Synthesis, structure, anion binding, and sensing by calix[4]pyrrole isomers

Article abstract of DOI:10.1021/ja0622150

The synthesis, structure, and anion binding properties of chromogenic octamethylcalix[4]pyrroles (OMCPs) and their N-confused octamethylcalix[4] pyrrole isomers (NC-OMCPs) containing an inverted pyrrole ring connected via α′- and β-positions are described

Full text of DOI:10.1021/ja0622150

Published on Web 08/16/2006
Synthesis, Structure, Anion Binding, and Sensing by
Calix[4]pyrrole Isomers
Ryuhei Nishiyabu,† Manuel A. Palacios,† Wim Dehaen,‡ and
Pavel Anzenbacher, Jr.*,†
Contribution from the Department of Chemistry and Center for Photochemical Sciences,
Bowling Green State UniVersity, Bowling Green, Ohio 43403, and Department of Chemistry,
Katholieke UniVersiteit LeuVen, Celestijnenlaan 200F, 3001 LeuVen, Belgium
Received April 14, 2006; E-mail: pavel@bgnet.bgsu.edu
Abstract: The synthesis, structure, and anion binding properties of chromogenic octamethylcalix[4]pyrroles
(OMCPs) and their N-confused octamethylcalix[4]pyrrole isomers (NC-OMCPs) containing an inverted pyrrole
ring connected via R′- and â-positions are described. X-ray diffraction analyses proved the structures of
two synthesized isomeric pairs of OMCPs and NC-OMCPs. The addition of anions to solutions of
chromogenic OMCPs and NC-OMCPs resulted in different colors suggesting different anion-binding
behaviors. The chromogenic NC-OMCPs showed significantly stronger anion-induced color changes
compared to the corresponding chromogenic OMCP, and the absorption spectroscopy titrations indicated
that chromogenic OMCPs and NC-OMCPs also possess different anion binding selectivity. Detailed NMR
studies revealed that this rather unusual feature stems from a different anion-binding mode in OMCPs and
NC-OMCPs, one where the â-pyrrole C-H of the inverted pyrrole moiety participates in the hydrogen-
bonded anion-NC-OMCP complex. Preliminary colorimetric microassays using synthesized chromogenic
calixpyrroles embedded in partially hydrophilic polyurethane matrices allow for observation of analyte-
specific changes in color when the anions are administered in the form of their aqueous solutions and in
the presence of weakly competing anions.
roles,19-27 calixbipyrrole,28-32 strapped calixpyrrole,33-37 calix-
pyrrole dimer,38 and cryptand-like calixpyrrole.39 Furthermore,
Introduction
Octamethylcalix[4]pyrrole (OMCP, 1), a tetrapyrrole mac-
rocycle first prepared by Baeyer,1 garnered significant attention
recently due to its ability to bind small anions2 and electroneutral
molecules.3 After the discovery of its supramolecular properties,1-8
OMCP has become the subject of numerous efforts devoted to
understanding, improving, and tuning the binding affinity and
selectivity toward anions by preparation of substituted OMCP9-18
and further related compounds such as expanded calix[n]pyr-
(14) Woods, C. J.; Camiolo, S.; Light, M. E.; Coles, S. J.; Hursthouse, M. B.;
King, M. A.; Gale, P. A.; Essex, J. W. J. Am. Chem. Soc. 2002, 124, 8644-
8652.
(15) Camiolo, S.; Gale, P. A. Chem. Commun. 2000, 1129-1130.
(16) Anzenbacher, P., Jr.; Jurs´ıkova´, K.; Lynch, V. M.; Gale, P. A.; Sessler, J.
L. J. Am. Chem. Soc. 1999, 121, 11020-11021.
(17) Bonomo, L.; Solari, E.; Toraman, G.; Scopelliti, R.; Latronico, M.; Floriani,
C. Chem. Commun. 1999, 2413-2414.
(18) Gale, P. A.; Sessler, J. L.; Lynch, V.; Sansom, P. I. Tetrahedron Lett. 1996,
37, 7881-7884.
(19) Sessler, J. L.; Cho, W.-S.; Gross, D. E.; Shriver, J. A.; Lynch, V. M.;
Marquez, M. J. Org. Chem. 2005, 70, 5982-5986.
† Bowling Green State University.
‡ Katholieke Universiteit Leuven.
(20) Piatek, P.; Lynch, V. M.; Sessler, J. L. J. Am. Chem. Soc. 2004, 126,
16073-16076.
(1) Baeyer, A. Ber. Dtsch. Chem. Ges. 1886, 19, 2184-2185.
(2) Gale, P. A.; Sessler, J. L.; Kra´l, V.; Lynch, V. J. Am. Chem. Soc. 1996,
118, 5140-5141.
(21) Cafeo, G.; Kohnke, F. H.; Parisi, M. F.; Nascone, R. P.; La Torre, G. L.;
Williams, D. J. Org. Lett. 2002, 4, 2695-2697.
(3) Allen, W. E.; Gale, P. A.; Brown, C. T.; Lynch, V. M.; Sessler, J. L. J.
Am. Chem. Soc. 1996, 118, 8, 12471-12472.
(22) Sessler, J. L.; Anzenbacher, P., Jr.; Shriver, J. A.; Jurs´ıkova´, K.; Lynch,
V. M.; Marquez, M. J. Am. Chem. Soc. 2000, 122, 12061-12062.
(23) Arumugam, N.; Jang, Y.-S.; Lee, C.-H. Org. Lett. 2000, 2, 3115-3117.
(24) Cafeo, G.; Kohnke, F. H.; La Torre, G. L.; White, A. J. P.; Williams, D.
