Oligoether-strapped calix[4]pyrrole: An ion-pair receptor displaying cation-dependent chloride anion transport

Article abstract of DOI:10.1002/chem.201103239

A ditopic ion-pair receptor (1), which has tunable cation- and anion-binding sites, has been synthesized and characterized. Spectroscopic analyses provide support for the conclusion that receptor 1 binds fluoride and chloride anions strongly and forms sta

Full text of DOI:10.1002/chem.201103239

DOI: 10.1002/chem.201103239
Oligoether-Strapped Calix[4]pyrrole: An Ion-Pair Receptor Displaying
Cation-Dependent Chloride Anion Transport
In-Won Park,[a] Jaeduk Yoo,[a] Bohyang Kim,[a] Suman Adhikari,[a] Sung Kuk Kim,[b]
Yerim Yeon,[b] Cally J. E. Haynes,[c] Jennifer L. Sutton,[c] Christine C. Tong,[c]
Vincent M. Lynch,[b] Jonathan L. Sessler,*[b] Philip A. Gale,*[c] and Chang-Hee Lee*[a]
2514
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2514 – 2523

FULL PAPER
Abstract: A ditopic ion-pair receptor
(1), which has tunable cation- and
anion-binding sites, has been synthe-
sized and characterized. Spectroscopic
analyses provide support for the con-
clusion that receptor 1 binds fluoride
and chloride anions strongly and forms
stable 1:1 complexes ([1·F]ꢀ and
[1·Cl]ꢀ) with appropriately chosen salts
of these anions in acetonitrile. When
the anion complexes of 1 were treated
with alkali metal ions (Li+, Na+, K+,
Cs+, as their perchlorate salts), ion-de-
pendent interactions were observed
that were found to depend on both the
choice of added cation and the initially
complexed anion. In the case of [1·F]ꢀ,
no appreciable interaction with the K+
ion was seen. On the other hand, when
this complex was treated with Li+ or
Na+ ions, decomplexation of the
bound fluoride anion was observed. In
contrast to what was seen with Li+,
Na+, K+, treating [1·F]ꢀ with Cs+ ions
gave rise to a stable, host-separated
ion-pair complex, [F·1·Cs], which con-
tains the Cs+ ion bound in the cup-like
portion of the calix[4]pyrrole. Different
complexation behavior was seen in the
case of the chloride complex, [1·Cl]ꢀ.
Here, no appreciable interaction was
observed with Na+ or K+. In contrast,
treating with Li+ produces a tight ion-
pair complex, [1·Li·Cl], in which the
cation is bound to the crown moiety. In
analogy to what was seen for [1·F]ꢀ,
treatment of [1·Cl]ꢀ with Cs+ ions
gives rise to a host-separated ion-pair
complex, [Cl·1·Cs], in which the cation
is bound to the cup of the calix[4]pyr-
role. As inferred from liposomal model
membrane transport studies, system 1
can act as an effective carrier for sever-
al chloride anion salts of Group 1 cat-
ions, operating through both symport
(chloride+cation co-transport) and an-
tiport (nitrate-for-chloride exchange)
mechanisms. This transport behavior
stands in contrast to what is seen for
simple
octamethylcalix[4]pyrrole,
Keywords: binding sites · calixar-
enes · crown compounds · ion pairs ·
ion transport · structure-activity re-
lationships
which acts as an effective carrier for
cesium chloride but does not operates
through a nitrate-for-chloride anion ex-
change mechanism.
Introduction
topic receptors that contain a cation complexation site com-
bined with an amide-derived anion-binding site.[2] The use of
thiourea groups[3] and macrocyclic phosphine oxide sites[4]
for ion-pair recognition have been reported recently. Vari-
ous other molecular frameworks have been developed for
ion-pair recognition; however, it is rare that recognition
takes place through formation of a contact ion pair. One of
the essential, and still unanswered, questions in the area of
ion-pair recognition is why different systems bind different
combinations of anions and cations in different ways. Re-
cently, we detailed three limiting binding modes for ion-pair
recognition,[5] namely, contact ion pair, solvent-separated
ion pair and dissociated ion pair, and have begun working
to understand how the combination of anion and cation,
combined with differences in receptor structure, favor one
mode or another.[5,6]
The simultaneous recognition and binding of cationic and
anionic guests by a single receptor, either as ion pairs or co-
bound salts, is a recognized challenge in supramolecular
chemistry. However, properly designed ion-pair receptors
are attractive for use in selective extraction/transportation
applications and for mimicking important biological func-
tions, such as ion transport across cellular membranes. Al-
though considerable effort has been devoted to the problem
of designing ion-pair receptors, the number of available sys-
tems remains limited and the determinants of concurrent
anion and cation recognition remain recondite.[1] Many of
the reported studies have concerned spatially separated di-
[a] I.-W. Park, J. Yoo, B. Kim, Dr. S. Adhikari, Prof. C.-H. Lee
Department of Chemistry
An attractive approach to the design and synthesis of ion-
pair receptors involves incorporating disparate binding sites
into a properly preorganized scaffold in such a way that that
the targeted anion and cation are held in close proximity
without allowing significant interaction between the individ-
ual recognition subunits. Ion-pair receptors designed accord-
ing to these principles are generally expected to display su-
perior binding properties (e.g., higher affinities and greater
selectivity) than linked systems that simply serve to bind
pairs of dissociated ions, rather than associated ion pairs.
Currently, the definition of associated ion pair is operational
in the context of receptor design and can refer to constructs
in which the anion and cation are 1) in close contact, 2) sol-
vent separated, or 3) bound in a convergent fashion as the
result of receptor-mediated interactions.[7]
Kangwon National University
Chun Cheon, 200-701 (Korea)
Fax : (+82)33-253-7582
E-mail: chhlee@kangwon.ac.kr
[b] Dr. S. K. Kim, Y. Yeon, Dr. V. M. Lynch, Prof. J. L. Sessler
Department of Chemistry and Biochemistry
1 University Station-A5300
The University of Texas at Austin
Austin, Texas 78712-0165 (USA)
Fax : (+1)512-471-5009
E-mail: sessler@mail.utexas.edu
[c] Dr. C. J. E. Haynes, J. L. Sutton, Dr. C. C. Tong, Prof. P. A. Gale
School of Chemistry
University of Southampton
Southampton, SO17 1BJ (UK)
Fax : (+44)23-80596805
E-mail: philip.gale@soton.ac.uk
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103239.
