Boronic acid-based bipyridinium salts as tunable receptors for monosaccharides and α-hydroxycarboxylates

Article abstract of DOI:10.1021/ja066567i

Several novel diboronic acid-substituted bipyridinium salts were prepared and, using a fluorescent reporter dye, were tested for their ability to selectively bind various monosaccharides and α-hydroxycarboxylates in an aqueous medium. The fluorescence sensing mechanism relies on the formation of a ground-state charge-transfer complex between the dye and bipyridinium. An X-ray crystal structure of this complex is described herein. Glucose selectivity over fructose and galactose was achieved by designing the bipyridinium-based receptors to be capable of attaining a 1:1 receptor/substrate stoichiometry via cooperative diboronic acid binding.

Full text of DOI:10.1021/ja066567i

Published on Web 01/17/2007
Boronic Acid-Based Bipyridinium Salts as Tunable Receptors
for Monosaccharides and r-Hydroxycarboxylates
Soya Gamsey, Aaron Miller, Marilyn M. Olmstead,† Christine M. Beavers,†
Lacie C. Hirayama, Sulolit Pradhan, Ritchie A. Wessling, and Bakthan Singaram*
Department of Chemistry and Biochemistry, UniVersity of California, Santa Cruz, California
95064, and Department of Chemistry, UniVersity of California, DaVis, California 95616
Received September 11, 2006; E-mail: singaram@chemistry.ucsc.edu
Abstract: Several novel diboronic acid-substituted bipyridinium salts were prepared and, using a fluorescent
reporter dye, were tested for their ability to selectively bind various monosaccharides and R-hydroxycar-
boxylates in an aqueous medium. The fluorescence sensing mechanism relies on the formation of a ground-
state charge-transfer complex between the dye and bipyridinium. An X-ray crystal structure of this complex
is described herein. Glucose selectivity over fructose and galactose was achieved by designing the
bipyridinium-based receptors to be capable of attaining a 1:1 receptor/substrate stoichiometry via cooperative
diboronic acid binding.
Boronic acids are known to reversibly bind diols with high
affinity to form cyclic esters.1 As a result, many boronic acid-
containing molecules have been utilized as chemosensors for
the recognition of carbohydrates.2 Specifically, there has been
much interest focused on the design of boronic acid-containing
fluorescent glucose sensors that operate under physiological
conditions.3 If optimized, such sensors can be implantable, and
used to continuously monitor glucose concentrations in preterm
infants,4 patients in the Intensive Care Unit,5 and in people
suffering from diabetes.6
In order to develop a successful in vivo boronic acid-based
glucose sensor, an important criterion that must be met is the
preferential binding of glucose over other physiologically
significant monosaccharides such as fructose. This is a chal-
lenging objective since phenylboronic acids display a higher
binding affinity for fructose.7 To increase glucose selectivity,
† Department of Chemistry, University of California at Davis.
(1) (a) Lorand, J. P.; Edwards, J. O. J. Org. Chem. 1959, 24, 769-774. (b)
Kuivila, H. G.; Keough, A. H.; Soboczenski, E. J. J. Org. Chem. 1954, 19,
780-783. (c) Sugihara, J. M.; Bowman, C. M. J. Am. Chem. Soc. 1958,
80, 2443-2446.
Figure 1. Portion of the structure of (4,4′-m-BBV2+)3(HPTS3-)2. The
(2) For reviews see: (a) Wang, W.; Gao, X.; Wang, B. Curr. Org. Chem.
BBV2+ molecule containing N1 and N2 has no crystallographic symmetry,
2002, 6, 1285-1317. (b) Striegler, S. Curr. Org. Chem. 2003, 7, 81-102.
whereas the top molecule, containing N3, has crystallographic inversion
symmetry. The HPTS3- species lies between these two cations with the
sulfonate groups staggered with respect to the basic π-stacking motif.
(c) James, T. D.; Shinkai, S. Top. Curr. Chem. 2002, 218, 159-200.
(3) (a) Czarnik, A. W. Fluorescent Chemosensors for Ion and Molecule
Recognition; American Chemical Society: Washington, DC, 1993; Vol.
