solution to immediately turn lemon-yellow. It was stirred
at room temperature under a nitrogen atmosphere, and then methanol
(1 mL) was added to quench the alkylating agent, causing the reaction to
lose its colour. It was taken to dryness under reduced pressure and puri-
fied by column chromatography (silica, 4% methanol in chloroform) to
give 3·BF4 as a white powder. Yield: 0.223 g (95%).
Triazolium axle 4·Cl: The hexafluorophosphate salt of 4 (0.057 g,
0.050 mmol) was dissolved in 4:1 acetone/water with warming. After
being cooled to room temperature, it was repeatedly passed down an
Amberliteꢄ ClÀ anion exchange column. The acetone was removed under
reduced pressure, and the aqueous phase was extracted with dichlorome-
thane (3”20 mL), the combined organic fractions were dried (magnesi-
um sulfate) and taken to dryness to give 4·Cl as a white powder. Yield:
0.055 g (100%).
1H NMR (CDCl3): d=11.06 (s, 1H, trz-H+), 8.31 (d, 3J=8.8 Hz, 2H, Ar-
H adjacent to trz+), 7.89 (d, 3J=8.8 Hz, 2H, Ar-H adjacent to trz+),
7.50–7.55 (m, 4H, Ar-H), 7.22–7.27 (m, 12H, Ar-H), 7.05–7.10 (m, 12H,
Ar-H), 4.44 (s, 3H, trz+-CH3), 1.30 ppm (s, 54H, tBu-H); LRESI-MS
(pos.): 1056.67, calcd for [C77H90N3]+: 1056.71. LRESI-MS (neg.):
1126.65, calcd for [C77H90N3·2Cl]À: 1126.66.
3
1H NMR (CDCl3): d=8.94 (s, 1H, trz+-H), 7.70 (d, J=8.9 Hz, 2H, Ar-H
3
adjacent to trz+), 7.49 (d, J=8.9 Hz, 2H, Ar-H), 7.21–7.28 (m, 12H, Ar-
H), 7.15 (d, 3J=8.9 Hz, 2H, Ar-H), 7.03–7.08 (m, 12H, Ar-H), 6.85 (d, J
d,e=8.9 Hz, 2H, Ar-H adjacent to O), 5.47 (s, 2H, trz+-CH2-O), 4.45 (s,
3H, trz+-CH3), 1.30 (s, 27H, tBu-H), 1.29 ppm (s, 27H, tBu-H);
13C NMR (CDCl3): d=154.7, 152.9, 149.2, 148.6, 143.9, 142.0, 141.1,
133.4, 132.9, 130.8, 130.7, 124.7, 124.3, 120.4, 113.3, 77.4, 63.9, 63.2, 39.3,
34.5, 34.5, 31.5, 31.5 ppm; 19F NMR (CDCl3): d=À152.9 ppm (m);
HRESI-MS (pos.): 1086.7223, calcd for [C78H92N3O]+: 1086.7235.
Vinyl-appended nitrile 20: 2-(Allyloxy)ethyl methanesulfonate (4.15 g,
23.0 mmol) and 4-cyanophenol (3.29 g, 27.6 mmol) were suspended in dry
acetonitrile (100 mL). Potassium carbonate (6.36 g, 46.0 mmol) was
added and the suspension heated to reflux under a nitrogen atmosphere
for three days. It was taken to dryness under reduced pressure, the crude
product was dissolved in dichloromethane (100 mL) and washed with
Triazolium axle 3·Cl: The hexafluorophosphate salt of
3 (0.047 g,
0.040 mmol) was dissolved in 4:1 acetone/water with warming. After
17762
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17751 – 17765

Halide Recognition by Triazolium Containing Rotaxanes
FULL PAPER
HCl(aq.) (1.0m, 2”100 mL), brine (100 mL), dried (magnesium sulfate)
and taken to dryness under reduced pressure. The crude oil was purified
by column chromatography (1:1 petrol/dichloromethane) to give 20 as
a colourless oil. Yield: 4.16 g (89%).
