Selective redox-active molecular receptors for K+ and Ag +

Article abstract of DOI:10.1021/ol303308e

Two new tetrathiafulvalene based receptors in which the favorable redox properties of the tetrathiafulvalene unit are coupled to either a benzo-crown (X = O) or a dithiabenzo-crown (X = S) ether binding site were designed and synthesized as receptors for

Full text of DOI:10.1021/ol303308e

ORGANIC
LETTERS
2013
Vol. 15, No. 7
1452–1455
Selective Redox-Active Molecular
Receptors for Kþ and Agþ
Karina R. Larsen,† Carsten Johnsen,† Ole Hammerich,‡ and Jan O. Jeppesen*,†
Department of Physics, Chemistry, and Pharmacy, University of Southern Denmark,
Campusvej 55, DK-5230, Odense M, Denmark, and Department of Chemistry,
University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen Ø, Denmark
joj@sdu.dk
Received January 18, 2013
ABSTRACT
Two new tetrathiafulvalene based receptors in which the favorable redox properties of the tetrathiafulvalene unit are coupled to either a benzo-
crown (X = O) or a dithiabenzo-crown (X = S) ether binding site were designed and synthesized as receptors for Kþ and Agþ. The receptors display
a good (Kþ, X = O) to strong (Agþ, X = S) affinity toward the cation and a high discrimination against other metal cations.
The advent of supramolecular chemistry1 has aroused
the interest of chemists of many different persuasions in the
development ofmolecularreceptorscapableofrecognizing
specific chemicalsubstrates. Redox-activereceptors2 are of
particular interest because their affinity toward different
substrates are reflected by the redox properties of the
receptor allowing supramolecular systems capable of per-
forming controlled uptake and release of substrates to be
constructed. Redox-active molecular receptors are typi-
cally designed to allow the detection of substrates by
binding-induced changes in the redox properties and are
generally being built by the covalent association of an
appropriate binding site and a redox-active unit. In the
context of redox-active receptors, tetrathiafulvalene3
(TTF) constitutes a unique building block on account of
the fact that it can exist in three stable redox states (i.e., as
neutral TTF, as the radical cation TTF•þ, and the dica-
tionic TTF2þ) and TTF and its derivatives have been used
as the redox-active unit in a number of anion4 and cation5
responsive receptors. A majority of the cation responsive
receptors involve a binding site made of a crown-ether
cavity or a coordinating acyclic ether moiety.6 Although
a huge amount of work has been devoted to create TTF
(4) (a) Park, J. S.; Bejger, C.; Larsen, K. R.; Nielsen, K. A.; Jana, A.;
Lynch, V. M.; Jeppesen, J. O.; Kim, D.; Sessler, J. L. Chem. Sci. 2012, 3,
2685–2689. (b) Nielsen, K. A. Tetrahedron Lett. 2012, 53, 5616–5618.
(5) (a) Christensen, K.; Becher, J.; Hansen, T. K. Tetrahedron Lett.
† University of Southern Denmark.
‡ University of Copenhagen.
(1) (a) Lehn, J.-M. Supramolecular Chemistry; VCH: Weinheim,
Germany, 1995. (b) Stoddart, J. F. Nat. Chem. 2009, 1, 14–15.
(2) (a) Clare, J. P.; Statnikov, A.; Lynch, V.; Andrew L. Sargent,
A. L.; Sibert, J. W. J. Org. Chem. 2009, 74, 6637–6646. (b) Yang, F.; Liu,
Z.; Hong, B.; Guo, H. J. Incl. Phenom. Macrocycl. Chem. 2012, 72,
183–188.
ꢀ
1992, 33, 3035–3038. (b) Le Derf, F.; Levillain, E.; Trippe, G.; Gorgues,
ꢀ
A.; Salle, M.; Sebastian, R. M.; Caminade, A. M.; Majoral, J. P. Angew.
ꢀ
Chem., Int. Ed. 2001, 40, 224–227. (c) Trippe, G.; Levillain, E.; Le Derf,
ꢀ
F.; Gorgues, A.; Salle, M.; Jeppesen, J. O.; Nielsen, K.; Becher, J. Org.
ꢀ
ꢁ
ꢀ
Lett. 2002, 4, 2461–2464. (d) Trippe, G.; Le Derf, F.; Frere, P.; Salle, M.
ꢀ
Tetrahedron Lett. 2008, 49, 5452–5454. (e) Canevet, D.; Salle, M.;
Zhang, G.; Zhang, D.; Zhu, D. Chem. Commun. 2009, 2245–2269.