J. Angew. Chem., Int. Ed. 2000, 39, 1496-1498.
(4) Custelcean, R.; Delmau, L. H.; Moyer, B. A.; Sessler, J. L.; Cho, W.-S.;
Gross, D.; Bates, G. W.; Brooks, S. J.; Light, M. E.; Gale, P. A. Angew.
Chem., Int. Ed. 2005, 44, 2537-2542.
(5) Nielsen, K. A.; Cho, W.-S.; Jeppesen, J. O.; Lynch, V. M.; Becher, J.;
Sessler, J. L. J. Am. Chem. Soc. 2004, 126, 16296-16297.
(6) Gale, P. A.; Sessler, J. L.; Kra´l, V. Chem. Commun. 1998, 1-8.
(7) Gale, P. A.; Anzenbacher, P., Jr.; Sessler, J. L. Coord. Chem. ReV. 2001,
222, 57-102.
(25) Cafeo, G.; Kohnke, F. H.; La Torre, G. L.; White, A. J. P.; Williams, D.
J. Chem. Commun. 2000, 1207-1208.
(26) Turner, B.; Botoshansky, M.; Eichen, Y. Angew. Chem., Int. Ed. 1998, 37,
2475-2478.
(27) Gale, P. A.; Genge, J. W.; Kra´l, V.; McKervey, M. A.; Sessler, J. L.; Walker,
A. Tetrahedron Lett. 1997, 38, 8443-8444.
(8) Sessler, J. L.; Andrievsky, A.; Gale, P. A.; Lynch, V. Angew. Chem., Int.
Ed. Engl. 1996, 35, 2782-2785.
(28) Sessler, J. L.; An, D.; Cho, W.-S.; Lynch, V. M.; Yoon, D.-W.; Hong,
S.-J.; Lee, C.-H. J. Org. Chem. 2005, 70, 1511-1517.
(29) Sessler, J. L.; An, D.; Cho, W.-S.; Lynch, V.; Marquez, M. Chem.sEur.
J. 2005, 11, 2001-2011.
(9) Miyaji, H.; An, D.; Sessler, J. L. Supramol. Chem. 2001, 13, 661-669.
(10) Anzenbacher, P., Jr.; Try, A. C.; Miyaji, H.; Jurs´ıkova´, K.; Lynch, V. M.;
Marquez, M.; Sessler, J. L. J. Am. Chem. Soc. 2000, 122, 10268-10272.
(11) Anzenbacher, P., Jr.; Jurs´ıkova´, K.; Shriver, J. A.; Miyaji, H.; Lynch, V.
M.; Sessler, J. L.; Gale, P. A. J. Org. Chem. 2000, 65, 7641-7645.
(12) Gale, P. A.; Sessler, J. L.; Allen, W. E.; Tvermoes, N. A.; Lynch, V. Chem.
Commun. 1997, 665-666.
(30) Sessler, J. L.; An, D.; Cho, W.-S.; Lynch, V.; Marquez, M. Chem. Commun.
2005, 540-542.
(31) Sessler, J. L.; An, D.; Cho, W.-S.; Lynch, V. J. Am. Chem. Soc. 2003,
125, 13646-13647.
(13) Warriner, C. N.; Gale, P. A.; Light, M. E.; Hursthouse, M. B. Chem.
Commun. 2003, 1810-1811.
(32) Sessler, J. L.; An, D.; Cho, W.-S.; Lynch, V. Angew. Chem., Int. Ed. 2003,
42, 2278-2281.
9
11496
J. AM. CHEM. SOC. 2006, 128, 11496-11504
10.1021/ja0622150 CCC: $33.50 © 2006 American Chemical Society

Calix[4]pyrrole Isomers
A R T I C L E S
OMCP has been applied in fabrication of HPLC supports for
anion separation,40 ion extractant,41 and, most notably, optical42-46
and electrochemical47-51 anion sensors. This is mainly because
calix[4]pyrroles, despite binding anions via hydrogen bonds,
show unique affinity for biologically important anions such as
chloride and phosphate, even in the presence of competing
media such as water and electrolytes.42 Recently, N-confused
calix[4]pyrrole (NC-OMCP, 2), an isomer containing an inverted
pyrrole ring connected via R′- and â-positions, was also reported
by Dehaen.52
Figure 1. Structures of Azo-1, Azo-2, TCE-1, and TCE-2.
Scheme 1
Our recent work on electrophilic aromatic substitution of
OMCP led us to investigate the reactions of OMCP with various
electrophiles, including tetracyanoethene and diazonium salts
to obtain chromogenic OMCP derivatives.42,53
In their independent studies, Dehaen and co-workers54 also
synthesized a series of azo-substituted NC-OMCPs (e.g., Azo-
2) and demonstrated changes in color of these chromogenic NC-
OMCPs in the presence of anions that indicated formation of
anion-receptor complexes. We have been intrigued by the
possibility to synthesize pairs of regular and N-confused calix-
[4]pyrrole receptors and sensors and compare their anion-binding
behaviors because we believed that both OMCPs and their
corresponding NC-OMCPs could show unique features stem-
ming from a different structure and different pKa of hydrogen-
bond donors.55,56
Here we report on syntheses, structures, and anion-binding
properties of chromogenic NC-OMCPs together with corre-
sponding OMCPs. Previously we have synthesized chromogenic
OMCPs such as TCE-142 and Azo-157 by electrophilic substitu-
tion reaction and studied their anion binding properties.