1H NMR (CDCl3): d=7.55–7.60 (m, 2H, Ar-H), 6.95–7.00 (m, 2H, Ar-
H), 5.86–6.00 (m, 1H, O-CH2-CH=CH2), 5.20–5.34 (m, 2H, O-CH2-CH=
CH2), 4.18 (t, 3J=4.7 Hz, 2H, Ph-O-CH2-CH2-O), 4.07–4.10 (m, 2H, O-
CH2-CH=CH2), 3.82 ppm (t, J=4.7 Hz, 2H, Ph-O-CH2-CH2-O);
13C NMR (CDCl3): d=62.2, 134.4, 134.0, 119.3, 117.7, 115.4, 104.2, 72.5,
68.2, 67.9 ppm; HRESI-MS (pos.): 226.0837, calcd for [C12H13NO2·Na]+:
226.0838.
Rotaxane 17·PF6: Rotaxane 17·Cl (0.017 g, 0.010 mmol) was taken up in
dichloromethane (20 mL), washed with NH4PF6(aq.) (7”20 mL) and water
(2”20 mL). It was dried thoroughly in vacuo to give 17·PF6 as a white
powder. Yield: 0.018 g (100%).
1H NMR (CDCl3):d=8.90 (s, 2H, ext. Ar-H), 8.43 (s, 1H, trz+-H), 8.25
(s, 1H, int. Ar-H), 7.53 (d, 3J=8.6 Hz, 2H, Ar-H adjacent to trz+), 7.36
3
(d, J=8.6 Hz, 2H, Ar-H), 7.19–7.26 (m, 12H, Ar-H), 7.01–7.07 (m, 14H,
Ar-H), 6.85 (d, 3J=8.8 Hz, 2H, Ar-H-O-CH2), 6.41 (d, 3J=8.8 Hz, 4H,
hydroquinone Ar-H), 6.16 (d, 3J=8.8 Hz, 4H, hydroquinone Ar-H), 5.49
(brs, 2H, CH=CH), 4.63 (s, 2H, Ar-O-CH2-trz+), 4.21–4.28 (m, 4H,
amide-CH2-CH2-O-Ph), 3.56–3.96 (m, 19H, trz+-CH3, amide-CH2-CH2-
O-Ph, Ph-O-CH2-CH2-O, Ph-O-CH2-CH2-O, O-CH2-CH=CH), 1.33 (s,
27H, tBu-H), 1.30 ppm (s, 27H, tBu-H); 13C NMR (CDCl3): d=164.6,
154.8, 153.3, 152.8, 152.2, 149.3, 149.0, 148.7, 144.0, 142.9, 142.3, 139.9,
136.5, 132.9, 132.7, 132.0, 130.8, 130.7, 130.5, 130.5, 129.7, 126.9, 124.8,
124.4, 120.4, 115.3, 114.2, 113.5, 70.7, 69.3, 67.7, 66.4, 64.0, 63.3, 40.6, 34.6,
Vinyl-appended amine 21: The nitrile 20 (2.03 g, 10.0 mmol) was dis-
solved in dry diethyl ether (50 mL) and added dropwise to a suspension
of lithium aluminium hydride (0.76 g, 20 mmol) in dry diethyl ether
(50 mL). Once addition was complete, the suspension was heated to
reflux under a nitrogen atmosphere for three hours. It was then cooled in
an ice bath, and quenched by the addition of magnesium sulfate decahy-
drate (3.22 g, 10.0 mmol). After bubbling had ceased, the suspension was
filtered and the colourless filtrate was taken to dryness under reduced
pressure. It was dissolved in 225:24:1 dichloromethane/methanol/triethyl
amine and passed through a short pad of silica, and then evaporated to
dryness under reduced pressure. Thorough drying in vacuo gave 21 as
a colourless oil that solidified on standing. Yield: 1.67 g (81%).
1H NMR (CDCl3): d=7.19–7.24 (m, 2H, Ar-H), 6.87–6.92 (m, 2H, Ar-
H), 5.88–6.01 (m, 1H, O-CH2-CH=CH2), 5.18–5.34 (m, 2H, O-CH2-CH=
CH2), 4.07–4.14 (m, Ph-O-CH2-CH2-O, O-CH2-CH=CH2), 3.78–3.81 ppm
(m, 4H, Ph-CH2-NH2, Ph-O-CH2-CH2-O); 13C NMR (CDCl3): d=157.8,
135.9, 134.7, 128.3, 117.5, 114.8, 72.5, 68.7, 67.6, 46.0 ppm; HRMS (EI/
FI): 207.1294, calcd for [C12H17NO2]: 207.1259.