538. (b) Fang, H.; Kaur, G.; Wang, B. J. Fluoresc. 2004, 14, 481-489.
(c) Cao, H.; Heagy, M. D. J. Fluoresc. 2004, 14, 569-584. (d) Pickup,
J. C.; Hussain, F.; Evans, N. D.; Rolinski, O. J.; Birch, D. J. S. Biosens.
Bioelectron. 2005, 20, 2555-2565.
the concept of bidentate binding via diboronic acids has been
utilized.8 In this approach, the sensor molecule contains two
arylboronic acid moieties spatially disposed in a way that allows
for cooperative binding to a single glucose molecule. For a given
(4) Cornblath, M.; Reisner, S. H. N. Engl. J. Med. 1965, 273, 378-381.
(5) (a) Van den Berghe, G. Int. J. Obes. 2002, 26, S3-S8. (b) Van den Berghe,
G.; Wilmer, A.; Hermans, G.; Meersseman, W.; Wouters, P. J.; Milants,
I.; Van Wijngaerden, E.; Bobbaers, H.; Bouillon, R. N. Engl. J. Med. 2006,
354, 449-461. (c) Finney, S. J.; Zekveld, C.; Elia, A.; Evans, T. W.
J. Am. Med. Assoc. 2003, 290, 2041-2047.
(8) (a) Tsukagoshi, K.; Shinkai, S. J. Org. Chem. 1991, 56, 4089-4091. (b)
Shiomi, Y.; Saisho, M.; Tsukagoshi, K.; Shinkai, S. J. Chem. Soc., Perkin
Trans. 1 1993, 2111-2117. (c) James, T. D.; Sandanayake, K. R. A. S.;
Shinkai, S. Angew. Chem., Int. Ed. Engl. 1994, 106, 2287-2289. (d) Eggert,
H.; Frederiksen, J.; Morin, C.; Norrild, J. C. J. Am. Chem. Soc. 1999, 64,
3846-3852. (e) Yang, W.; He, H.; Drueckhammer, D. G. Angew. Chem.,
Int. Ed. 2001, 40, 1714-1718. Arimori, S.; Ward, C. J.; James, T. D.
Tetrahedron Lett. 2002, 43, 303-305.
(6) (a) Wentholt, I. M.; Vollebregt, M. A.; Hart, A. A.; Hoekstra, J. B.; DeVries,
J. H. Diabetes Care 2005, 28, 2871-2876. (b) Wilson, Darrell, M.; Block,
J. Diabetes Technol. Ther. 2005, 7, 788-791. (c) Heinemann, L.;
Schmelzeisen-Redeker, G. Diabetologia 1998, 41, 848-854. (d) Kerr, D.
Int. J. Clin. Pract. 2001, 43-46. (e) Garg, S. K.; Hoff, H. K.; Chase,
H. P. Endocrinol. Metab. Clin. North Am. 2004, 33, 163-173.
(7) Springsteen, G.; Wang, B. Tetrahedron 2002, 58, 5291-5300.
9
1278
J. AM. CHEM. SOC. 2007, 129, 1278-1286
10.1021/ja066567i CCC: $37.00 © 2007 American Chemical Society

Boronic Acid-Based Glucose Sensors
A R T I C L E S
Scheme 1. Proposed Glucose-Sensing Mechanism
Scheme 2. Boronic Acid-Based Bipyridinium Salts for Analyte Recognitiona
a Reagents and conditions: (i) Pd(OAc)2, PPh3, Na2CO3, p-dioxane, 95 °C, 4 h, 64% (2), 70% (3); (ii) 2- or 3-bromomethylphenylboronic acid (2.5
equiv), DMF, 70 °C, 48 h, 80% (3,3′-o-), 76% (3,3′-m-), 77% (3,4′-o-), 76% (3,4′-m-).
diboronic acid-containing scaffold, the spacing between the
boronic acids could be adjusted through synthetic modifications
in order to create a glucose-specific binding pocket. However,
in many systems, the boronic acids are covalently linked to a
reporter (a fluorophore or chromophore),2 and such modifica-
tions to the receptor can prove difficult or impossible to achieve
without altering the photophysical properties of the reporter as
well.
enhanced glucose selectivity and sensitivity. The different
receptors were studied in terms of analyte selectivity (monosac-
charides and R-hydroxycarboxylates), quenching strengths, and
reduction potentials.