Particularly TCE-1 demonstrated its strong anion binding
affinity and selectivity as well as ability to act as colorimetric
sensors for anions.42 For this study, we have developed a new
high-yielding synthesis of NC-OMCP, adopted the above
modifications for the preparation of NC-OMCP-based conge-
ners, and compared anion-binding properties of the chromogenic
OMCP and NC-OMCP as well as the parent OMCP and NC-
OMCP. The structures of the isomer pairs Azo-1, Azo-2, TCE-
1, and TCE-2 are shown in Figure 1.
(33) Lee, C.-H.; Lee, J.-S.; Na, H.-K.; Yoon, D.-W.; Miyaji, H.; Cho, W.-S.;
Sessler, J. L. J. Org. Chem. 2005, 70, 2067-2074.
(34) Panda, P. K.; Lee, C.-H. J. Org. Chem. 2005, 70, 3148-3156.
(35) Panda, P. K.; Lee, C.-H. Org. Lett. 2004, 6, 671-674.
(36) Lee, C.-H.; Na, H.-K.; Yoon, D.-W.; Won, D.-H.; Cho, W.-S.; Lynch, V.
M.; Shevchuk, S. V.; Sessler, J. L. J. Am. Chem. Soc. 2003, 125, 7301-
7306.
(37) Yoon, D.-W.; Hwang, H.; Lee, C.-H. Angew. Chem., Int. Ed. 2002, 41,
1757-1759.
(38) Sato, W.; Miyaji, H.; Sessler, J. L. Tetrahedron Lett. 2000, 41, 6731-
6736.
(39) Bucher, C.; Zimmerman, R. S.; Lynch, V.; Sessler, J. L. J. Am. Chem.
Soc. 2001, 123, 9716-9717.
(40) Sessler, J. L.; Gale, P. A.; Genge, J. W. Chem.sEur. J. 1998, 4, 1095-
1099.
Results and Discussion
Both the regular octamethylcalix[4]pyrrole (1) and N-con-
fused octamethylcalix[4]pyrrole (2) were obtained from con-
densation of acetone and pyrrole in a MeOH/MeSO3H system
in 40% and 6% yield, respectively (Scheme 1).2,52 For the
purpose of this study an adaptation of this method yielding
multigram yields of the starting NC-OMCP was developed. This
methodology, described in the Supporting Information, provides
reliable yields of 65% and 15% for OMCP and NC-OMCP,
respectively.
To obtain unambiguous evidence supporting the structural
assignment of 2, we attempted the preparation of crystals of 2
for X-ray diffraction analysis from several organic solvents, but
the very similar sizes of C-H and N-H pyrrole moieties in 2
resulted mostly in highly disordered crystals as the inverted
(41) Levitskaia, T. G.; Marquez, M.; Sessler, J. L.; Shriver, J. A.; Vercouter,
T.; Moyer, B. A. Chem. Commun. 2003, 2248-2249.
(42) Nishiyabu, R.; Anzenbacher, P., Jr. J. Am. Chem. Soc. 2005, 127, 8270-
8271.
(43) Miyaji, H.; Sato, W.; Sessler, J. L. Angew. Chem., Int. Ed. 2000, 39, 1777-
1780.
(44) Miyaji, H.; Sato, W.; Sessler, J. L.; Lynch, V. M. Tetrahedron Lett. 2000,
41, 1369-1373.
(45) Anzenbacher, P., Jr.; Jurs´ıkova´, K.; Sessler, J. L. J. Am. Chem. Soc. 2000,
122, 9350-9351.
(46) Miyaji, H.; Anzenbacher, P., Jr.; Sessler, J. L.; Bleasdale, E. R.; Gale, P.
A. Chem. Commun. 1999, 1723-1724.
(47) Nielsen, K. A.; Jeppesen, J. O.; Levillain, E.; Becher, J. Angew. Chem.,
Int. Ed. 2003, 42, 187-191.
(48) Gale, P. A.; Bleasdale, E. R.; Chen, G. Z. Supramol. Chem. 2001, 13, 557-
563.
(49) Gale, P. A.; Hursthouse, M. B.; Light, M. E.; Sessler, J. L.; Warriner, C.
N.; Zimmerman, R. S. Tetrahedron Lett. 2001, 42, 6759-6762.
(50) Kra´l, V.; Sessler, J. L.; Shishkanova, T. V.; Gale, P. A.; Volf, R. J. Am.
Chem. Soc. 1999, 121, 8771-8775.
(51) Sessler, J. L.; Gebauer, A.; Gale, P. A. Gazz. Chim. Ital. 1997, 127, 723-
726.
(52) Depraetere, S.; Smet, M.; Dehaen, W. Angew. Chem., Int. Ed. 1999, 38,
3359-3361.
(55) Nechita, M. T.; Lotrean, S.; Radecki, J.; Radecka, H.; Depreatere, S.;
Dehaen, W. Pol. J. Food. Nutr. Sci. 2003, 12 (SI 2), 81-87.
(56) Piotrowski, T.; Radecka, H.; Radecki, J.; Depraetere, S.; Dehaen, W. Mater.
Sci. Eng., C 2001, C18, 223-228.
(57) Wex, B.; Anzenbacher, P., Jr. Abstracts of papers; 223rd National Meeting
of the American Chemical Society, Orland, FL, Apr 7-11, 2002; American
Chemical Society: Washington, DC, 2002; ORGN-245.
(53) Nishiyabu, R.; Anzenbacher, P., Jr. Abstract Book; 13th International
Symposium on Supramolecular Chemistry, Notre Dame, IN, Jul 25-30,
2004; P-3-48.
(54) Gu, R.; Depraetre, S.; Kotek, J.; Budka, J.; Wagner-Wysiecka, E.; Biernat,
J. F.; Dehaen, W. Org. Biomol. Chem. 2005, 3, 2921-2923.
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11497

A R T I C L E S
Nishiyabu et al.