34.5, 31.5, 31.5, 29.9 ppm; 19F NMR (CDCl3): d=À69.8 ppm (d, JP, F
=
716 Hz); HRESI-MS (pos.): 1707.9565, calcd for [C110H127N6O11]+:
1707.9567.
Rotaxane 18·Cl: The axle 4·Cl (0.044 g, 0.040 mmol) and macrocycle pre-
cursor 16 (0.039 g, 0.060 mmol) were dissolved in dry dichloromethane
(10 mL). Grubbsꢀ II catalyst (0.0039 g, 10% by weight) was added and
the mixture was stirred at room temperature under a nitrogen atmos-
phere for 3 days. Purification by preparative TLC (2% methanol in di-
chloromethane) gave slightly impure rotaxane. This was recrystallised
from boiling methanol/chloroform (25 mL, ~4:1), and the resulting solid
was isolated by filtration and washed with cold methanol (8 mL) to give
pure 18·Cl as a white powder. Yield: 0.039 g (57%).
1H NMR (CDCl3): d=10.46 (s, 1H, trz+-H), 9.82 (s, 1H, int. Ar-H), 9.16
(s, 2H, amide-H), 8.76 (s, 2H, ext Ar-H), 8.09 (d, 3J=8.8 Hz, 2H, Ar-H
adjacent to trz+), 7.53 (d, 3J=8.6 Hz, Ar-H adjacent to trz+), 7.39 (d,
3J=8.6 Hz, 2H, Ar-H), 7.22–7.32 (m, 14H, Ar-H), 7.12 (d, 3J=8.6 Hz,
6H, Ar-H), 6.93 (d, 3J=8.6 Hz, 6H, Ar-H), 6.16 (d, 3J=9.1 Hz, 4H, hy-
droquinone Ar-H), 6.05 (d, 3J=9.1 Hz, 4H, hydroquinone Ar-H), 5.90
(brs, 2H, CH=CH), 4.24 (s, 3H, trz+-CH3), 3.54–3.99 (m, 20H, amide-
CH2-CH2-O-Ph, amide-CH2-CH2-O-Ph, Ph-O-CH2-CH2-O, Ph-O-CH2-
CH2-O, O-CH2-CH=CH), 1.35 (s, 27H, tBu-H), 1.33 ppm (s, 27H, tBu-
H); LRESI-MS (pos.): 1677.89, calcd for [C109H125N6O10]+: 1677.95.
Vinyl-appended macrocycle precursor 19: 5-Nitroisophthalic acid
(0.422 g, 2.00 mmol) was suspended in thionyl chloride (10 mL). Pyridine
(5 drops) were added and the colourless suspension heated to reflux
under a nitrogen atmosphere, overnight. Excess thionyl chloride was re-
moved under reduced pressure and the crude acid chloride was taken up
in dry dichloromethane (30 mL). Triethyl amine (0.83 mL, 0.61 g,
6.0 mmol) was added, followed by 21 (0.871 g, 4.20 mmol) in dry di-
chloromethane (10 mL). The reaction was stirred at room temperature
under a nitrogen atmosphere for 3 h, and then washed with 10% HCl(aq.)
(2”40 mL) and brine (40 mL). It was dried (magnesium sulfate) and pu-
rified by column chromatography (3% methanol in chloroform) to give
19 as an off-white solid. Yield: 1.08 g (91%).
Rotaxane 18·PF6: Rotaxane 18·Cl (0.031 g, 0.018 mmol) was taken up in
dichloromethane (20 mL), and washed with NH4PF6(aq.) (7”20 mL) and
water (2”20 mL). It was dried thoroughly in vacuo to give 18·PF6 as
a white powder. Yield: 0.035 g (100%).