Results and Discussion
The sensing ensemble is composed of the anionic fluorescent
dye, 8-hydroxypyrene-1,3,6-trisulfonic acid trisodium salt (HPTS,
This problem can be avoided by employing a modular
approach to sensor design, where the receptor and the reporter
moieties exist as covalently discrete entities. To this end, our
laboratory has developed a two-component sensing system
comprising a fluorescent dye that serves as the reporter unit
and a diboronic acid-containing molecule that acts dually as a
fluorescence quencher and a saccharide receptor.9 We report
herein that tuning the receptor unit by synthetically manipulating
the spacing between the diboronic acids provides greatly
(9) (a) Camara, J. N.; Suri, J. T.; Cappuccio, F. E.; Wessling, R. A.; Singaram,
B. Tetrahedron Lett. 2002, 43, 1139-1141. (b) Suri, J. T.; Cordes, D. B.;
Cappuccio, F. E.; Wessling, R. A.; Singaram, B. Langmuir 2003, 19, 5145-
5152. (c) Suri, J. T.; Cordes, D. B.; Cappuccio, F. E.; Wessling, R. A.;
Singaram, B. Angew. Chem., Int. Ed. 2003, 42, 5857-5859. (d) Cappuccio,
F. E.; Suri, J. T.; Cordes, D. B.; Wessling, R. A.; Singaram, B. J. Fluoresc.
2004, 14, 521-533. (e) Cordes, D. B.; Gamsey, S.; Sharrett, Z.; Miller,
A.; Thoniyot, P.; Wessling, R. A.; Singaram, B. Langmuir 2005, 21, 6540-
6547. (f) Cordes, D. B.; Miller, A.; Gamsey, S.; Sharrett, Z.; Thoniyot, P.;
Wessling, R.; Singaram, B. Org. Biomol. Chem. 2005, 3, 1708-1713. (g)
Gamsey, S.; Baxter, N. A.; Sharrett, Z.; Cordes, D. B.; Olmstead, M. M.;
Wessling, R. A.; Singaram, B. Tetrahedron 2006, 62, 6321-6331.
9
J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1279

A R T I C L E S
Gamsey et al.
Figure 2. View of the unit cell contents of (BBV2+)3(HPTS3-)2 showing the donor-acceptor π-stacking arrangement of the molecules edge-on. Solvate
methanol and water molecules are omitted for clarity.
1), and a boronic acid-appended viologen,10 such as 4,4′-m-
BBV (Scheme 1). It is postulated that Coulombic attraction
between 1 and 4,4′-m-BBV results in the formation of a ground-
state complex which facilitates electron transfer from the dye
to the viologen, leading to a decrease in the fluorescence
intensity of the dye. When glucose is added to the system, the
boronic acids are converted to tetrahedral anionic glucoboronate
esters, which effectively neutralize the cationic viologen, thus
greatly diminishing its quenching efficiency, and an increase
in fluorescence intensity is observed. To help substantiate the
putative glucose-induced dissociation mechanism of the dye and
receptor, we have used 11B NMR to monitor the change in
charge of boron from neutral (sp2) to anionic (sp3) upon addition
of glucose at pH 7.4.9g
been reported. Due to their electron-deficient nature, viologens
function as acceptors in many charge-transfer (CT) complexes.14
Baptista and co-workers maintain that HPTS and methyl
viologen form a ground-state complex on the basis of UV-
vis, fluorescence, and laser flash photolysis studies.11a Regarding
the initial quenching mechanism in our system, we have
proposed, on the basis of UV-vis absorbance studies, that 1
and 4,4′-m-BBV form a distinct complex.9e,g From single-crystal
X-ray analysis, we now report the formation of a CT complex
in the solid state (Figures 1 and 2).