Figure 3. Tautomerization of TCE-2 to its cyclic form TCE-2C explains
1
three broad H NMR resonances attributed to NH moieties and one sharp
singlet attributed to the imine dNH.
Figure 2. X-ray crystal structure of 2 (thermal ellipsoids are scaled to the
30% probability level; the methyl hydrogen atoms have been removed for
clarity). (Left) Structure of the receptor 2. (Right) Inclusion of the two
DMSO molecules that stabilize the conformation of the receptor in the
crystal.
Scheme 2
Figure 4. X-ray crystal structures of TCE-1 (left) and TCE-2C (right)
(ORTEP, the displacement ellipsoids are scaled to the 50% probability level;
the hydrogen atoms have been removed for clarity).
pyrrole was not always found in a crystallographically unique
position. This, together with the lack of stability of 2 due to a
free reactive R-pyrrole moiety,52 made it difficult to grow X-ray
quality crystals. After numerous, largely unsuccessful attempts
we realized that binding between 2 and a suitable hydrogen bond
acceptor could lock the receptor in a bulkier and more spatially
unique complex. Consequently diffraction grade crystals of 2
were grown in dimethyl sulfoxide to obtain hydrogen bonded
complex 2‚2DMSO. The single-crystal X-ray analysis clearly
showed the structure of 2 in which one pyrrole ring is linked
through its R′- and â-positions instead of R- and R′-positions
in 1 (Figure 2).
Figure 5. X-ray crystal structures of Azo-1 (left) and Azo-2 (right)
(ORTEP, the displacement ellipsoids are scaled to the 50% probability level;
the methyl hydrogen atoms have been removed for clarity).
The crystal structure revealed that TCE-2C contains the bicyclic
pyrrolizin-3-ylideneamine moiety (Figure 4).
Azo-substituted OMCP(Azo-1) and NC-OMCP (Azo-2) were
also prepared by electrophilic substitution with 4-nitrobenzene-
diazonium tetrafluoroborate.53,54,57 Electrophilic substitution of
2 with 4-nitrobenzenediazonium tetrafluoroborate in THF in the
presence of triethylamine yielded Azo-2 in 30% yield, while
the same reaction using 1 gave a complex mixture of products
including Azo-1 in lower yield (9%). The observed lower yield
of Azo-1 could be attributed to a large number of equally
reactive pyrrole â-positions as well as to lower reactivity of
the â-position in 1 compared to the R-position in 2.60,61 The
single crystals of Azo-1 and Azo-2 were obtained from N,N-
dimethylformamide and dichloromethane/ether, respectively.
The crystal structures of Azo-1 and Azo-2 are shown in Figure
5.
The attachment of dye precursors to 2 also converts the
originally colorless species into chromogenic materials, which
show changes in color upon binding of an anion. Both the naked-
eye observation and the absorption spectroscopy revealed
different anion binding behaviors of chromogenic OMCPs and
NC-OMCPs (Figures 6 and 7).
Tricyanoethylene substituted 1 (TCE-1) was prepared by
electrophilic substitution of 1 with tetracyanoethylene.42 The
same reaction using 2 gave TCE-2 (Scheme 2) where the
tricyanoethylene moiety was attached at the more electron-rich
R-position. This reaction yielded only one product (36% yield)
as a purple solid.
1
Interestingly, H NMR spectrum of the obtained compound
assumed to be TCE-2 showed only three broad pyrrole NH
singlet signals (8.53, 9.19, and 9.45 ppm in DMSO-d6) instead
of four broad pyrrole NH singlets. The disappearance of the
R-pyrrole CH proton signal indicated substitution of the
R-pyrrole position with the tricyanoethylene moiety, and a mass
spectrum of this compound showed a molecular ion peak
corresponding with that of TCE-2. The NMR spectra suggested
that the obtained compound was not TCE-2 but its cyclic form,
resulting from intramolecular cyclization of the tricyanoethylene
moiety with inverted pyrrole nitrogen (Figure 3).58,59 Indeed,
in the NMR spectrum of the obtained compound there was one
sharp singlet at 11.16 ppm that could be assigned to the imine
dNH proton in TCE-2C.
The synthesized isomeric pairs were tested for their ability
to bind anions, first by using strongly binding (F-) and then by
weakly binding (Cl-) anions. Absorption spectra of TCE-1 in
the presence of fluoride in DMSO showed a moderate spectral
change and a pink-to-orange color change, while that of chloride
showed a small spectral change without a noticeable color
The final confirmation of the structure assignment of TCE-
2C came from single-crystal X-ray analysis. The crystal structure
of TCE-2C is shown in Figure 2 with the structure of TCE-1.
(58) Fares, V.; Flamini, A.; Poli, N. J. Am. Chem. Soc. 1995, 117, 11580-
11581.
(59) Collange, E.; Flamini, A.; Poli, R. J. Phys. Chem. A 2002, 106, 200-208.
(60) Depraetere, S.; Dehaen, W. Tetrahedron Lett. 2003, 44, 345-347.
(61) Bulter, A. R.; Shepherd, P. T. J. Chem. Soc., Perkin, Trans. 2 1980, 113-
116.
9
11498 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006

Calix[4]pyrrole Isomers
A R T I C L E S
Figure 6. Absorption spectra of TCE-1 (left) and TCE-2C (right), both 5.0 × 10-5 M in DMSO, in the absence of anion (black), with 10 equiv of fluoride
(red), and with an excess of chloride (green). Insets: Solutions of TCE-1 and TCE-2C in the absence of anion and with fluoride (10 equiv) and chloride
(excess). Bu4N+ (TBA) salts of anions were used.