1H NMR (CDCl3): d=8.66 (d, 4J=1.4 Hz, 2H, ext. Ar-H), 8.52 (t, 4J=
1.4 Hz, 1H, int Ar-H), 7.42 (3J=5.4 Hz, 2H, amide-H) 7.11 (d, 3J=
8.6 Hz, 4H, CH2-Ar-H), 6.69 (d, 3J=8.6 Hz, 4H, O-Ar-H), 5.84–5.98 (m,
1H NMR ([D6]DMSO): d=9.12 (s, 1H, int. Ar-H), 9.07 (s, 1H, trz+-H),
8.80 (s, 2H, ext. Ar-H), 7.44–7.47 (m, 4H, Ar-H adjacent to trz+), 7.32–
3
3
7.39 (m, 14H, Ar-H), 7.24 (d, J=8.7 Hz, 2H, Ar-H), 7.11 (d, J=8.5 Hz,
6H, Ar-H) 7.03 (d, 3J=8.5 Hz, 6H, Ar-H), 6.35 (d, 3J=9.0 Hz, 4H, hy-
droquinone Ar-H), 6.20 (d, 3J=9.0 Hz, 4H, hydroquinone Ar-H), 5.65
(brs, 2H, CH=CH), 4.24 (s, 3H, trz+-H), 3.81–3.88 (m, 8H, amide-CH2,
O-CH2-CH=CH), 3.52–3.68 (m, 12H, amide-CH2-CH2-O-Ph, Ph-O-CH2-
CH2-O, Ph-O-CH2-CH2-O), 1.27 ppm (s, 54H, tBu-H); 13C NMR
([D6]DMSO): d=163.7, 152.1, 152.0, 148.2, 148.2, 147.8, 143.1, 143.0,
142.5, 135.6, 131.6, 131.3, 131.1, 129.9, 128.6, 128.2, 125.4, 124.7, 124.7,
124.3, 119.9, 118.9, 114.9, 114.4, 70.0, 68.3, 67.3, 67.0, 66.5, 63.4, 63.3, 34.1,
34.1, 31.1, 31.1 ppm; HRESI-MS (pos.): 1677.9452, calcd for
[C109H125N6O10]+: 1677.9452.
3
2H, O-CH2-CH=CH2), 5.17–5.33 (m, 4H, O-CH2-CH=CH2), 4.43 (d, J=
5.4 Hz, 4H, amide-CH2-Ph), 4.03–4.06 (m, 4H, O-CH2-CH=CH2), 3.98 (t,
3J=4.7 Hz, 4H, Ph-O-CH2-CH2-O), 3.73 ppm (t, 3J=4.7 Hz, 4H, Ph-O-
CH2-CH2-O); 13C NMR (CDCl3): d=164.4, 158.3, 148.3, 136.2, 134.4,
131.2, 129.8, 129.4, 124.9, 117.9, 114.7, 72.5, 68.6, 67.4, 44.0 ppm; HRESI-
MS (pos.): 612.2321, calcd for [C32H35N3O8·Na]+: 612.2316.
Rotaxane 17·Cl: The triazolium axle 3·Cl (0.034 g, 0.030 mmol) and mac-
rocycle precursor 16 (0.029 g, 0.045 mmol) were dissolved in dry dichloro-
methane (10 mL). Grubbsꢀ II catalyst (0.0029 g, 10% by weight) was
added and the mixture was stirred under a nitrogen atmosphere for
7 days. Purification by preparative TLC (3% methanol in dichlorome-
thane) gave slightly impure rotaxane. This was recrystallised from boiling
Rotaxane 22·Cl: The axle 4·Cl (0.033 g, 0.030 mmol) and macrocycle pre-
cursor 19 (0.027 g, 0.045 mmol) were dissolved in dry dichloromethane
(5 mL). Grubbsꢀ II catalyst (5.4 mg 20% by weight) was added, and the
mixture was stirred at room temperature under a nitrogen atmosphere
for 3 days. It was taken to dryness and purified by preparative TLC (3%
methanol in dichloromethane) to give slightly impure rotaxane (yield at
this point ~55%). This was recrystallised from boiling methanol (4 mL)
to give a white solid, which was isolated by filtration, and washed with
cold methanol (2”1 mL) to give pure 22·Cl as a white powder. Yield:
0.020 g (40%).
methanol/dichloromethane (10 mL, ~9:1) to give 17·Cl as
a white
powder. Yield: 0.027 g (51%).