The asymmetric unit contains one 8-hydroxypyrene-1,3,6-
trisulphonate (HPTS3-) molecule, 1.5 BBV2+ molecules, one
methanol, and one water molecule. The overall stoichiometry
is 2(HPTS3-):3(BBV2+). The BBV2+ molecule that is entirely
contained within the asymmetric unit has one boronic acid (B2)
that has undergone an esterification with methanol (bottom
structure, Figure 1). The dihedral angle between the two pyridine
rings of the viologen is 10.8°. The second half of the other
Several instances of viologens serving as fluorescence
quenchers for organic dyes,11 porphyrins,12 and polymers13 have
(10) In general, the term viologen refers to any diquaternized bipyridyl salt.
(11) (a) De Borba, E. B.; Amaral, C. L. C.; Politi, M. J.; Villalobos, R.; Baptista,
M. S. Langmuir 2000, 16, 5900-5907. (b) Nakashima, K.; Kido, N.
Photochem. Photobiol. 1996, 64, 296-302. (c) Sun, L.-C.; Yang, Y.-F.;
He, J.-J.; Shen, T. Dyes Pigm. 1995, 28, 275-279. (d) Takenaka, S.;
Shigemoto, N.; Kondo, H. Supramol. Chem. 1998, 9, 47-56.
(12) (a) Blondeel, G.; De Keukeleire, D.; Harriman, A.; Milgrom, L. R. Chem.
Phys. Lett. 1985, 118, 77-82. (b) Harriman, A. Inorg. Chim. Acta 1984,
88, 213-216. (c) Kusumoto, Y.; Watanabe, J.; Kurawaki, J.; Satake, I.
Chem. Lett. 1987, 1417-1420. (d) Kusumoto, Y.; Uchikoba, M. Chem.
Lett. 1991, 1985-1988. (e) Yanuck, M. D.; Schmehl, R. H. Chem. Phys.
Lett. 1985, 122, 133-138.
(13) (a) Chen, L.; McBranch, D. W.; Wang, H.-L.; Helgeson, R.; Wudl, F.;
Whitten, D. G. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12287-12292. (b)
Gaylord, B. S.; Wang, S.; Heeger, A. J.; Bazan, G. C. J. Am. Chem. Soc.
2001, 123, 6417-6418. (c) Wang, D.; Wang, J.; Moses, D.; Bazan, G. C.;
Heeger, A. J.; Park, J.-H.; Park, Y.-W. Synth. Met. 2001, 119, 587-588.
(d) Wang, D.; Wang, J.; Moses, D.; Bazan, G. C.; Heeger, A. J. Langmuir
2001, 17, 1262-1266. (e) Fan, C.; Hirasa, T.; Plaxco, K. W.; Heeger, A.
J. Langmuir 2003, 19, 3554-3556. (f) Miyashita, T.; Ohsawa, M.; Matsuda,
M. Macromolecules 1986, 19, 585-588.
(14) (a) Pia, E.; Toba, R.; Chas, M.; Peinador, C.; Quintela, J. M. Tetrahedron
Lett. 2006, 47, 1953-1956. (b) Monk, P. M. S.; Hodgkinson, N. M.;
Partridge, R. D. Dyes Pigm. 1999, 43, 241-251. (c) Matos, M. S.; Gehlen,
M. H. Spectrochim. Acta, Part A 1998, 54A, 1857-1867. (d) Monk,
P. M. S.; Hodgkinson, N. M. Electrochim. Acta 1997, 43, 245-255. (e)
Tsukahara, K.; Kaneko, J.; Miyaji, T.; Abe, K. Tetrahedron Lett. 1996,
37, 3149-3152. (f) Yoon, K. B. Chem. ReV. 1993, 93, 321-339. (g) Alvaro,
M.; Ferrer, B.; Fornes, V.; Garcia, H. Chem. Phys. Chem. 2003, 4, 612-
617.
9
1280 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Boronic Acid-Based Glucose Sensors
A R T I C L E S
Figure 3. Stern-Volmer plot of HPTS (4 × 10-6 M in pH 7.4 phosphate
buffer, λex ) 460 nm and λem ) 510 nm) with increasing concentrations of
BBVs.