Figure 7. Absorption spectra of Azo-1 (left) and Azo-2 (right), both 5.0 × 10-5 M in DMSO, in the absence of anion (black), with 10 equiv of fluoride
(red), and with an excess of chloride (green). Insets: Solutions of Azo-1 and Azo-2 in the absence of anion, with fluoride (10 equiv), and with chloride
(excess). Bu4N+ (TBA) salts of anions were used.
change (Figure 6, left).42 The solution of TCE-2C in DMSO
Table 1. Association Constants for TCE-1, TCE-2c, Azo-1, and
Azo-2 and Anions in DMSO at 22 °C Determined by Absorption
showed a vivid pink color. The addition of fluoride to TCE-
2C resulted in enhanced absorption in the 500 nm region, with
a corresponding dramatic pink-to-orange color change. On the
other hand, the addition of chloride into the TCE-2C solution
resulted in a small spectral change but an easily discernible pink-
to-purple color change (Figure 6, right).
Spectroscopy Titrationsa
Absorption spectroscopic titrations of Azo-1 and Azo-2 with
anions were also performed. In preliminary qualitative experi-
ments, the addition of fluoride and chloride anions into DMSO
solutions of Azo-1 resulted in the formation of two new maxima
located at 385 and 467 nm and yellow-to-orange color changes
(Figure 7, left). Conversely, as Dehaen reported, titrations of
Azo-2 with fluoride showed a large bathochromic shift with a
dramatic orange-to-blue change in color, possibly due to the
deprotonation of the inverted pyrrole NH.54 Furthermore, Azo-2
showed upon the addition of chloride an orange-to-red color
change (Figure 7, right). The quantitative titration experiments
allowed us to determine the apparent binding constants of the
present chromogenic OMCPs and NC-OMCPs with anions, and
the respective Kas values are listed in Table 1.
The data in Table 1 suggest that the sensors based on the
regular OMCP receptor (TCE-1 and Azo-1) favor spherical
anions such as fluoride and chloride. This is rationalized by
the symmetry of the respective sensor-anion complex. The
symmetrical conelike conformation of the OMCP receptor
appears to be more suited for binding of the spherical anions.
In contradistinction, sensors based on the N-confused OMCP
receptor (TCE-2/2C and Azo-2) cannot adopt the almost
perfectly symmetrical conelike conformation. We believe that
a Anions were used in the form of their Bu4N+ (TBA) salts. The errors
in all fits are <15%. b See ref 42. c Association constants were calculated
on the assumption that pyrophosphate forms a dimer in DMSO.62
the twisted conelike conformation available to the N-confused
OMCP receptor and the corresponding sensors is better suited
for anions that can bind through more than one atom, for
example, two oxygen atoms in the delocalized carboxylate anion.
Thus comparison of binding affinities for TCE-1 and TCE-
2C shows that both compounds strongly bind fluoride (F) and
acetate (AcO) and both bind chloride (Cl) weakly. TCE-1, like
other regular calix[4]pyrroles, does not bind dihydrogenphos-
phate (H2P) strongly, but shows high affinity for hydrogenpy-
rophosphate (HPP). By comparison, TCE-2C that adopts a
distorted conelike conformation binds dihydrogenphosphate as
well as acetate with high affinity. The specific order of the
binding affinities is as follows: TCE-1: F> HPP > AcO >
H2P > Cl and TCE-2C: AcO > F > H2P > Cl (while the
binding of pyrophosphate follows a complex pattern that cannot
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11499

A R T I C L E S
Nishiyabu et al.
Figure 8. 1H NMR titrations (selected regions) of 1.0 × 10-2 M solutions of (a) 1 and (b) 2 with chloride in DMSO-d6. The inverted pyrrole 1H resonances
in 2 are labeled as follows: NHi (red), R-CHi (green), and â-CHi (blue).
be easily numerically evaluated). Similarly, both Azo-1 and
Azo-2 show affinity for fluoride, acetate, and hydrogenpyro-
phosphate, albeit weaker compared to the TCE-sensors. The
specific order of the binding affinities is as follows: Azo-1: F
> HPP > AcO > H2P > Cl and Azo-2: AcO > F > HPP >
H2P > Cl.
state and the respective receptor-anion complex. The fact that
the changes in color, understood as a change in absorption
wavelengths as well as a change in the absorption coefficient,
which are not the same for all of the sensor-anion combinations,
adds another dimension to the potential application of the
reported sensors as they may be perhaps used for qualitative
substrate determination in a naked-eye regime. Table 1 therefore
not only shows the actual values of affinity constants but also
provides a qualitative indication of the colors of the sensors
and their complexes with anions.
Originally, we believed that the different affinities and
selectivities between OMCP and NC-OMCP-based sensors for
anions are due to the inverted-pyrrole effects on adopting a
symmetrical cone conformation of the receptor, while the
chromophore responsible for color change is attached to the
R-position, which is closer to the pyrrole NH hydrogen bond
donor then it is in the regular OMCP-based sensors where the
chromophore is attached to the â-position. Later we realized
that the dramatic changes in color response are unlikely to be
due to a simple distance effect but most likely due to a different
distribution of the frontier molecular orbitals and the corre-
sponding π-π* transitions in the chromophores. This feature
is discussed later in this paper.