1H NMR (CDCl3): d=10.12 (brs, 1H, trz+-H), 9.86 (brs, 1H, int Ar-H),
9.16 (brs, 2H, amide-H), 8.86 (s, 2H, ext Ar-H), 7.71 (brs, 2H, Ar-H ad-
jacent to trz+), 7.23–7.30 (m, 18H, Ar-H), 7.09 (d, 3J=8.0 Hz, 6H, Ar-
H), 6.97 (d, 3J=8.0 Hz, 6H, Ar-H), 6.21 (brs, 8H, hydroquinone Ar-H),
5.81 (brs, 2H, CH=CH), 4.74 (s, 2H, Ph-O-CH2-trz+), 4.30 (s, 3H, trz+
-CH3), 3.53–4.11 (m, 20H, amide-CH2-CH2-O-Ph, amide-CH2-CH2-O-Ph,
Ph-O-CH2-CH2-O, Ph-O-CH2-CH2-O, O-CH2-CH=CH), 1.35 (s, 27H,
tBu-H), 1.30 ppm (s, 27H, tBu-H); LRESI-MS (pos.) 1707.78, calcd for
[C110H127N6O11]+: 1707.96.
1H NMR (1:1 CDCl3/CD3OD): d=9.33 (t, 3J=5.8 Hz, 2H, amide-H),
9.22 (s, 1H, trz+-H), 8.90 (s, 1H, int. Ar-H), 8.80 (s, 2H, ext. Ar-H), 7.69
3
3
(d, J=8.8 Hz, 2H, Ar-H adjacent to trz+), 7.42 (d, J=8.8 Hz, 2H, Ar-H
Chem. Eur. J. 2013, 19, 17751 – 17765
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17763

P. D. Beer et al.
3
adjacent to trz+), 7.26–7.32 (m, 14H, Ar-H), 7.21 (d, J=8.8 Hz, 2H, Ar-
[14] H. Juwarker, J. M. Lenhardt, D. M. Pham, S. L. Craig, Angew.
Chem. 2008, 120, 3800–3803; Angew. Chem. Int. Ed. 2008, 47, 3740–
3743.
[15] H. Juwarker, J. M. Lenhardt, J. C. Castillo, E. Zhao, S. Krishnamur-
thy, R. M. Jamiolkowski, K. H. Kim, S. L. Craig, J. Org. Chem. 2009,
74, 8924–8934.
[16] Y. Hua, A. H. Flood, Chem. Soc. Rev. 2010, 39, 1262–1271.
[17] Y. Hua, A. H. Flood, J. Am. Chem. Soc. 2010, 132, 12838–12840.
[18] Y. Hua, R. O. Ramabhadran, E. O. Uduchi, J. A. Karty, K. Raghava-
chari, A. H. Flood, Chem. Eur. J. 2011, 17, 312–321.
[19] Y. Li, A. H. Flood, J. Am. Chem. Soc. 2008, 130, 12111–12222.
[20] Y. Li, A. H. Flood, Angew. Chem. 2008, 120, 2689–2692; Angew.
Chem. Int. Ed. 2008, 47, 2649–2652.
[21] R. O. Ramabhadran, Y. Hua, Y. Li, A. H. Flood, K. Raghavachari,
Chem. Eur. J. 2011, 17, 9123–9129.
[22] Y. Wang, F. Bie, H. Jiang, Org. Lett. 2010, 12, 3630–3633.
[23] Q.-Y. Cao, T. Pradhan, S. Kim, J. S. Kim, Org. Lett. 2011, 13, 4386–
4389.
[24] M. R. Krause, R. Goddard, S. Kubik, J. Org. Chem. 2011, 76, 7084–
7095.
[25] V. Haridas, S. Sahu, P. Venugopalan, Tetrahedron 2011, 67, 727–733.
[26] H. Zheng, W. Zhou, J. Lv, X. Yin, Y. Li, H. Liu, Y. Li, Chem. Eur. J.
2009, 15, 13253–13262.
H), 7.07–7.14 (m, 12H, Ar-H), 6.80 (d, 3J=8.6 Hz, amide-CH2-Ar-H-O),
5.92 (d, 3J=8.6 Hz, amide-CH2-Ar-H-O), 5.70 (brs, 2H, CH=CH), 4.42
(t, J=5.8 Hz, 4H, amide-CH2-Ph), 4.12 (s, 3H, trz+-CH3), 3.86 (brs, 4H,
3
O-CH2-CH=CH), 3.64 (brs, 4H, Ph-O-CH2-CH2-O), 3.53 (brs, 4H, Ph-
O-CH2-CH2-O), 1.30 (s, 27H, tBu-H), 1.29 ppm (s, 27H, tBu-H); HRESI-
MS (pos.): 1617.9253, calcd for [C107H121N6O8]+: 1617.9240.