Table 1. Static (Ks) and Dynamic (V) Quenching Constants for
BBVs with HPTS (4 × 10-6 M in pH 7.4 Phosphate Buffer, λex
460 nm and λem ) 510 nm)
)
Figure 4. Benzyl viologens used in cyclic voltammetry.
Table 2. Static (Ks) and Dynamic (V) Quenching Constants for
Benzyl Viologens with HPTS (4 × 10-6 M in pH 7.4 Phosphate
Buffer, λex ) 460 nm and λem ) 510 nm) and Formal Potentials
(Eo′) Measured by CV in pH 7.4 Phosphate Buffer (Scan Rate )
100 mV/s)
1
1
1
1
BBV
K
s (M-
)
V (M-
)
BBV
K
s (M-
)
V (M-
)
4,4′-o- 8900 ( 200 2900 ( 100
4,4′-m- 8100 ( 100 2950 ( 100
3,4′-o- 6600 ( 200 1650 ( 50
3,4′-m- 7500 ( 200 1700 ( 100
3,3′-o- 4300 ( 200 140 ( 30
3,3′-m- 5000 ( 100 700 ( 50
1
1
K
s (M-
)
V (M-
)
Eo
′
(mV vs NHE)
4,4′-BV
3,4′-BV
3,3′-BV
15000 ( 1000
9500 ( 500
7400 ( 300
2300 ( 200
2800 ( 300
2200 ( 200
-420
-650
BBV2+ molecule is generated by a center of inversion, giving
two coplanar pyridine rings.
-1100
The BBV2+ and HPTS3- molecules are arranged in alternating
layers that are almost parallel; however, the HPTS3- molecule,
with three deprotonated sufonates is too sterically hindered for
a perpendicular stacking arrangement (Figure 2). The molecules
adopt a slipped stacking motif that permits the distance between
the layers to range between 3.25 and 3.46 Å, which is typical
for a donor-acceptor CT complex of this type.15 X-ray
structures of CT complexes involving similar viologen acceptors
have been reported.16
3 shows a Stern-Volmer plot generated from the fluorescence
quenching of HPTS by the receptors. The calculated Stern-
Volmer constants, Ks and V, indicating the degree of static and
dynamic quenching, respectively, are summarized in Table 1.
The quenching efficiencies of the viologens were found to be
of the order: 4,4′-BBVs > 3,4′-BBVs > 3,3′-BBVs. To help
verify our suspicion that these observed differences in quenching
behavior are due to the electrochemical properties of the
bipyridimium core, cyclic voltammetry studies were carried out
on model compounds (benzyl viologens).
We have previously shown that Stern-Volmer quenching
constants correlate with association constants, determined by
Benesi-Hildebrandt treatment of UV-vis absorbance data, and
therefore provide a measure of the strength of the dye-receptor
complex.9b,e The degree of quenching is also an indication of
the electron-accepting abilities of the viologens and should
therefore correlate with their reduction potentials.17 To confirm
this relationship, the reduction potentials of benzyl viologens
(BV), 4,4′-BV, 3,4′-BV, and 3,3′-BV (Figure 4), were deter-
mined by cyclic voltammetry (CV) and were compared to their
quenching constants. The Stern-Volmer values for the BVs
(Table 2) are higher than those of the boronic-acid containing
derivatives,18 but the quenching trend (4,4′ > 3,4′ > 3,3′) within
each series is the same. The measured formal potentials (Eo′)
of the BVs correspond with their Stern-Volmer values, in that
the Eo′ values become increasingly more positive with increasing
To optimize the system in terms of glucose sensitivity and
selectivity, six bipyridinium-based receptors, shown in Scheme
2, were synthesized. The 4,4′-bipyridyl-based receptors (4,4′-
o- and 4,4′-m-BBV) have been previously studied and shown
to display monosaccharide selectivity of the order fructose .
galactose > glucose.9g The relative ratios of the binding
constants were parallel to those reported for monoboronic acids
which is expected since, for both structures, the intramolecular
distance between the two boronic acids is too great to allow
for cooperative binding with one glucose molecule. The four
other receptors (3,4′-o-, 3,4′-m-, 3,3′-o-, and 3,3′-m-BBV),
however, were designed to have shorter intramolecular B-B
distances, and therefore each theoretically possesses the ability
to form a 1:1 binding scenario with glucose.