Interestingly, both N-confused sensors (TCE-2/2C and Azo-
2) show a dramatic increase in selectivity toward the acetate
anion over chloride and phosphate compared to their counter-
parts (TCE-1 and Azo-1) as evidenced by the respective
selectivity coefficients expressed as ratios of binding constants
KAcO-/KCl- and KAcO-/KH PO -. For example, the affinity for
2
4
acetate over chloride increased more than 50 times for TCE-
2C vs regular TCE-1 and 28 times for Azo-2 vs the regular
Azo-1. The aspect of carboxylate-chloride selectivity is of
potential practical importance, as this feature may allow sensing
of carboxylate in biological milieu where the chloride and
phosphate usually abound in high concentrations.
From the combination of the spectroscopic investigations and
binding data expressed as relative affinity constants (Table 1),
one can see that the relative strength in the respective color
transitions observed for the sensors and their complexes with
anions does not necessarily correlate with the receptor-anion
affinities (Kas). That is most likely because the extent of the
color transitions relates to the magnitude of the change in the
transition dipole moment between the sensor in the resting state
and in the complex. The magnitude of the change in the
transition dipole moment does not generally correlate with the
thermodynamic association constants but rather with the degree
of difference in the frontier orbital density between the resting
Further insight into anion complexation was achieved by
examination of anion binding of 1 and 2 by 1H NMR titrations.
Titration of 1 by chloride showed a concerted downfield shift
of four pyrrole NHs and an upfield shift of pyrrole â-proton
signals corresponding to the formation of a symmetrical cone
conformation (Figure 8a).2,6,12,13 This is explained by the pyrrole
NHs appearing close to the anion, while the â-protons face away
9
11500 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006

Calix[4]pyrrole Isomers
A R T I C L E S
Figure 9. 1H NMR titrations (selected regions) of 1.0 × 10-2 M solutions of (a) Azo-1 and (b) Azo-2 with chloride in DMSO-d6. The inverted pyrrole 1H
resonances in Azo-2 are labeled as follows: NHi (red) and â-CHi (blue).
from the anion and π-electrons of the other pyrroles. To our
1
surprise, H NMR titration of 2 revealed a different pattern:
the three noninverted pyrroles show the usual behavior (a
downfield shift of NHs and upfield shift of the pyrrole â-H
signals), while the inverted pyrrole shows a concerted upfield
shift of the inverted pyrrole NH signal as well as a strong
downfield shift of the â-pyrrole proton (Figure 8b).1H NMR
titrations of Azo-1, Azo-2, TCE-1 (see Supporting Information),
and TCE-2C (see Supporting Information) with chloride show
the same type of behavior as 1 and 2, respectively, suggesting
the same binding behavior (Figure 9).
Figure 10. Calculated structures of 1‚TMAF (calix[4]pyrrole‚
tetramethylammonium fluoride) and 2‚TMAF complexes with 1:1 binding
mode. Gaussian 0364 was used for all calculations presented. For geometry
optimizations the first guesses were fully optimized by PM3 semiempirical
Hamiltonian. Then, the models were reoptimized at the B3LYP/6-31G*
level of theory. The TMA+ is omitted here for clarity (for more details see
Supporting Information).
These results imply that the NC-OMCP receptor of 2, Azo-
2, and TCE-2C bind anions via a different binding mode
compared to OMCP and OMCP-based sensors. Further insight
into the binding mode was obtained from the density functional
theory (DFT) study. From the combination of NMR data, and
DFT energy-minimized structures, it appears that this binding
mode, new in calixpyrrole-anion binding, utilizes the â-pyrrole
C-H in close interaction with the anion.49,63 Density functional
theory (DFT) calculations were carried out for both OMCP and
NC-OMCP and their respective complexes with tetramethyl-
ammonium fluoride (TMAF). The energy-optimized structures
of complexes 1‚TMAF and 2‚TMAF are shown on computer-
generated models in Figure 10.
NC-OMCPs prefer a more symmetrical, close to conelike
conformation involving hydrogen-bond-like interaction to a less
acidic â-pyrrole CH, rather than adopting the conformation
utilizing all four pyrrole NHs. The DFT calculations also lend
support to the experimental observation showing a spatial
arrangement for the 2‚TMAF complex with an energy minimum
for the conelike conformation. Conversely, DFT geometrical
optimizations aimed at determination of an energy minimum
with all four NHs participating in the fluoride anion binding
failed as such conformations are too strained. The binding of
basic anions such as a fluoride anion was also investigated by
1H NMR titration experiments of 1 and 2 combined with DFT
calculations. Titration of 1 by fluoride showed a decrease in
intensity of the NH proton signal of free 1, while a new,
downfield-shifted NH proton signal corresponding to the
receptor-fluoride complex appeared. In these titrations coupling
between those protons of 1 and the bound fluoride was observed.
The respective coupling constant is 39.9 Hz.2,6,12,13 The 1H NMR
titration of 2 with fluoride afforded results very similar to those
One might think that 2 would prefer distortion of the
receptor-anion complex to make all four pyrrole NHs available
for hydrogen bonding. We found that the anion complexes of
(62) Chu, F.; Flatt, L. S.; Anslyn, E. V. J. Am. Chem. Soc. 1994, 116, 4194-
4204.
(63) Camiolo, S.; Gale, P. A.; Hursthouse, M. B.; Light, M. E. Org. Biomol.
Chem. 2003, 1, 741-744.
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11501

A R T I C L E S
Nishiyabu et al.
Figure 11. (a) Calculated DFT-GIAO magnetic shields for NC-OMCP and NC-OMCP‚TMAF in 1:1 binding mode at the B3LYP/6-31G* level of theory.
1
(b) H NMR of 2 (1.0 × 10-2 M) in DMSO-d6 before and after the addition of 1.85 equiv of TBAF. For details see Supporting Information.