Rotaxane 22·PF6: Rotaxane 22·Cl (0.013 g, 0.0080 mmol) was taken up in
dichloromethane (20 mL), washed with NH4PF6(aq.) (7”20 mL) and water
(3”20 mL). It was dried thoroughly in vacuo to give 22·PF6 as a white
powder. Yield: 0.011 g (79%).
1H NMR (1:1 CDCl3/CD3OD): d=9.03 (s, 1H, trz+-H), 8.80 (s, 2H, ext.
Ar-H), 8.31 (s, 1H, int. Ar-H), 7.40 (d, 3J=8.4 Hz, 2H, Ar-H adjacent to
trz+), 7.28–7.32 (m, 14H, Ar-H), 7.13–7.16 (m, 16H, Ar-H), 6.85 (d, 3J=
8.5 Hz, 4H, amide-CH2-Ar-H-O), 6.02 (d, 3J=8.5 Hz, 4H, amide-CH2-
Ar-H-O), 5.45 (brs, 2H, CH=CH), 4.28–4.42 (m, 4H, amide-CH2-Ph),
3.88 (s, 3H, 3 ” H f), 3.50–3.72 (m, 12H, Ph-O-CH2-CH2-O-CH2-CH=),
1.29 ppm (s, 54H, tBu-H); 13C NMR (1:1 CDCl3/CD3OD): d=165.5,
157.9, 153.3, 153.0, 149.7, 149.6, 149.4, 143.9, 143.8, 143.6, 137.1, 133.1,
132.6, 132.5, 132.1, 131.0, 130.9, 130.3, 130.1, 128.7, 127.7, 125.7, 125.4,
125.3, 125.0, 121.0, 119.4, 114.0, 71.3, 69.8, 67.7, 64.8, 64.7, 43.7, 39.2, 34.9,
34.9, 31.7, 31.6 ppm; 19F NMR (1:1 CDCl3/CD3OD): d=À73.7 ppm (d,
J
J
P, F =711 Hz); 31P NMR (1:1 CDCl3/CD3OD): d=À144.5 ppm (septet,
P, F =711 Hz); LRESI-MS (pos.): 1617.88, calcd for [C107H121N6O8]+:
[27] Y. Li, M. Pink, J. A. Karty, A. H. Flood, J. Am. Chem. Soc. 2008,
130, 17293–17295.
1617.92.
[28] S. Lee, Y. Hua, A. H. Flood, Org. Lett. 2010, 12, 2100–2102.
[29] A. Kumar, P. S. Pandey, Org. Lett. 2008, 10, 165–168.
[30] K. Ohmatsu, M. Kiyokawa, T. Ooi, J. Am. Chem. Soc. 2011, 133,
1307–1309.
[31] R. K. Chhatra, A. Kumar, P. S. Pandey, J. Org. Chem. 2011, 76,
9086–9089.
Acknowledgements
[32] 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.
[33] L. C. Gilday, N. G. White, P. D. Beer, Dalton Trans. 2012, 41, 7092–
7097.
[34] N. G. White, P. D. Beer, Beilstein J. Org. Chem. 2012, 8, 246–252.
[35] G. T. Spence, M. B. Pitak, P. D. Beer, Chem. Eur. J. 2012, 18, 7100–
7108.
[36] B. Schulze, C. Friebe, M. D. Hager, W. Günther, U. Kçhn, B. O.
Jahn, H. Gçrls, U. Schubert, Org. Lett. 2010, 12, 2710–2713.
[37] N. G. White, S. Carvalho, V. FØlix, P. D. Beer, Org. Biomol. Chem.
2012, 10, 6951–6959.
[38] J. Cia, B. P. Hay, N. J. Young, X. Yang, J. L. Sessler, Chem. Sci. 2013,
4, 1560–1567.