Chem. Eur. J. 2012, 18, 2514 – 2523
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2515

C.-H. Lee, J. L. Sessler, P. A. Gale et al.
Results and Discussion
cation binding and a calix[4]pyrrole for anion recognition.
However, as has been demonstrated in several studies since
2005, the calix[4]pyrrole skeleton is not completely innocent
as an anion binding agent;[9] in certain circumstances large
cations can be accommodated in the pyrrole-derived cup
present in the pyrrole-bound cone conformation (Figure 1).
However, such in-cup cation complexation has not been ob-
served in the case of smaller cations, such as K+ and Na+.
Thus, points of interest are whether the combined use of a
crown ether strap and a calix[4]pyrrole core would allow for
ion-pair recognition of NaCl, KCl, and CsCl; whether the
nature of the complex would differ (e.g., crown- or cup-
based recognition of M+ (M+ =Cs+, K+, or Na+)); and
how these presumed recognition effects would translate into
differences in transport phenomenon. In this latter context,
we were particularly keen to see if the crown-strapped
calix[4]pyrrole 1 could be used to effect the concurrent
transport of K+ and Clꢀ or Na+ and Clꢀ in our standard lip-
osomal model. Such cotransport is not observed with simple
octamethylcalix[4]pyrrole, even in the presence of a good
K+ transporter, such as valinomycin.[10]
Recently, we described
a
calix[4]arene–calix[4]pyrrole
hybrid system, in which all three of these limiting binding
modes were observed within the same receptor depending
on the choice of anion and cation.[5] Herein, we report the
synthesis, characterization, and ion-pair-recognition proper-
ties of a new hybrid receptor (1; Figure 1) that contains both
Figure 1. Schematic line drawing of crown-ether-strapped calix[4]pyrrole
ion-pair receptor 1.
The synthesis of receptor 1 is summarized in Scheme 1.
Tosylated pentaethylene glycol 3 was reacted with meso-(p-
hydroxyphenyl-methyl)dipyrromethane 2 to give the corre-
sponding podand-like dipyrromethane 4 in 74% yield.
Lewis acid-catalyzed condensation of 4 with acetone (as the
solvent) afforded receptor 1 in 10% yield.[11,12]
Receptor 1 was characterized by standard spectroscopic
means. Although efforts to obtain a diffraction-grade single
crystal of free host 1 were not successful, crystals of the
CsCl complex of receptor 1 ([1·CsCl]) suitable for X-ray
analysis could be obtained by slow evaporation from a mix-
ture of chloroform and ethanol. The resulting structure re-
vealed that in [1·CsCl] one molecule of ethanol forms a hy-
drogen bond with the bound cesium cation (Figure 2). The
cesium ion is also complexed by four of the six oxygen
atoms of the oligoether moiety, as inferred from the pres-
ence of Cs+···O contacts, the lengths of which vary from
3.052 to 3.438 ꢁ. However, the bound chloride anion is hy-
drogen bonded to the four pyrrole NH protons as well as to
a bound ethanol molecule. The separation between the chlo-
an anion-binding site and a cation-binding site held in close
proximity. This system, the first to our knowledge that is
based on the direct strapping of a calix[4]pyrrole core with
an oligoether, was found to form ion-pair complexes with
halide anion salts, the nature and stability of which was
found to depend on the choice of Group 1 counterion. As
judged from through-liposomal and bulk U-tube type model
membrane studies, this same system was found to be an ef-
fective carrier for chloride anions, the efficiency of which
was again found to depend on the choice of counterion. Of
particular note is that chloride anion transport was seen
when NaCl, KCl, or CsCl was used as the chloride anion
source, whereas chloride transport was only observed in the
case of CsCl with simple octamethylcalix[4]pyrrole.[8] Based
on the transport studies, it is inferred that system 1 can func-
tion both as a Group 1 cation+chloride anion symport car-
rier and a chloride-for-nitrate exchanger (antiport carrier).
The design of receptor 1 is based on two well-established
recognition motifs, namely, a crown ether (oligoether) for
Scheme 1. Synthesis of receptor 1.
2516
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2514 – 2523

Oligoether-Strapped Calix[4]pyrrole
FULL PAPER
studies of anion binding were made in acetonitrile by using
isothermal titration calorimetry (ITC), a technique that has
been applied extensively to the study of calix[4]pyrrole-
based anion binding in recent years.[13] Table 1 summarizes
Table 1. Association constants measured by ITC as determined from the
titration of receptor 1 with various anions (all studies carried out by
using the corresponding tetrabutylammonium salts) in CH3CN. In these
studies, [1]=0.25 mm. Estimated errors are ꢁ10%.
K1 [mꢀ1
]
K2 [mꢀ1
1.49ꢂ105
]
Fꢀ[a]
4.88ꢂ108
9.70ꢂ104
2.09ꢂ103
2.56ꢂ105
3.89ꢂ103
1.29ꢂ102
2.86ꢂ102
Clꢀ
Brꢀ
AcOꢀ
H2PO4
ꢀ
ꢀ
HSO4
ꢀ
NO3
Figure 2. A) Single-crystal X-ray structure of the cesium chloride com-
plex of receptor 1. Hydrogen atoms have been removed for clarity and
displacement ellipsoids are scaled to the 50% probability level. Note that
two different complexation modes are seen for the two cesium cations.
One of these involves complexation by the oligoether, whereas the other
involves binding of the cation within the tetrapyrrolic “cup” present in
the cone conformation of calix[4]pyrrole. B) The resulting extended
framework (dark gray spheres: Cs+, light gray spheres: Clꢀ).