Before binding studies were carried out, the quenching
characteristics for each of the receptors were established. Figure
(15) Yoon, K. B.; Kochi, J. K. J. Am. Chem. Soc. 1989, 111, 1128-1130.
(16) (a) Kidowaki, M.; Tamaoki, N. Chem. Commun. 2003, 290-291. (b)
Yoshikawa, H.; Nishikiori, S.-I. J. Chem. Soc., Dalton Trans. 2005, 3056-
3064. (c) Nishikiori, S.-I.; Yoshikawa, H.; Sano, Y.; Iwamoto, T. Acc.
Chem. Res. 2005, 38, 227-234. (d) Yoshikawa, H.; Nishikiori, S.-I.; Ishida,
T. J. Phys. Chem. B 2003, 107, 9261-9267. (e) Willner, I.; Eichen, Y.;
Rabinovitz, M.; Hoffman, R.; Cohen, S. J. Am. Chem. Soc. 1992, 114,
637-644. (f) Sundaresan, T.; Wallwork, S. C. Acta Crystallogr., Sect. B
1972, 28, 2474-2480. (g) Robinson, P. D.; Smith, K. R.; Vermeulen,
L. A. Acta Crystallogr., Sect. C 2002, C58, o632-o634.
(17) (a) Amouyal, E.; Zidler, B.; Keller, P.; Moradpour, A. Chem. Phys. Lett.
1980, 74, 314-317. (b) Keller, P.; Moradpour, A.; Amouyal, E.; Zidler,
B. J. Mol. Catal. 1981, 12, 261-263.
(18) It is proposed that the addition of boronic acids alters the electronics as
well as the sterics of the viologens, thereby affecting their static and dynamic
quenching constants. See refs 9b,g.
9
J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1281

A R T I C L E S
Gamsey et al.
Table 3. Apparent Monosaccharide Binding Constants (Kb)
Determined for the BBVs (5 × 10-4 M) with HPTS (4 × 10-6 M in
Phosphate Buffer, λex ) 460 nm and λem ) 510 nm) and Their
Monosaccharide Selectivity Ratios
1
K
b (M-
)
selectivity ratios
glucose
galactose
fructose
glu/fru
glu/gal
4,4′-o-
4,4′-m-
3,4′-o-
3,4′-m-
3,3′-o-
3,3′-m-
27 ( 5
14 ( 5
37 ( 5
25 ( 5
800 ( 100
300 ( 50
0.03
0.05
0.33
0.16
1.7
0.73
0.56
3.7
1.9
11
260 ( 50
65 ( 10
70 ( 10
35 ( 5
800 ( 100
400 ( 50
1900 ( 200
53 ( 10
180 ( 50
45 ( 5
1100 ( 200
500 ( 50
0.11
1.2
quenching strength (Table 2). This indicates that the quenching
efficiencies (measured by Ks and V) of the viologens are
dependent upon their electron-accepting abilities (measured by
Eo′). Preliminary CV studies indicate that the boronic acid-
containing viologens (BBVs) display a similar trend (data not
shown). A detailed investigation of the electrochemical behavior
of the BBVs in the presence and absence of various saccharides
is currently underway.
The apparent binding affinities of each of the BBVs for
glucose were determined by measuring the degree of fluores-
cence enhancement obtained when increasing amounts of
glucose (0-35 mM) were added to an aqueous solution (pH
7.4 phosphate buffer, 0.1 M ionic strength) of 1 (4 × 10-6 M)
and BBV (5 × 10-4 M).19 The calculated binding constants
(Kb) show that 3,3′-o-BBV has a binding affinity for glucose
several orders of magnitude higher than the other BBVs (Table
3). Further inspection of the binding constants reveals a general
trend, in which the ortho-substituted bipyridyls have higher
affinities than their corresponding meta-derivatives. Most likely,
this is due to an intramolecular B f N charge interaction,
previously described,9g which is unique to ortho-substituted
benzylboronic acid viologens, and which stabilizes boronate
ester formation, thus resulting in higher binding constants.