We thank Oxford Chemical Crystallography for instrumentation, Dia-
mond Lightsource for an award of beamtime on Beamline I19, Dr. David
Allan (Diamond Lightsource) for assistance with the structure of 17·BF4,
Dr. Amber Thompson (University of Oxford) for helpful discussions, and
Dr. Nick Rees (University of Oxford) for assistance with ROESY NMR
spectroscopy. N.G.W. thanks the Clarendon Fund, Trinity College Oxford
and the University of Oxfordꢀs Vice-Chancellorsꢀ Fund for financial sup-
port. S.C. thanks FCT for the postdoctoral grant SFRH/BPD/42357/2007.
P.J.C. acknowledges project New Strategies Applied to Neuropathological
Disorders (CENTRO-07-ST24-FEDER-002034), co-financed by QREN,
Mais Centro- Programa Operacional Regional do Centro e Uni¼o Eu-
ropeia/Fundo Europeu de Desenvolvimento Regional.
[39] M. R. Sambrook, P. D. Beer, J. A. Wisner, R. L. Paul, A. R. Cowley,
J. Am. Chem. Soc. 2004, 126, 15364–15365.
[40] M. D. Lankshear, P. D. Beer, Acc. Chem. Res. 2007, 40, 657–668.
[41] J. A. Wisner, P. D. Beer, M. G. B. Drew, M. R. Sambrook, J. Am.
Chem. Soc. 2002, 124, 12469–12476.
[1] P. D. Beer, P. A. Gale, Angew. Chem. 2001, 113, 502–532; Angew.
Chem. Int. Ed. 2001, 40, 486–516.
[2] J. L. Sessler, P. A. Gale, W.-S. Cho, Anion Receptor Chemistry, RSC,
2006.
[42] L. M. Hancock, L. C. Gilday, S. Carvalho, P. J. Costa, V. FØlix, C. J.
Serpell, N. L. Kilah, P. D. Beer, Chem. Eur. J. 2010, 16, 13082.
[43] G. T. Spence, P. D. Beer, Acc. Chem. Res. 2013, 46, 571–586.
[44] D. J. Cram, Angew. Chem. 1988, 100, 1041–1052; Angew. Chem. Int.
Ed. Engl. 1988, 27, 1009–1112.
[3] C.-H. Lee, S. Lee, H. Yoon, W.-D. Jang, Chem. Eur. J. 2011, 17,
13898–13903.
[4] Y. Wang, J. Xiang, H. Jiang, Chem. Eur. J. 2011, 17, 613–619.
[5] V. Haridas, S. Sahu, P. P. P. Kumar, Tetrahedron Lett. 2011, 52, 6930–
6934.
[45] A. Metzger, V. M. Lynch, E. V. Anslyn, Angew. Chem. 1997, 109,
911–914; Angew. Chem. Int. Ed. Engl. 1997, 36, 862–865.
[46] P. V. Santacroce, J. T. Davis, M. E. Light, P. A. Gale, J. C. Iglesias-
Sµnchez, P. Prados, R. Quesada, J. Am. Chem. Soc. 2007, 129, 1886–
1887.
[47] M. P. Hughes, B. D. Smith, J. Org. Chem. 1997, 62, 4492–4499.
[48] K. H. Choi, A. D. Hamilton, J. Am. Chem. Soc. 2001, 123, 2456–
2457.
[6] F. Garcꢁa, M. R. Torres, E. Matesanz, L. Sµnchez, Chem. Commun.
2011, 47, 5016–5018.
[7] Y.-J. Li, L. Xu, W.-L. Yang, H.-B. Liu, S.-W. Lai, C.-M. Che, Y.-L.
Li, Chem. Eur. J. 2012, 18, 4782–4790.
[8] N. G. White, P. D. Beer, Chem. Commun. 2012, 48, 8499–8501.
[9] S. C. Picot, B. R. Mullaney, P. D. Beer, Chem. Eur. J. 2012, 18, 6230.
[10] G. Yu, Z. Zhang, C. Han, M. Xue, Q. Zhou, F. Huang, Chem.
Commun. 2012, 48, 2958–2960.
[49] M. J. Chmielewski, J. Jurczak, Chem. Eur. J. 2005, 11, 6080–6094.
[50] H. W. Gibson, S. H. Lee, P. T. Engen, P. Le Cavalier, J. Sze, Y. X.
Shen, M. Bheda, J. Org. Chem. 1993, 58, 3748–3756.