[a] Treated mathematically as two separate and successive binding events.
the resulting association constants Ka, measured by ITC (all
anions were studied as the corresponding TBA salts). On
the basis of these studies, it is concluded that the fluoride
anion interacts exceptionally well with receptor 1 (Ka =
4.88ꢂ108 mꢀ1). Although further study will be required to
elucidate fully the determinants underlying this exceptional-
ly high affinity, we ascribe it to a combination of hydrogen-
bond interactions (as found in many fluoride anion recep-
tors) and stabilizing anion–p interactions that are possible in
the present instance as the result of the two phenyl groups
present within the receptor. Evidence for a second binding
event, characterized by K2 =1.49ꢂ105 mꢀ1, was also obtained
when excess fluoride anion was added. These results provide
support for the notion that receptor 1 provides relatively
preorganized binding sites that allow for both cooperative
hydrogen-bonding interactions with one or two fluoride
anions (as a function of relative concentration) and anion–p
interactions. In the event, it is important to underscore that
receptor 1 binds the fluoride anion remarkably well, at least
in acetonitrile.
ride anion and the pyrrole nitrogen atoms is nearly constant
ꢀ
ꢀ
ꢀ
(3.23 ꢁ<Cl ···N<3.29 ꢁ). The Cl ···O Et separation was
3.143 ꢁ, whereas the Cs+···Clꢀ separation was 5.407 ꢁ. The
relatively short contact between the Clꢀ ion and the bound
ethanol is consistent with the notion that for this particular
salt, complexation of the anion and cation as a solvent-sepa-
rated ion pair is energetically more favorable than binding
the corresponding putative contact ion pair.[6]
In the structure of complex [1·CsCl], a cup-bound cesium
ion is also seen. It interacts with a chloride anion bound in a
neighboring calix[4]pyrrole. Because the crown-bound
cesium ion also interacts with a neighboring crown ether
moiety, an extended superstructure is produced in the solid
state (see below).
Preliminary solution-phase anion-binding studies of recep-
tor 1 were carried out in CD3CN by using H NMR spec-
The effect of the counterion was explored by treating pre-
formed fluoride complex [1·F]ꢀ (TBA counterion) with vari-
ous alkaline metal ion salts. A mixture of acetonitrile/meth-
anol (9:1 v/v), which was chosen due to solubility considera-
tions, was used for these studies. As shown in Figure 3, the
spectrum of the fluoride-bound complex [1·F]ꢀ shows typical
signal changes associated with slow complexation/decom-
1
troscopy. To check the interactions between Group 1 cations
and the oligoether moiety, receptor 1 was subjected to titra-
tions with various alkaline metal perchlorate salts (Li+,
Na+, K+, and Cs+). However, no appreciable changes in
chemical shifts were observed (see the Supporting Informa-
tion). This finding provides support for the conclusion that
the
ꢀ
plexation. Upfield shifts were seen for the Ar H signal, a
finding considered consistent with the presence of possible
anion–p interactions.
oligoether moiety does not complex these metal cations in
ꢀ
acetonitrile. In contrast to what was seen for PF6 cation
Upon addition of Cs+ (as the perchlorate salt), a slight
downfield shift for the pyrrole-b-H protons was observed
(Figure 3). Conversely, the pyrrole NH protons were ob-
served to undergo a modest shift to higher field. Little dis-
cernable change in the signals ascribable to the oligoether
moiety was observed. These results are consistent with the
Cs+ being bound to the cup of the calix[4]pyrrole in its cone
conformation, thus forming a receptor-mediated ion-pair
salts, significant chemical shift changes were observed when
receptor 1 was titrated with fluoride ions (as the tetrabuty-
lammonium (TBA) salt). Titrations involving other anions
also revealed changes in the calix[4]pyrrole chemical shifts
characteristic of anion binding, with the extent of the
changes at a given number of anion equivalents being found
to depend on the specific anion employed. Quantitative
Chem. Eur. J. 2012, 18, 2514 – 2523
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2517

C.-H. Lee, J. L. Sessler, P. A. Gale et al.
observed. Rather, a cavity-bound ion-pair complex contain-
ing both the Li+ and Clꢀ ions formed. Evidence for this
1
complexation mode comes from the H NMR spectra shown
in Figure 4. The chemical shift of the pyrrole NH proton
signal undergoes little appreciable change upon addition of
Li+ ions. Conversely, the chemical shift of the ether bridges
was gradually shifted to lower field upon addition of Li+
ion. These observations are most readily rationalized in
terms of Li+ ions binding to the oligoether moiety and the
chloride remaining co-bound by the pyrrole NH protons;
the net result is the formation of a cavity-bound ion-pair
complex.
Preformed complex [1·Cl]ꢀ was also treated with different
alkaline metal cations (as the perchlorate salts) under analo-
gous solution-phase conditions. Again, the resulting changes
in chemical shift allowed insights into the complexation be-
havior to be inferred. As can be seen from an inspection of
Figure 5, the addition of one equivalent of Cs+ to [1·Cl]ꢀ re-
sulted in a significant downfield shift in the b-pyrrolic
proton signals but little change in the NH proton signals.
Figure 3. Partial 1H NMR spectra of complex [1·F]ꢀ (TBA salt) observed
upon treatment with various alkaline perchlorate salts. a) Free 1,
b) [1·F]ꢀ (1+TBAF (1.5 equiv)), c) [1·F]ꢀ +CsClO4 (5 equiv), d) [1·F]ꢀ +
KClO4 (5 equiv), e) [1·F]ꢀ +NaClO4 (5 equiv), f) [1·F]ꢀ +LiClO4
(5 equiv) in CD3CN/CD3OD (9:1 v/v).
complex. In the case of K+, a
significant downfield shift in
the signals ascribed to the oli-
goether moiety was observed.
Concomitantly, the resonance
of the pyrrole NH protons was
found to shift to higher field.
These observations provide
spectroscopic support for the
conclusion that the ion pair
ACHTUNGTRENNUNG
[K+·Fꢀ] is bound within the re-
ceptor cavity.