(6) (a) Hansen, T. K.; Jørgensen, T.; Stein, P. C.; Becher, J. J. Org.
Chem. 1992, 57, 6403–6409. (b) Dieing, R.; Morisson, V.; Moore, A. J.;
Goldenberg, L. M.; Bryce, M. R.; Raoul, J.-M.; Petty, M. C.; Garin, J.;
Saviron, M.; Lednev, I. K.; Hester, R. E.; Moore, J. N. J. Chem. Soc.,
Perkin Trans. 2 1996, 1587–1593. (c) Benhaoua, C.; Mazari, M.; Mercier,
€
(3) (a) Schukat, G.; Fanghanel, E. Sulfur Rep. 1996, 18, 1–294. (b)
Bryce, M. R. J. Mater. Chem. 2000, 10, 589–598. (c) Nielsen, M. B.;
Lomholt, C.; Becher, J. Chem. Soc. Rev. 2000, 29, 153–164. (d) Segura,
J. L.; Martın, N. Angew. Chem., Int. Ed. 2001, 40, 1372–1409. (e) Inagi,
S.; Naka, K.; Chujo, Y. J. Mater. Chem. 2007, 17, 4122–4135. (f) Martın,
N.; Sanchez, L.; Herranz, M. a. A.; Illescas, B.; Guldi, D. M. Acc. Chem.
Res. 2007, 40, 1015–1024. (g) Lorcy, D.; Bellec, N.; Fourmigue, M.;
ꢀ
ꢀ
ꢀ
ꢀ
N.; Le Derf, F.; Salle, M. New. J. Chem. 2008, 32, 913–916and references
Avarvari, N. Coord. Chem. Rev. 2009, 253, 1398–1438.
cited therein.
r
10.1021/ol303308e
Published on Web 03/19/2013
2013 American Chemical Society

based receptors with high sensitivity toward cations,
selectivity has received much less attention.
Scheme 1. Synthesis of the Receptor Molecules 1 and 2
Here, we describe the synthesis and characterization of
two unique TTF based receptors in which the favorable
redox properties of the monopyrrolo-TTF7 (MPTTF) unit
are coupled to either a benzo-crown (1) or a dithiabenzo-
crown (2) ether binding site. Subsequently, binding studies
between the receptors 1 and 2 and a variety of cations are
described and it is demonstrated that the oxygen/sulfur
atoms attached to the benzo moiety in the redox-active
receptors control the selectivity toward Kþ and Agþ.
The MPTTF crown ethers 1 and 2 were synthesized as
illustrated in Scheme 1.8 Cathecol 3 and dithiole benzene 4
were alkylated with the glycol derivative 5 in MeCN in the
presence of K2CO3 and KI to afford compounds 6/7 as pale
yellow oils in high yields (83/96%). The two alcohol groups
in 6 and 7 were tosylated using TsCl in CH2Cl2 in the
presence of 4-dimethylaminopyridine (DMAP) and Et3N
affording 8 or 9 in 78 or 83% yield, respectively. Subsequent
treatment of 8 or 9 with 4,5-bis(2-cyanoethylthio)-1,3-
dithiole-2-thione9 (10) under high-dilution conditions using
CsOH as the base gave the macrocyclic crown ether 11 or 12
in approximately 25% yield. Finally, cross-couplings of 11
and 12 with 5-tosyl-(1,3)-dithiolo[4,5-c]pyrrole-2-one10 (13)
in neat (EtO)3P gave the MPTTF crown ethers 1 and 2 in 41
and 46% yields, respectively.
The MPTTF receptor molecules 1 and 2 were isolated as
analytically pure yellow solids and high-resolution electro-
spray ionization mass spectrometry (HiRes-ESI-MS) of 1
and 2 produced peaks with the exact masses of m/z =
799.0520 and 853.9961, respectively, corresponding to the
expected molecular ions M•þ (calcd M•þ = 799.0559 and
853.9999, respectively).
receptors 1 and 2 and different cations were carried out in a
1:1 mixture of CH2Cl2/MeCN (UVꢀvis spectroscopy and
cyclic voltammetry) or in a 1:1 mixture of CDCl3/CD3CN
(NMR spectroscopy).