[51] P. R. Ashton, D. Philp, N. Spencer, J. F. Stoddart, J. Chem. Soc.
Chem. Commun. 1992, 1124–1128.
[11] J. L. Sessler, J. J. Cai, H.-Y. Gong, X. P. Yang, J. F. Arambula, B. P.
Hay, J. Am. Chem. Soc. 2010, 132, 14058–14060.
[12] M. G. Fisher, P. A. Gale, J. R. Hiscock, M. B. Hunthouse, M. E.
Light, F. P. Schmidtchen, C. C. Tong, Chem. Commun. 2009, 3017–
3019.
[13] R. M. Meudtner, S. Hecht, Angew. Chem. 2008, 120, 5004–5008;
Angew. Chem. Int. Ed. 2008, 47, 4926–4930.
[52] K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975, 16,
4467.
17764
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17751 – 17765

Halide Recognition by Triazolium Containing Rotaxanes
FULL PAPER
[53] V. V. Rostovtsev, L. G. Green, V. V. Fokin, K. B. Sharpless, Angew.
Chem. 2002, 114, 2708–2711; Angew. Chem. Int. Ed. 2002, 41, 2596–
2599.
[54] C. W. Tornoøe, C. Christensen, M. Meldal, J. Org. Chem. 2002, 67,
3057–3064.
[64] A. W. Goetz, M. J. Williamson, D. Xu, D. Poole, S. L. Grand, R. C.
Walker, J. Chem. Theory Comput., 2012, 8, 1542–1555.
[65] S. Le Grand, A. W. Goetz, R. C. Walker, Comp. Phys. Comm. 2013,
184, 374–380.
[66] J. Wang, R. M. Wolf, J. W. Caldwell, P. A. Kollman, D. A. Case, J.
Comput. Chem. 2004, 25, 1157–1174.
[55] V. Aucagne, K. D. Hanni, D. A. Leigh, P. J. Lusby, D. B. Walker, J.
Am. Chem. Soc. 2006, 128, 2186–2187.
[67] I. S. Young, T. E. Cheatham III, J. Phys. Chem. B 2008, 112, 9020–
9041.
[56] M. J. Hynes, J. Chem. Soc. Dalton Trans. 1993, 311–312.
[57] This peak was monitored as it moved the most during the course of
the titration and hence gave data with the lowest errors.
[58] N. G. White, P. D. Beer, Supramol. Chem. 2012, 24, 473–480.
[59] This solvent was used as 1:1 CDCl3/CD3OD was too competitive to
obtain useful binding data.
[60] H. Hagelin, J. S. Murray, P. Politzer, T. Brinck, M. Berthelot, Can. J.
Chem. 1995, 73, 483–488.
[61] J. S. Murray, K. E. Riley, P. Politzer, T. Clark, Aust. J. Chem. 2010,
63, 1598–1607.
[68] B. Roux, Comp. Phys. Comm. 1995, 91, 275–282.
[69] M. R. Sambrook, P. D. Beer, M. D. Lankshear, R. F. Ludlow, J. A.
Wisner, Org. Biomol. Chem. 2006, 4, 1529–1538.
[70] B.-Y. Lee, S. R. Park, H. B. Jeon, K. S. Kim, Tetrahedron Lett. 2006,
47, 5105–5109.
[71] H. Nowell, S. A. Barnett, K. E. Christensen, S. J. Teat, D. R. Allan,
J. Synchrotron Radiat. 2012, 19, 435–441.
[72] CrysAlis Pro, Agilent Technologies, 2011.
[73] L. Palatinus, G. Chapuis, J. Appl. Crystallogr. 2007, 40, 786–790.
[74] P. W. Betteridge, J. R. Carruthers, R. I. Cooper, K. Prout, D. J.
Watkin, J. Appl. Crystallogr. 2003, 36, 1487.
[62] Y. Ma, K. Gross, P. Hollingsworth, P. Seybold, J. Murray, J. Mol.
Model. 2004, 10, 235–239.
[63] D. A. Case et al., 2012, AMBER12, University of California, San
Francisco (full reference is given in the Supporting Information).
Received: August 7, 2013
Published online: November 22, 2013
Chem. Eur. J. 2013, 19, 17751 – 17765
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17765