The upfield shift seen for the
pyrrole NH signal is consistent
with the intuitively appealing
conclusion that the hydrogen
bonding between the pyrrole
NH protons and the Fꢀ ion
become weaker as the interac-
Scheme 2. Effect of treating complex [1·F]ꢀ (TBA salt) with perchlorate salts of various alkaline metal ions.
tion between the co-bound ions, K+ and Fꢀ, becomes stron-
ger. On this basis, it is suggested that preformed complex
[1·F]ꢀ forms two different kinds of ion-pair complex depend-
ing on whether the counterion is Cs+ or K+. In contrast, the
addition of smaller cations, specifically Li+ or Na+, serves
to break preformed complex [1·F]ꢀ, as evidenced by a resto-
ration of signals in the NMR spectrum originally seen for
free 1. Presumably, these cations mediate this effect by bind-
ing so tightly with the fluoride anion under these solvent
ꢀ
conditions that they outcompete and thus rupture the N
H···Fꢀ hydrogen-bonding interactions present in [1·F]ꢀ.
These varying cation-dependent processes are summarized
in Scheme 2.
In contrast with the above, when preformed chloride com-
plex [1·Cl]ꢀ (TBA salt) was treated with LiClO4, no evi-
dence of decomplexation of the bound chloride anion was
Figure 4. 1H NMR spectral changes observed when complex [1·Cl]ꢀ (top
trace) is treated with a) 1, b) 3, c) 5, and d) 10 equivalents of LiClO4 in
CD3CN.
2518
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2514 – 2523

Oligoether-Strapped Calix[4]pyrrole
FULL PAPER
was mixed with equimolar
quantities of various test anions
(acetate, phosphate, chloride,
and fluoride, as their corre-
sponding tetrabutylammonium
salts)
in
[D3]acetonitrile/
[D4]MeOH (9:1), the fluoride
anion complex was formed in
preference and to the exclusion
of the other possible complexes.
Support for the suggestion that
the fluoride complex is formed
exclusively (within the limits of
1H NMR spectroscopic analy-
sis) comes from noting that
under these conditions the pyr-
role NH proton signal shifts to
lower field and displays the
19F–1H splitting typical of
a
Figure 5. 1H NMR spectra of a) free 1, b) [1·Cl]ꢀ (TBA salt), c) [1·Cl]ꢀ +Cs+, d) [1·Cl]ꢀ +K+, c) [1·Cl]ꢀ +Na+,
and d) [1·Cl]ꢀ +Li+ ([M+]=5 equiv, perchlorate salts) in CD3CN.
calix[4]pyrrole fluoride anion
complex undergoing slow ex-
change.
This cation-induced behavior is ascribed to cation
ACHTUNGTRENNUNG
(Cs+)–p(pyrrole ring) interactions. To an extent this assign-
Successive treatment of the preformed [1·F]ꢀ complex
with three equivalents of LiClO4 resulted in complete de-
complexation of the receptor-bound fluoride anion, as infer-
ment is correct, the observed chemical shifts are consistent
with the cesium cation being bound to the cup of the
calix[4]pyrrole moiety, rather than by the oligoether moiety
as in the case of Li+ (see above). Finally, addition of K+ or
Na+ resulted in no appreciable change in the chemical
shifts, although peak broadening was observed. This finding
is interpreted in terms of partial decomplexation of the
bound chloride anion occurring with these two cations.
1
red from H NMR spectral changes. These changes lead us
to propose that under these solvent conditions (chosen to
accommodate a wide range of anion salts) a combination of
Coulombic interactions between Li+ and Fꢀ and solubiliza-
tion by the relatively polar medium dominate over hydro-
gen-bonding interactions involving the pyrrolic NH protons
and the Fꢀ anion. As a result, decomplexation of the fluo-
ride anion occurs. However, when the chloride complex
[1·Cl]ꢀ is treated with LiClO4 under identical conditions, the
cation (Li+) binds to the oligoether moiety of the host mole-
cule. As a result an ion pair complex is stabilized and de-
complexation of the bound chloride anion does not occur.
The cavity-bound ion-pair complex (Li+Clꢀ) is presumably
1
Taken in concert, the cation-dependent H NMR spectral
changes are consistent with the nature of the prebound
anion complex, [1·Cl]ꢀ or [1·F]ꢀ (both as the corresponding
TBA salt), having a significant role in determining the
chemical events that transpire upon subsequent treatment
with an alkali cation (perchlorate salt) and the nature of the
ion-pair complex (if any) that is
ultimately formed. The chemis-
try occurring in the case of
[1·F]ꢀ
is
summarized
in
Scheme 2, whereas that pro-
posed for [1·Cl]ꢀ is shown in
Scheme 3. In the case of the
latter anion complex, it is infer-
red that only the upper, crown-
like cavity of the chloride-
bound receptor [1·Cl]ꢀ can ac-
commodate the Li+ ion effec-
tively, whereas the lower
calix[4]pyrrole cup allows for
effective Cs+ complexation.
Neither Na+ nor K+ bind well
to either cation binding site and
thus partial anion decomplexa-
Scheme 3. Proposed cation-induced changes that occur when complex [1·Cl]ꢀ is treated with perchlorate salts
tion occurs. When receptor 1 of various alkaline metal ions in CD3CN.
Chem. Eur. J. 2012, 18, 2514 – 2523
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2519

C.-H. Lee, J. L. Sessler, P. A. Gale et al.
stabilized in part by the ditopic nature of the receptor,
which allows for energetically favorable receptor-ion con-
tacts with the co-bound ions, as well as favorable Li+–Clꢀ
interactions.
When a mixture of acetate, chloride, and phosphate
anions (containing no fluoride anions) was treated with re-
ceptor 1 under the above mixed-solvent conditions, the
greatest spectral changes were seen in the case of the ace-
tate anion. On this basis, we conclude that the acetate anion
is bound in preference over the chloride and phosphate
anions. When this mixture, containing the pre-formed ace-
tate complex [1·OAc]ꢀ as the dominant species, is treated
with LiClO4, decomplexation of the bound acetate anion
occurs and is replaced by the chloride anion, which is bound
in the form of the Li+–Clꢀ ion pair, as inferred from the
nature of the 1H NMR spectral signals observed. This
cation-mediated control of anion recognition is a unique
feature of receptor 1 that could find use in various applica-
tions involving, for example, anion separations.