A comparison of the 1H NMRspectraof1and 2recorded
in CDCl3 at 298 K reveals (Figures S1 and S2) significant
differences. In 1, the resonances associated with the aro-
matic H1 and H2 protons (Scheme 1) were observed as a 4 H
multiplet at δ = 6.89 ppm, whereas the H1 and H2 protons
in 2 were observed as two 2 H multiplets resonating at δ =
7.14 and 7.33 ppm, respectively, an observation that can be
accounted for by the different substitutions of the benzene
ring (i.e., oxygen vs sulfur). In compound 2, two triplets
each integrating to 4 H were observed in the aliphatic region
between 2.75 and 3.25 ppm, a feature that was assigned to
the two chemically different SCH2 groups present in 2,
whereas only one 4 H triplet was observed in the
2.75ꢀ3.25 ppm region for 1. The remaining aromatic and
aliphatic protons in 1 and 2 all appeared at the expected
chemical-shift values (see Supporting Information).
Initial evidence for the interactions between the recep-
tors 1 and 2 and Kþ/Agþ came from UVꢀvis absorption
spectroscopy. When KPF6 was added to a solution
(CH2Cl2/MeCN (1:1), 298 K) of the receptor 1 significant
spectral changes were observed. Figure 1a shows the
spectral changes of 1 upon addition of 0ꢀ7 equiv of
KPF6. As the concentration of KPF6 was increased, the
intensity of the absorption band at 350 nm decreases, while
the intensity at 410 nm increases leading to the formation
of two isosbestic points at 320 and 375 nm, respectively.
This observation is consistent with the formation of only
two absorption species, the free receptor 1 and the 1•Kþ
complex. The resulting binding profile (Figure 1a, insert) was
subjected to a standard nonlinear (1:1 host/guest) curve fitting
analysis11 from which a binding constant (Ka) of 2000 Mꢀ1
((20%) for the 1•Kþ complex was obtained. Similar titra-
tion experiments were carried out with Agþ, Pb2þ, Hg2þ
,
To increase the solubility and dissociation of KPF6 and
AgPF6, the solution-state binding studies involving the
Cd2þ, Pd2þ, Zn2þ, Naþ, and Liþ. However, no perceptible
changes in the UVꢀvis absorption spectra were observed
indicating that the receptor 1 is highly selective toward Kþ.
UVꢀvis titration experiments carried out between the
(7) Jeppesen, J. O.; Becher, J. Eur. J. Org. Chem. 2003, 3245–3266.
(8) For solubility reasons, the MPTTF derivatives 1 and 2 were
substituted with a tosyl group.
(9) Svenstrup, N.; Rasmussen, K. M.; Hansen, T. K.; Becher, J.
Synthesis 1994, 8, 809–812.
receptor 2 and Kþ, Agþ, Pb2þ, Hg2þ, Cd2þ, Pd2þ, Zn2þ
,
Naþ, and Liþ reveal that only in the case of Agþ significant
(10) (a) Jeppesen, J. O.; Takimiya, K.; Jensen, F.; Becher, J. Org.
Lett. 1999, 1, 1291–1291. (b) Jeppesen, J. O.; Takimiya, K.; Jensen, F.;
Brimert, T.; Nielsen, K.; Thorup, N.; Becher, J. J. Org. Chem. 2000, 65,
5794–5805.
(11) Connors, K. A. Binding Constants; John Wiley & Son: New York,
1987.
Org. Lett., Vol. 15, No. 7, 2013
1453

A comparison (Figure 2) of the 1H NMR spectra
(400 MHz, 298 K) recorded in a 1:1 mixture of CDCl3
and CD3CN of the free receptor 2 and the receptor 2 in the
presence of AgPF6 reveals significant chemical-shift differ-
ences for several of the resonances associated with the
protons in the receptor 2, which indicates that the Agþ ion
is complexed to 2. In particular, the resonances associated
withthearomaticprotons(Scheme1) are downfieldshifted
from δ = 7.25 ppm to δ = 7.62 ppm (Δδ = þ0.37 ppm)
and from δ=7.42ppmtoδ=7.95ppm(Δδ=þ0.53 ppm)
for H2 and H1, respectively, whereas the aliphatic H4 and H7
protons (Scheme 1) are upfield shifted from δ = 3.75 ppm
to δ = 3.27 ppm (Δδ = ꢀ0.48 ppm) and δ = 3.60 (Δδ =
ꢀ0.15 ppm) indicating that these protons are highly
affected by the complexation of Agþ inside the cavity of
the dithiabenzo-crown ether moiety. Addition of more than
1 equiv of Agþ did not cause any significant changes in the
1H NMR spectra of 2, thus confirming that 2 forms a 1:1
complex with Agþ. As a consequence of the strong binding
(vide supra) between the receptor 2 and Agþ and the fact
1
that the H NMR titration experiments were carried out
with the concentration of 2 20 times higher than in the case
of the corresponding UVꢀvis titration experiments, it was
not possible to obtain a reliable Ka value for the binding
between 2 and Agþ using 1H NMR spectroscopy, because
the binding profile tended toward linearity with increasing
Agþ concentration, which made the nonlinear curve fitting
procedure unreliable13 for estimating the Ka value.