The differences in the fluorescence emission of HPTS with
4,4′-o-BBV (which cannot cooperatively bind glucose) versus
that with 3,3′-o-BBV upon glucose addition are shown in Figure
5. These two receptors share in common the o-boronic acid
positioning but differ in their quenching strengths. To balance
out the quenching differences, a 5 × 10-4 M concentration of
3,3′-o-BBV was used to quench the fluorescence emission of
HPTS (Figure 5a), while a lower concentration of 4,4′-o-BBV
(2 × 10-4 M) was used to achieve the same degree of quenching
(Figure 5b). When each of these dye/quencher solutions was
titrated with glucose, very different sensing profiles were
obtained. The binding curve for 3,3′-o-BBV displays an early
saturation and a slight sigmoidal shape,20 whereas 4,4′-o-BBV
gives a nearly linear response within the same glucose concen-
tration range (Figure 5, insets).
Figure 5. Change in fluorescence emission spectrum of HPTS (4 × 10-6
M in pH 7.4 phosphate buffer, λex ) 460 nm and λem ) 510 nm) in the
presence of (a) 3,3′-o-BBV (5 × 10-4 M), and (b) 4,4′-o-BBV (2 × 10-4
M) upon addition of glucose. Dotted line ) dye only, red line ) dye and
quencher, black lines ) dye, quencher, and glucose. Insets are the binding
isotherms: relative fluorescence increase (F/F0) of HPTS in the presence
of quencher plotted as a function of glucose concentration.
m-BBV) exhibit saccharide selectivity of the order: fructose
. galactose > glucose, where the fructose/glucose binding
constant ratio is about 25:1 and the galactose/glucose ratio is
about 1.5:1. As mentioned earlier, these ratios are similar to
those obtained for monoboronic acids. The glucose selectivity
of 3,3′-m-BBV and the 3,4′-bipyridyl-based viologens (3,4′-o-
and 3,4′-m-BBV) were improved relative to that of the 4,4′-
BBVs. These receptors display monosaccharide selectivity of
the order fructose . glucose > galactose. The receptor 3,3′-
m-BBV showed a slight preference for glucose over galactose
(1.2:1), and a 10:1 preference for fructose over glucose. The
selectivity ratios for 3,4′-m-BBV were a bit better, giving a 2:1
preference for glucose over galactose and a 6:1 preference for
fructose over glucose. The receptor 3,4′-o-BBV was better still,
displaying a 4:1 preference for glucose over galactose, and only
a 3:1 preference for fructose over glucose.
Remarkably, 3,3′-o-BBV displayed an enhanced selectivity
for glucose over fructose and galactose. For this receptor, the
apparent binding constant for glucose is almost 2-fold higher
than that of fructose, and 11-fold higher than that of galactose.
It is important to note that, although the apparent binding
constant of 3,3′-o-BBV was found to be higher for glucose than
for fructose, the magnitude of fluorescence enhancement
obtained upon addition of monosaccharide was greater for
fructose.21
To test for monsaccharide selectivity, the affinities of each
of the BBVs for glucose, galactose, and fructose were compared
(Table 3). The 4,4′-bipyridyl-based viologens (4,4′-o- and 4,4′-
(19) All of the receptors were tested at this concentration. However, since the
receptors have different quenching strengths and are used in molar excess
relative to dye, the amount of the dye-receptor complex formed may vary
for each receptor studied.
(20) (a) It is likely that the sigmoidal binding behavior may occur as a result of
the large excess of [BBV] relative to that of [HPTS]; see: Metzger, A.;
Anslyn, E. V. Angew. Chem., Int. Ed. 1998, 37, 649-652. (b) The [BBV]/
[HPTS] ratio is large (125) for all the receptors studied, but the sigmoidal
binding curve shape is most pronounced for 3,3′-o-BBV.