1
for the H NMR titration of 2 with chloride (see Supporting
Information). Here too, the concerted shifts of the three
noninverted pyrrole NHs together with the â-pyrrole C-H
resonance confirm the presence of a C-H-anion interaction
contributing to the receptor-fluoride complex.
bathochromic shifts, we have observed a hypsochromic shift in
absorption spectra of TCE-2C in the presence of 1 equiv of
fluoride. We assume that this is because the deprotonation of
the imine moiety results in the simultaneous opening of the
pyrrolizine moiety to form TCE-2 (Figure 3).59
In the absence of the X-ray structure confirming the proposed
binding mode, we performed an extensive DFT study aimed at
interpretation of shifts in NMR spectra corresponding to the
energy-minimized model for NC-OMCP in the resting state and
in the complex with anions. Here we show the DFT-GIAO
magnetic shields calculated for 2 and 2‚TMAF in the energy
minimum (a conelike conformation involving the â-pyrrole
hydrogen). The calculation shows the same tendency (shifts of
the 1H-resonances) upon complexation of 2 with TMAF
observed in the 1H NMR titration experiments for 2 with TBAF
in DMSO-d6 (Figure 11).
Deprotonation of Azo-2 by fluoride was also proposed by
Dehaen, based on the similarity in shapes of absorption spectra
of Azo-2 in the presence of fluoride and hydroxide.54 1H NMR
titrations of Azo-2 with fluoride showed that the azo-substituted
pyrrole NH signal as well as the three noninverted pyrrole NH
signals broadened and disappeared (see Supporting Information).
While the disappearance of the azo-substituted pyrrole NH and
the bathochromic shift in the absorption spectrum suggest
deprotonation, the disappearance of the three regular pyrrole
NHs indicates rapid interconversion of multiple conformations
compared to NMR time scale. The course of 1H NMR titration
of Azo-1 and Azo-2 with fluoride is included in the Supporting
Information.
Among the most important issues in anion sensing is the fact
that anions occur mostly as solutes in water or aqueous media.
While a number of anion sensors that are soluble in water-
miscible solvents such as DMSO or MeCN tolerate a small
percentage of water in the medium, the fact remains that most
simple receptors and sensors that utilize hydrogen bonding to
establish the complex do not operate successfully in water.65
We have previously demonstrated that some polyurethanes
are capable of extracting anions from water and aqueous
buffers.42,66 Such materials doped with anion sensors utilizing
hydrogen bonding to establish receptor-anion complexes may
then be successfully used in conjunction with polymer matrices
thereby yielding anion sensors compatible with pure water.42,66
Based on this premise, we have developed a sensing assay using
80 (8 × 10) wells (230 ( 10 µm Ø, 230 ( 10 µm deep, Figure
13) microwell plates. Polyurethane doped with sensors were
solution-cast to create ca. a 50-100 µm coating in each
respective hole. Figure 13 shows a 4 × 8 section of the assay
Specifically, all three NH resonances from the three regular
pyrroles (∆0) and the â-hydrogen of the inverted pyrrole (pink
dot) shift downfield, while the â-hydrogens of the three regular
pyrroles (black dot) and the NH resonance from the inverted
pyrrole (green) show an upfield shift. These shifts in resonances
were calculated from the structure corresponding to energy
minima shown in Figure 10 and are in excellent agreement with
the data recorded for N-confused calixpyrrole and its fluoride
complex. We feel that these studies lend additional credibility
to the binding mode anion proposed for N-confused calix[4]-
pyrroles.
1
Interestingly, the H NMR titrations of TCE-2C and Azo-2
with fluoride showed a slightly different behavior of chemical
shift changes from that of 2 with fluoride. Upon the addition
of fluoride to a TCE-1 solution, four NH proton singlet signals
of a free TCE-1 disappeared and four new NH proton doublet
signals appeared corresponding to the formation of a complex
(Figure 12a).42 On the other hand, titrations of TCE-2C with
fluoride showed that the imine proton signal disappeared and
three pyrrole NH signals in noninverted pyrrole rings did not
show appreciable changes in chemical shifts (Figure 12b). We
believe that this is due to deprotonation of the imine proton by
a fluoride anion. The high acidity of the imine hydrogen is
indirectly confirmed by its downfield shift (11.16 ppm in
DMSO-d6). While the deprotonation usually results in significant
(64) Frisch, M. J. et. al. Gaussian 03, revision B.04; Gaussian, Inc.: Pittsburgh,
PA, 2003.
(65) Steed, J. W.; Atwood, J. L. Supramolecular Chemistry; Wiley: Chichester,
2000.
(66) Aldakov, D.; Palacios, M. A.; Anzenbacher, P., Jr. Chem. Mater. 2005,
17, 5238-5241.
9
11502 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006

Calix[4]pyrrole Isomers
A R T I C L E S
Figure 12. 1H NMR titrations (selected regions) of 1.0 × 10-2 M solutions of (a) TCE-1 and (b) TCE-2C with fluoride in DMSO-d6. The inverted pyrrole
1H resonances in TCE-2C are labeled as follows: imine NHi (red) and â-CHi (black).
Figure 13. Microassay utilizing Azo-1, Azo-2, TCE-1, and TCE-2C doped
polyurethane. The respective anions were administered in pure water (200
nL, 10 mM).
that accommodates sensors Azo-1, Azo-2, TCE-1, and TCE-
2C (horizontal lines) and blank (water), and six different
Figure 14. Competitive microassay utilizing Azo-1, Azo-2, TCE-1, and
TCE-2C doped polyurethane. Assay A: Sensing of acetate (0-20 mM)
administered as an aqueous solution containing 1 mM H2PO4-. Assay B:
Sensing of acetate (0-20 mM) administered as an aqueous solution
containing 1 mM Cl-.
aqueous anionic analytes (fluoride, chloride, dihydrogenphos-
phate, hydrogenpyrophosphate, acetate, and benzoate) in the
vertical lines. The anions used in this assay were added as
aqueous solutions (200 nL, 10 mM). Interestingly, changes in
color observed in the DMSO-water solutions (Table 1)
correspond in most cases to the color changes observed in the
microassay where purely aqueous anion solutions were used.