Several modified calix[4]pyrroles, including certain strap-
ped calixpyrroles, have previously been observed to function
as chloride transporters across lipid bilayer membranes.[14]
Accordingly, we studied the transport properties of receptor
1 by using the test alkali cation salts CsCl, RbCl, KCl, and
NaCl. The most highly hydrated member of the series, LiCl,
was excluded from this study. For these studies, unilamellar
1-palmitoyl-2-oleoylphophatidylcholine (POPC) vesicles
were loaded with the salts in question and then suspended
in an external solution containing either NaNO3 or Na2SO4
buffered to pH 7.2 with 5 mm sodium phosphate salts. A
sample of receptor 1 (4% molar carrier to lipid) was added
in the form of a DMSO solution and the resultant Clꢀ anion
efflux was monitored using a chloride selective electrode.
As discussed below, these studies revealed that receptor 1,
in contrast to simple calix[4]pyrrole,[13] is capable of trans-
porting chloride from within liposomes containing sodium,
potassium and rubidium chloride, with the fastest release
being observed in the case of the larger Group 1 metal cat-
ions (Figure 6).
Figure 6. Chloride efflux promoted by receptor 1 (4 mol% with respect
to lipid) from unilamellar POPC vesicles loaded with MCl (489 mm; M=
Na, K, Rb or Cs) buffered to pH 7.2 with sodium phosphate salts. The
vesicles were dispersed in Na2SO4 (167 mm) buffered to pH 7.2 with
sodium phosphate salts (5 mm). Each point represents the average of
three trials.
cation were being co-transported. On the other hand, when
NaNO3 is used, chloride anion transport would be expected
via a chloride-for-nitrate exchange or “antiport” mechanism
(as well as by default a cation+anion cotransport mecha-
nism). Thus, by studying carriers using both conditions qual-
itative insights into the relative importance of these two lim-
iting mechanisms may be obtained.
As can be seen from an inspection of Figures 6 and 7, re-
ceptor 1 can function as a carrier through a cation+anion
cotransport mechanism and through anion exchange. As im-
plied above, more effective transport was observed for the
more lipophilic cations (Cs+ >Rb+ >K+ >Na+). However,
Two sets of conditions were used to assess the ability of
receptor 1 to act as a chloride anion carrier. In the first the
solution on the outside of the liposome was loaded with
NaNO3, whereas in the second Na2SO4 was used. The
doubly negatively charged sulfate anion is highly hydrophilic
(DG
(SO42ꢀ)=ꢀ1080 kJmolꢀ1)[14] and was not expected to
be soluble in the liposomal membrane. In contrast, the ni-
trate anion is much more hydrophobic (DG
ꢀ
hydrA
ꢀ300 kJmolꢀ1) and is able to diffuse through the liposomal
barrier when transport is mediated by a carrier so as to
maintain charge neutrality across the membrane. Thus, as
detailed in previous reports,[15] the use of Na2SO4 in the
extra-liposomal medium was expected to allow for an assess-
ment of carrier function under conditions of cation+anion
co-transport. In the specific case of experiments involving
receptor 1, an appreciable increase in chloride anion con-
centration outside of the vesicle as a function of time would
only be expected if both the chloride anion and its counter
Figure 7. Chloride efflux promoted by 1–5 mol% of receptor 1 from uni-
lamellar POPC vesicles loaded with NaCl (489 mm) buffered to pH 7.2
with sodium phosphate salts. The vesicles were dispersed in NaNO3
(489 mm) buffered to pH 7.2 with sodium phosphate salts (5 mm). Each
point represents the average of three trials.
2520
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2514 – 2523

Oligoether-Strapped Calix[4]pyrrole
FULL PAPER
even when the anion-exchange mechanism is not operative,
a modest level of transport was seen in the case of NaCl
(Figure 6). In accord with expectations, this latter transport
rate increases when nitrate-for-chloride exchange processes
are operative (Figure 7). Here, in accord for what is predict-
ed for a carrier-based transport mechanism, the rate of
transport increased as the concentration of the receptor
within the membrane was increased. Control experiments
confirmed, in agreement with prior studies involving elabo-
rated calixpyrrole transport agents, that a test crown ether,
alone or in conjunction with calix[4]pyrrole, did not act as
an effective carrier for the chloride anion under NaCl-for-
NaNO3 exchange conditions (see the Supporting Informa-
tion). On this basis, we conclude that receptor 1 is an effec-
tive carrier for the chloride anion and one that benefits, at
least in part, from an ability to effect the co-transport of
both an anion (chloride) and a cation (preferentially Cs+,
but also other alkali cations).
Qualitative support for the conclusion that receptor 1
functions as carrier through a cation+anion cotransport
mechanism came from bulk membrane transport studies.
These involved a U-tube setup in which a dichloromethane
layer (10 mL) containing receptor 1 (1 mm) was used to sep-
arate two aqueous phases (5 mL each). The first of these
two aqueous phases (Aq. 1) consisted of 0.5m MCl (M=Cs,
K, Na) in 20 mm phosphate buffer (pH 7), whereas the
second (Aq. 2) consisted of either 0.5m NaNO3 or Na2SO4
in 20 mm phosphate buffer (pH 7).
cup of the calix[4]pyrrole moiety. Conversely, the K+ forms
a cavity-bound ion-pair complex. The bound fluoride anion
in [1·F]ꢀ is decomplexed upon addition of NaClO4 or
LiClO4. The receptor–chloride complex [1·Cl]ꢀ, on the other
hand, displays quite different cation-binding properties. The
addition of CsClO4 results in formation of an cup-bound ion
pair. Support for this latter conclusion comes from a single
crystal X-ray diffraction analysis of the CsCl complex of 1,
which revealed such a binding mode, along with an in-cavity
Cs+–Clꢀ ion-pair binding. The Li+ ion forms a cavity-bound
ion-pair complex, whereas the addition of Na+ and K+ in-
duces partial decomplexation (all cations were added as the
corresponding perchlorate salts). This behavior stands in
marked contrast to what was seen for the earlier system.[5]
The advantages of the present receptor is that its ion-recog-
nition properties can be modulated by an appropriate
choice of both the anion and cation. These findings led to
the consideration that receptor 1 could act as a versatile ion
transporter, and liposomal model membrane studies re-
vealed that could act as a chloride+cation co-transporter,
as well as a chloride-for-nitrate ion exchanger. Support for
these key conclusions came from U-tube bulk transport ex-
periments.