Figure 1. (a) UVꢀvis spectra recorded in CH2Cl2/MeCN (1:1) at
298 K of a 0.1 mM solution of 1 upon titration with Kþ. The insert
shows the binding profile obtained using the 410 nm band as a probe
and its nonlinear curve fit to a 1:1 model. (b) UVꢀvis spectra recorded
in CH2Cl2/MeCN (1:1) at 298 K of a 0.1 mM solution of 2 upon
titrationwithAgþ. The insert shows the binding profile obtained using
the 420 nm band as a probe and its nonlinear curve fit to a 1:1 model.
1
Similar H NMR spectroscopic titration experiments
wereconductedwith solutions containing Kþ, Pb2þ, Hg2þ
,
Cd2þ, Pd2þ, Zn2þ, Naþ, and Liþ. However, no perceptible
changes in the 1H NMR spectra were observed supporting
the notion that the 2 is highly selective toward Agþ.
changes were observed in the UVꢀvis absorption spectra.
Titration (Figure 1b) of AgPF6 into a solution (CH2Cl2/
MeCN (1:1), 298 K) of the receptor 2 shows that the intensity
of the absorption band at 370 nm decreases while the intensity
at 420 nm increases with the concomitant formation of two
isosbestic points at 328 and 385 nm, respectively. This feature
indicates the presence of only two absorption species, the free
receptor 2 and the complex 2•Agþ. A nonlinear curve fitting
analysis11 of the resulting binding profile (Figure 1b, insert)
afforded a Ka value of 700 000 Mꢀ1 ((20%) for the 2•Agþ
complex indicating that the receptor 2is both highly sensitive12
and selective toward Agþ.
The binding between receptor 1and Kþ, Agþ, Pb2þ, Hg2þ
,
Cd2þ, Pd2þ, Zn2þ, Naþ, and Liþ were also studied using
1H NMR spectroscopy. As observed in the UVꢀvis titration
experiments (vide supra), only the addition of KPF6 to a
solution (CDCl3/CD3CN (1:1), 298 K) of the receptor 1
produced observable changes in the 1H NMR spectra
(Figure S20). However, the observed chemical-shift differ-
ences for the formation of the 1•Kþ complex were smaller
and required a higher concentration of the cation as
1
Figure 2. Partial H NMR spectra (400 MHz, 298 K) recorded in a 1:1 mixture of CDCl3/CD3CN of the receptor 2 (2 mM) upon
titration with Agþ.
1454
Org. Lett., Vol. 15, No. 7, 2013

compared to the formation of the 2•Agþ complex indicating
that the 1•Kþ complex is weaker than the 2•Agþ complex.
UVꢀvis and 1H NMR spectroscopic studies revealed that 1
and 2 are able to complex cations in solution. However, these
receptors were also designed to allow the binding events to be
followed via cation-induced changes in the electrochem-
ical properties of the MPTTF units. Cyclic voltammetry
(CV) was used to probe the changes in the redox potentials of
irreversible formation of the 2•Agþ complex, the simulations
did not allow for the evaluation of the equilibrium constant
for its formation. However, good agreement between
the experimental and the simulated data was obtained for
values of Ka larger than approximately 10 000 Mꢀ1. When
considered in concert, the electrochemical investigations
support the spectroscopic analyses and provide further
evidence that 1 and 2 are highly selective toward Kþ and
Agþ, respectively.
1 and 2 upon complexation with Kþ, Agþ, Pb2þ, Hg2þ
,
Cd2þ, Pd2þ, Zn2þ, Naþ, and Liþ. In the case of 1, only the
addition of Kþ produced perceptible changes in the cyclic
voltammograms (CVs) recorded in a 1:1 mixture of CH2Cl2
and MeCN at 298 K, whereas receptor 2 only produced
observable changes in the CVs upon addition of AgPF6 in
accordance with the results obtained from the UVꢀvis and
1H NMR spectroscopic investigations.