While the changes in color in the assay may be clearly observed
by a naked eye, one can use a variety of absorption spectroscopy
and imaging techniques to record the sensory response generated
by the array. Also, the techniques of evaluating the response in
the nonspecific sensor arrays67 utilizing pattern recognition67,68
may provide insight into an analysis of anionic analytes. The
color changes in the array utilizing calixpyrrole-based sensors
Azo-1, Azo-2, TCE-1, and TCE-2C suggest that these simple
materials may be potentially used for sensing of anions
administered as purely aqueous solutions.
colorimetric sensors for aqueous anions, for practical purposes
we were interested in the possibility of sensing the anions in
the presence of competing electrolytes. The selectivity coef-
ficients (see Table 1) suggest that it should be possible to sense
carboxylates in the presence of chloride or phosphate. However,
because these selectivity coefficients were derived from as-
sociation constants calculated from titration experiments using
only one anion, we performed the microarray experiments aimed
at sensing of acetate in the presence of dihydrogenphosphate
(H2P) (Figure 14, assay A) and chloride (Cl) (Figure 14, assay
B), respectively. Both assays confirmed that, indeed, acetate
may be sensed in water and in the presence of competing anionic
substrates. From Figure 14 it is clear that changes in color start
to be observable at an equimolar concentration of the target
analyte/competitor (assay A: column 2 vs column 1) and (assay
B: column 7 vs column 6). Further increase in the analyte (AcO)
compared to the competitor concentration (assay A: column
2-5), (assay B: column 7-10) shows a gradual increase in
the color change confirming that the change in color is due to
the presence of the acetate analyte. This result is also significant
While the previous microassay experiment indicated that
materials Azo-1, Azo-2, TCE-1, and TCE-2C can act as
(67) (a) Schena, M. Microarray Analysis; Willey-Liss: Hoboken, NJ, 2003. (b)
Vlasov, Y.; Legin, A.; Rudnitskaya, A.; Di Natale, C.; D’Amico, A. Pure
Appl. Chem. 2005, 77, 1965-1983.
(68) Lavigne, J. J.; Anslyn, E. V. Angew. Chem., Int. Ed. 2001, 40, 3118-
3130.
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11503

A R T I C L E S
Nishiyabu et al.
3-
-
for potential application of anion sensing in biological media,
where chloride and/or phosphate are often present. For example,
human blood plasma contains 2 mM phosphate (HPO42-) and
6 mM carboxylic acids/carboxylates.69
order F- > HP2O7 > AcO- > H2PO4 > Cl-, while the
materials based on NC-OMCP show the following order of
affinities: AcO- > F- > HP2O7 > H2PO4 > Cl-. This
change in the affinity order is understood in terms of symmetry
of the receptor and the receptor-anion complex, while the
difference in color responses between OMCP and NC-OMCP
derivatives is attributed to different distributions and energies
of the frontier molecular orbitals responsible for the intramo-
lecular partial charge transfer between the resting state and the
anion complex. The different affinities and color responses
between OMCP- and NC-OMCP-based sensors can be har-
nessed in anion sensing applications. Model colorimetric mi-
croassays were presented to sense selected anions administered
in the form of their purely aqueous solutions while providing a
clearly distinguishable change in color as a signaling output.
Preliminary competitive assays suggest that these microassays
may potentially be useful as anion sensors for carboxylates.
Activities toward application of the above sensors for carbox-
ylate sensing in biological milieu are currently under way in
our laboratory.
3-
-
Conclusion
Six years after being first described,52 N-confused octamethyl
calix[4]pyrrole was prepared on a multigram scale and unam-
biguously characterized by X-ray crystallography. As expected,
the inverted pyrrole of NC-OMCP reacts preferentially at the
unsubstituted R-position in electrophilic aromatic substitutions.
Most importantly, the anion binding studies using N-confused
calix[4]pyrroles revealed different anion selectivity compared
to calix[4]pyrrole. Detailed NMR studies indicated that this
rather unusual feature stems from a different binding mode, in
which the anion is bound via three regular pyrrole NHs and
appears to be in close contact with the â-CH of the inverted
pyrrole in a conelike conformation. The materials utilizing the
N-confused calix[4]pyrrole receptor exhibit a different binding
preference compared to the parent octamethylcalix[4]pyrrole
receptors. Thus a small difference in structure induces significant
changes in supramolecular behavior. Finally, two pairs of
chromogenic calix[4]pyrrole isomers were synthesized and
characterized including X-ray crystallography. The anion bind-
ing assays revealed an interesting response of these potential
anion sensors to various anions and was shown to follow the
binding behavior observed for the respective parent receptors.
Thus, regular OMCP-based materials show the anion affinity
Acknowledgment. P.A. gratefully acknowledges support from
the Alfred P. Sloan Foundation, BGSU (Technology Innovations
Enhancement grant), Kraft Foods, Inc., and the NSF (NER No.
0304320, SENSOR No. 0330267)
Supporting Information Available: X-ray crystallographic
data (CIF files), 1H NMR and absorption spectroscopic titration
data for anion binding study (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
(69) Schmidt, R. F.; Thews, G. Human Physiology, 2nd ed.; Springer-Verlag:
Berlin, 1989.
JA0622150
9
11504 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006