Experimental Section
1
General methods and materials. H NMR spectra were recorded by using
a 300 or 400 MHz NMR spectrometer with TMS as the internal standard.
Chemical shifts are reported in parts per million (ppm). Peak multiplici-
ties are given as the following abbreviations: s, singlet; brs, broad singlet;
d, doublet; t, triplet; m, multiplet. 13C NMR spectra were proton decou-
pled and recorded by using a 100 MHz NMR spectrometer with TMS as
the internal standard. High-resolution mass spectra were obtained by
using a Voyager-DE STR MALDI-TOF mass spectrometer. The bulk
membrane transport data were obtained by using Orion Ion-Selective
Solid-State Combination Electrodes (Thermo Scientific). All other chem-
icals and solvents were purchased from commercial sources and were
used as such, unless otherwise mentioned. Column chromatography was
performed over silica gel. Compounds 2 and 3 were synthesized by using
slightly modified procedures previously reported in the literature.[11,16]
All titrations (UV/Vis, fluorescence and ITC) were performed by using
HPLC-grade CH3CN purchased from Aldrich.
The amount of chloride anion transferred to Aq. 2 was
then monitored as a function of time by recording data
points at regular intervals over a period of about 24 h (see
Figures S17–S23 in the Supporting Information). It was
found that if Aq. 1 contained either CsCl or KCl, the rela-
tive rate of chloride anion transport (roughly 50 and 40%
increase in [Clꢀ] after 1 d for Cs+ and K+, respectively) was
essentially independent of the salt (NaNO3 or Na2SO4) in
the receiving phase (see Figure S17 in the Supporting Infor-
mation). In contrast, if Aq. 1 contained NaCl then roughly
twice as much Clꢀ was transferred if the receiving phase
contained NaNO3 rather than Na2ꢀSO4 (roughly 40 and 25%
increase in [Clꢀ] after 1 d for NO3 and SO42ꢀ, respectively).
This observation is consistent with an antiport mechanism
playing a major role for the more highly hydrated cation,
Na+, than for K+ or Cs+.
Compound 4:
A mixture of 5-(p-hydroxyphenyl)-5-methyldipyrrome-
thane 3 (480 mg, 1.92 mmol), K2CO3 (1.27 g), and pentaethylene glycol
ditosylate 2 (530 mg, 0.96 mmol) was dissolved in acetonitrile and heated
at reflux for 12 h. After cooling, the solvent was removed in vacuo and
the remaining solid was dissolved in methylene chloride, then washed
with water and dilute acid. The solvent was removed in vacuo and the re-
sulting solid was purified by column chromatography over silica gel
(CH2Cl2/EtOAc 5:1) to give compound 4 as a white solid (yield 651 mg,
78%). 1H NMR (400 MHz, CDCl3, 258C, TMS): d=7.81 (brs, 4H), 6.99
(d, J=8.64 Hz, 4H), 6.79 (d, J=8.85 Hz, 4H), 6.63–6.62 (m, 4H), 6.15–
6.13 (m, 4H), 5.95–5.93 (m, 4H), 4.07 (t, J=4.84 Hz, 4H), 3.80 (t, J=
4.84 Hz, 4H), 3.69–3.67 (m, 4H), 3.65–3.63 (m, 4H), 3.62 (s, 4H),
1.99 ppm (s, 6H); 13C NMR (100 MHz, CDCl3, 258C, TMS): d=157.8,
140.0, 138.17, 128.9, 117.3, 114.5, 108.6, 106.5, 71.2, 71.0, 70.1, 67.8, 44.5,
29.4 ppm; MS (MALDI-TOF): m/z calcd for C42H50N4O6: 706.37 [M]+;
found: 709.42 [M+3H]+.
Conclusion
Oligoether-strapped ion-pair receptor 1, which has both a
strong anion-binding site and two possible cation recogni-
tion sites, has been synthesized and characterized by stan-
dard spectroscopic means, as well as by the single crystal X-
ray diffraction analyses. The 1H NMR spectroscopic analyses
reveal that in solution in acetonitrile, receptor 1 forms a
strong anion complex with fluoride and chloride anions. The
receptor–fluoride complex [1·F]ꢀ only forms an ion-pair
complex with Cs+ ions, in which the cesium ion binds to the
OligoetherACTHNUTRGNE(UNG O-6) strapped calix[4]pyrrole 1: Compound 4 (1.4 mmol) was
dissolved in acetone (150 mL) and BF3·OEt2 (7.0 or 8.4 mmol) was
added. The reaction mixture was stirred for 2 h at RT. The reaction was
quenched by adding aqueous NaOH (1 n), then extracted several times
Chem. Eur. J. 2012, 18, 2514 – 2523
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2521

C.-H. Lee, J. L. Sessler, P. A. Gale et al.
with CH2Cl2. The combined organic extracts were dried (Na2SO4) and
the solvent was removed in vacuo, The residual solid was purified by
column chromatography (silica gel; CH2Cl2/EtOAc4:1) to give 1 as a
white solid (yield 0.110 g, 10%). 1H NMR (300 MHz, CDCl3, 258C,
TMS): d=7.03 (brs, 4H), 6.96 (d, J=8.85 Hz, 4H), 6.83 (d, J=8.88 Hz,
4H), 5.96–5.94 (m, 4H), 5.88–5.86 (m, 4H), 4.16 (t, J=5.02 Hz, 4H), 3.88
(t, J=5.02 Hz, 4H), 3.75–3.73 (m, 4H), 3.70–3.69 (m, 4H), 3.68 (s, 4H),
1.94 (s, 6H), 1.52 (s, 6H), 1.39 ppm (s, 6H); 13C NMR (100 MHz, CDCl3,
258C, TMS): d=157.3, 139.1, 137.3, 136.9, 128.3, 114.1, 104.9, 104.2, 70.9,
70.8, 70.7, 69.7, 67.5, 44.2, 35.5, 30.2, 29.1, 29.0 ppm; MS (MALDI-TOF):
m/z calcd for C48H58N4O6: 786.44; found: 787.33 [M+H]+.