In summary we have designed and synthesized two new
redox-active MPTTF based receptors and demonstrated
how small modifications in the dibenzo crown ether ring
can have a huge impact on the selectivity toward cations.
This finding is undoubtedly not unimportant when it
comes to the future design of redox-active molecular
receptors based on TTF and crown ethers.
The progressive addition (Figure 3a) of KPF6 to a solution
of the receptor 1 in CH2Cl2/MeCN (1:1) at 298 K resulted in
a small anodic shift of the first oxidation potential asso-
ciated with the MPTTF unit, whereas the second oxidation
potential was unaffected by the presence of KPF6. These
results can be explained as follows: addition of KPF6 to a
solution of 1 leads to the formation of the 1•Kþ complex;
after the first oxidation of the MPTTF unit, decomplexation
takes place caused by Coulombic repulsion between the
mono-oxidized MPTTF•þ unit and Kþ. Consequently, the
second oxidation of the MPTTF•þ unit in the mixture of
uncomplexed receptor 1•þ and uncomplexed Kþ takes place
at the same potential as that of the second oxidation of 1.
A more distinct response in the CVs of receptor 2 upon
addition of Agþ wasobserved. Incontrast to the formation
ofthe 1•Kþ complex, the kinetics for the dissociation of the
2•Agþ complex were found (Figure 3b) to be slow on the
CV time scale, and both complexed and uncomplexed
species can be observed in the CVs recorded in CH2Cl2/
MeCN (1:1) at 298 K. Thus, the addition of between 0 and
1 molar equiv of AgPF6 to a solution of the receptor 2
allows the oxidation potential for both the free receptor 2
(anodic peak potential: E1ox = þ0.18 V) and the complex
2•Agþ (E1ox = þ0.34 V) to be determined.
Figure 3. (a) CVs of the receptor 1 (1.0 mM) recorded in a 1:1
mixture of CH2Cl2 and MeCN at 298 K by adding in increasing
quantities of a concentrated MeCN solution of KPF6. (b) CVs
of the receptor 2 (1.0 mM) recorded in a 1:1 mixture of CH2Cl2
and MeCN at 298 K by adding increasing quantities of
a concentrated MeCN solution of AgPF6. (The potentials
were referred to ferrocene, the supporting electrolyte was
n-Bu4NPF6, and the scan rate was 200 mV sꢀ1.)
ox
Therefore ΔE1 = þ0.16 V, which to the best of our
knowledge is one of the highest values observed so far for
cation responsive receptors based on TTF. The second
oxidation potential of 2 was only slightly affected by the
addition of AgPF6 indicating that the mono-oxidized
receptor 2•þ has a much lower affinity toward Agþ. Simula-
tion of the electrochemical titration data using DigiSim3.05
(BAS Inc.14) provided a Ka value of 2000 Mꢀ1 ((20%) for
the formation of the 1•Kþ complex. Owing to the essentially
Acknowledgment. This research was supported by the
Villum Foundation and The Danish Natural Science
Reasearch Council (FNU, Project No. 11-106744).
(12) A value of 700 000 Mꢀ1 for the formation of the 2•Agþ complex
is more than 2 orders of magnitude higher as compared to other TTF
based sensors for Agþ; see: (a) Le Derf, F.; Mazari, M.; Mercier, N.;
Supporting Information Available. Experimental proce-
dures for the synthesis of 1, 2, 6ꢀ9, 11, and 12; spectroscopic
characterization of 1 and 2; description of titration experi-
ments; and description of electrochemical measurements.
This material is available free of charge via the Internet at
http://pubs.acs.org.
ꢀ
Levillain, E.; Gorgues, A.; Salle, M.; Richomme, P.; Becher, J.; Garın,
G.; Orduna, J. Chem. Commun. 1999, 1417–1418. (b) Johnston, B.;
Goldenberg, L. M.; Bryce, M. R.; Kataky, R. J. Chem. Soc., Perkin
Trans. 2 2000, 189–190.
(13) (a) de Boer, J. A. A.; Reinhoudt, D. N. J. Am. Chem. Soc. 1985,
107, 5347–5351. (b) Fielding, L. Tetrahedron 2000, 56, 6151–6170.
(14) Rudolf, M.; Feldberg, S. W. DigiSim 3.03a; Bioanalytical
Systems, Inc.: West Lafayette, IN 47906 USA.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 7, 2013
1455