The disordered groups/ions were modeled in essentially the same fashion.
For example, the site occupancy factor for Cs2 was assigned the variable
x, whereas that of Cs2a was assigned to (1ꢀx). A common isotropic dis-
placement parameter was refined while refining variable x. In this way,
the site occupancy for Cs2 refined to 68(2)%. No geometric restraints
were applied to the Cs ions. Geometric restraints were applied to the
ꢀ
chloroform molecules such that the C Cl bonds and the Cl-C-Cl bond
angles were restrained to be equivalent. Similar geometric restraints
were applied to the other disordered groups. H atoms on ethanol and
chloroform were not included in the refinement model. The function
2
2
Sw(jFoj ꢀjFc j )2, in which w=1/[(s(Fo))2+
(0.0752ꢂP)2+
ACHUTGTNRENNUG ACHUTNTGREN(NUGN 3.33ꢂP)] and P=
(jFo j +2jFc j )/3, was minimized. wR(F2) refined to 0.180, with R(F)
equal to 0.0614 and a goodness of fit (S) of 1.58. Definitions used for cal-
culating R(F), wR(F2) and S are given below. The data were checked for
secondary extinction effects but no correction was necessary. Neutral
atom scattering factors and values used to calculate the linear absorption
coefficient are from the International Tables for X-ray Crystallography
(1992).[20] All X-ray structural figures were generated by using
SHELXTL/PC.[21] Tables of positional and thermal parameters, bond
lengths and angles, torsion angles, and figures are given in the Supporting
Information. CCDC-821894 (1) contains the supplementary crystallo-
graphic data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.ccdc.cam.a-
c.uk/data_request/cif.
2
2
Liposomal model membrane studies: POPC was supplied by Genzyme.
Chloride concentrations during transport experiments were determined
by using an Accumet chloride-selective electrode. Vesicles were prepared
as follows: A lipid film of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocho-
line (POPC) and cholesterol (0% or 30%) was formed from a solution
in chloroform under reduced pressure and dried under vacuum for at
least 6 h. The lipid film was rehydrated by vortexing with the metal chlo-
ride (MCl) salt solution in question (489 mm MCl, 5 mm phosphate buffer
at pH 7.2). The lipid suspension was then subjected to seven freeze–thaw
cycles and allowed to age for 30 min at RT before extruding 25 times
through a 200 nm polycarbonate membrane. The resulting unilamellar
vesicles were dialyzed against the external medium to remove unencap-
sulated MCl salts.
Unilamellar POPC vesicles containing MCl, prepared as described
above, were suspended in 489 mm NaNO3 or 162 mm Na2SO4 solution
buffered to pH 7.2 with sodium phosphate salts. The lipid concentration
per sample was 1 mm. A solution of the carrier molecule (10 mm) in
DMSO was added to start the experiment and the chloride efflux was
monitored by using a chloride-sensitive electrode. At 5 min, the vesicles
were lysed with polyoxyethylene(8) lauryl ether (50 mL, 0.232 mm in 7:1
water/DMSO v/v) and a total chloride reading was taken at 7 min.
Acknowledgements
This work was supported by an NRF grant funded by MEST (2009-
0087013). The Central Instrumentation Facility at KNU and the BK21
program are acknowledged for support. The work in Austin was support-
ed by the U.S. Department of Energy (grant DE-FG02-01ER15186 to
J.L.S.) and the Robert A. Welch Foundation (grant F-1018 to J.L.S.) The
work in Southampton was supported by the EPSRC and the NSF.
U-tube transport studies: U-tube studies were carried out according to
procedures detailed previously.[17] Briefly, a U-shaped tube was used to
separate two aqueous phases (Aq. 1 and Aq. 2; 5 mL each) from one an-
other by means of an intervening dichloromethane layer (10 mL). Aq. 1
consisted of MCl (0.5m, M=Na, K, Cs) in phosphate buffer (20 mm,
pH 7). Aq. 2 contained either NaNO3 or Na2SO4 (0.5m) in phosphate
buffer (20 mm, pH 7). The dichloromethane layer contained carrier 1 at a
concentration of 1 mm. Transport was monitored by taking 2 mL aliquots
from Aq. 2 at regular time intervals (typically every hour for several
hours and one at around 24 h), testing the chloride anion concentration
by means of a chloride-selective electrode, and returning the aliquot to
Aq. 2. The electrode, which was calibrated before and after use, provides
readout as mV, with lower numbers corresponding to greater chloride
anion concentrations.
[1] a) B. D. Smith in Ion-Pair Recognition by Ditopic Receptors, Macro-
cyclic Chemistry: Current Trends and Future Prospectives, (Eds: K.
Gloe, B. Antonioli), Kluwer, London, U.K., 2005, pp. 137–152;
b) M. T. Reetz, C. M. Niemeyer, K. Harms, Angew. Chem. 1991,
103, 1515–1517; Angew. Chem. Int. Ed. Engl. 1991, 30, 1472–1474;
c) P. D. Beer, P. K. Hopkins, J. D. McKinney, Chem. Commun. 1999,
1253–1254; d) S. Kubik, J. Am. Chem. Soc. 1999, 121, 5846–5855;
e) G. J. Kirkovits, J. A. Shriver, P. A. Gale, J. L. Sessler, J. Inclusion
Phenom. Macrocyclic Chem. 2001, 41, 69–75.
[2] A. Arduini, E. Brindani, G. Giorgi, A. Pochini, A. Secchi, J. Org.
Chem. 2002, 67, 6188–6194.
[3] G. Tumcharern, T. Tuntulani, S. J. Coles, M. B. Hursthouse, J. D. Kil-
burn, Org. Lett. 2003, 5, 4971–4974.
[4] P. B. Savage, S. K. Holmgren, S. H. Gellman, J. Am. Chem. Soc.
1994, 116, 4069–4070.
[5] S.-K. Kim, J. L. Sessler, D. E. Gross, C.-H. Lee, J.-S. Kim, V. M.
Lynch, H. D. Lætitia, P. H. Benjamin, J. Am. Chem. Soc. 2010, 132,
5827–5836.
[6] J. L. Sessler, S.-K. Kim, D. E. Gross, C.-H. Lee, J.-S. Kim, V. M.
Lynch, J. Am. Chem. Soc. 2008, 130, 13162–13166.
[7] S.-K. Kim, J. L. Sessler, Chem. Soc. Rev. 2010, 39, 3784–3809.
[8] C. C. Tong, R. Quesada, J. L. Sessler, P. A. Gale, Chem. Commun.
2008, 6321–6323.
X-ray
data
for
the
CsCl
complex
of
1
(C48H58N4O6·CsCl·CHCl3·0.5C2H5OH): Crystals grew as colorless prisms
by slow evaporation from chloroform and ethanol. The data crystal was
cut from a larger crystal and had approximate dimensions 0.30ꢂ0.20ꢂ
0.06 mm. The data were collected by using a Rigaku AFC12 diffractome-
ter with a Saturn 724+ CCD equipped with a graphite monochromator
with MoKa radiation (l=0.71073 ꢁ). A total of 869 frames of data were
collected by using w-scans with a scan range of 18 and a counting time of
30 s per frame. The data were collected at 100 K by using a Rigaku
XStream low-temperature device. Details of crystal data, data collection,
and structure refinement are listed in Table S1 in the Supporting Infor-
mation. Data reduction were performed by using Rigaku Americas Cor-
porationꢃs Crystal Clear v. 1.40.[18] The structure was solved by direct
methods using SIR97[20] and refined by full-matrix least-squares on F2
with anisotropic displacement parameters for the non-H atoms calculated
by using SHELXL-97.[19] The hydrogen atoms on carbon were calculated
in ideal positions with isotropic displacement parameters set to 1.2ꢂUeq
of the attached atom (1.5ꢂUeq for methyl hydrogen atoms). The two
molecules of chloroform were disordered. Portions of the ether linkages
were also disordered. One of the Cs ions was p-bound to the pyrrole
rings of an adjacent receptor molecule. This Cs ion, Cs2, was also disor-
dered. Finally, the methyl group of the ethanol molecule was disordered.
[9] a) R. Custelcean, L. H. Delmau, B. A. Moyer, J. L. Sessler, W.-S.
Cho, D. Gross, G. W. Bates, S. J. Brooks, M. E. Light, P. A. Gale,
Angew. Chem. 2005, 117, 2593–2598; Angew. Chem. Int. Ed. 2005,
44, 2537–2542; b) G. W. Bates, P. A. Gale, M. E. Light, CrystEng-
Comm 2006, 8, 300–302; c) G. W. Bates, P. A. Gale, M. Eight, Su-
pramol. Chem. 2008, 20, 23–28; d) D. E. Gross, F. P. Schmidtchen,
W. Antonius, P. A. Gale, V. M. Lynch, J. L. Sessler, Chem. Eur. J.
2008, 14, 7822–7827.
2522
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 2514 – 2523

Oligoether-Strapped Calix[4]pyrrole
FULL PAPER
[10] S. J. Moore, M. G. Fisher, M. Yano, C. C. Tong, P. A. Gale, Dalton
Trans. 2011, 40, 12017–12020.
[11] S. Inokuma, T. Yamamoto, J. Nishimura, Tetrahedron Lett. 1990, 31,
97–100.
[12] G. Cafeo, A. Kaledkowski, F. H. Kohnke, A. Messina, Supramol.
Chem. 2006, 18, 273–279.
[16] P. M. Marcos, S. Fꢄlixab, J. R. Ascenso, M. A. P. Seguradod, B.
Mellah, R. Abidi, V. Hubscher-Bruder, F. Arnaud-Neu, Supramol.
Chem. 2006, 18, 285–297.
[17] P. A. Gale, C. C. Tong, C. J. E. Haynes, O. Adeosun, D. E. Gross, E.
Karnas, E. M. Sedenberg, R. Quesada, J. L. Sessler, J. Am. Chem.
Soc. 2010, 132, 3240–3241.
[13] a) J. L. Sessler, D. E. Gross, W.-S. Cho, V. M. Lynch, F. P. Schmidtch-
en, G. W. Bates, M. E. Light, P. A. Gale, J. Am. Chem. Soc. 2006,
128, 12281–12288; b) D.-W. Yoon, H. Hwang, C.-H. Lee, Angew.
Chem. 2002, 114, 1835–1837; Angew. Chem. Int. Ed. 2002, 41, 1757–
1759; c) F. P. Schmidtchen, Org. Lett. 2002, 4, 431–434; d) J.-D. Yoo,
M.-S. Kim, S.-J. Hong, J. L. Sessler, C.-H. Lee, J. Org. Chem. 2009,
74, 1065–1069.
[18] CrystalClear 1.40 (2008), Rigaku Americas Corporation, The Wood-
lands, TX.
[19] International Tables for X-ray Crystallography, Vol. C, Tables 4.2.6.8
and 6.1.1.4, A. J. C. Wilson, editor, Boston: Kluwer Academic Press,
1992.
[20] R. Roznyatovskiy, V. Lynch, J. L. Sessler, Org. Lett. 2010, 12, 4424–
44270.
[14] Y. J. Marcus, J. Chem. Soc. Faraday Trans. 1987, 2995–2999.
[15] a) M. Yano, C. C. Tong, M. E. Light, F. P. Schmidtchen, P. A. Gale,
Org. Biomol. Chem. 2010, 8, 4356–4363; b) M. G. Fisher, P. A. Gale,
J. R. Hiscock, M. B. Hursthouse, M. E. Light, F. P. Schmidtchen,
C. C. Tong, Chem. Commun. 2009, 3017–3019.
[21] G. M. Sheldrick, SHELXL97: Program for the Refinement of Crystal
Structures, University of Gottingen, Germany 1994.
Received: October 13, 2011
Published online: February 1, 2012
Chem. Eur. J. 2012, 18, 2514 – 2523
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2523