9
1282 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Boronic Acid-Based Glucose Sensors
A R T I C L E S
Table 4. Apparent Analyte Binding Constants (Kb) Determined for
the BBVs (5 × 10-4 M) with HPTS (4 × 10-6 M in Phosphate
Buffer, λex ) 460 nm and λem ) 510 nm) and Their Tatrate/Malate
Selectivity Ratios.
1
K
b (M-
)
selectivity ratio
tartrate/malate
tartrate
malate
lactate
erythritol
4,4′-o-
4,4′-m-
3,4′-o-
3,4′-m-
3,3′-o-
3,3′-m-
160 ( 50
47 ( 10
230 ( 50
64 ( 10
60 ( 10
30 ( 5
6 ( 2 13 ( 5
5 ( 2 16 ( 5
2.7
1.6
2.3
4.3
7.8
4.2
100 ( 20 15 ( 5 20 ( 5
15 ( 5
4 ( 2 20 ( 5
1250 ( 100 160 ( 50 26 ( 5 38 ( 10
250 ( 50 60 ( 10 18 ( 5 34 ( 5
malate, tartrate, and erythritol (Figure 7). Erythritol, the diol
counterpart of tartrate, was used to directly compare bis(diol)
binding versus bis(R-hydroxycarboxylate) binding. The natural
product tartaric acid is an optically active compound that is
present in wine and other grape-based drinks. Some boronic
acid-based sensors have been used to detect and quantify tartrate
for possible applications in the beverage industry.25 Chiral
diboronic acid-based sensors have also been developed which
are able to discriminate between L- and D-tartrate enantiomers.26
Malic acid is structurally similar to tartaric acid and is also
present in wine. Thus, the ability to distinguish between these
two analytes is both useful and challenging. Lactate is present
in the body, and its levels increase during exercise or when the
body is under stress. When the body is at rest, the levels of
lactic acid are about 1-2 mM. However, these levels can rise
to 20 mM during intense physical exertion.
Figure 6. Energy-minimized structure of R-D-glucofuranose-1,2:3,5-
bisboronate of 3,3′-o-BBV modeled in the gas phase (semiempirical AM1
using Spartan). Carbon atoms are in gray, and oxygen atoms are in white.
The calculated apparent binding constants for the R-hydroxy-
carboxylates are shown in Table 4. Malate and tartrate both
elicited large fluorescence increases (see Supporting Informa-
tion) and moderate to high apparent binding constants with all
the BBVs, whereas erythritol and lactate consistently produced
weak fluorescence modulations and low binding constants.
Comparing each BBV’s selectivity ratio for tartrate and malate,
4,4′-o-, 4,4′-m-, and 3,4′-o-BBV all have tartrate/malate binding
constant ratios of about 2:1, while 3,3′-o-BBV displays a ratio
of 8:1. This suggests that 3,3′-o-BBV may cooperatively bind
tartrate. 3,4′-m- and 3,3′-m-BBV have tartrate/malate binding
constant ratios of about 4:1. It is noteworthy the fact that all of
the receptors displayed very low affinities for lactatesboth in
terms of binding constants and fluorescence signal enhance-
ments. This is a desirable attribute for a clinical glucose sensor
since the concentration of lactate in the blood can sometimes
surpass that of glucose and could therefore interfere with glucose
detection if the sensor was indiscriminate toward these two
analytes.
Figure 7. Structures of R-hydroxycarboxylates and erythritol used in
binding studies.
These fluorescence binding studies demonstrate the relatively
exceptional glucose affinity of 3,3′-o-BBV and suggest that this
may be due to its ability to bind one glucose molecule through
both boronic acid receptors. To verify that 3,3′-o-BBV can
indeed attain this type of binding mode without exorbitant
energy expenditures relative to the unbound receptor, molecular
modeling studies were performed. Although the form of glucose
that predominates in solution is the R-pyranose form, many
studies aimed at structure elucidation of diboronic acid-glucose
complexes reveal that the R-furanose form is likely bound.22
We thus modeled R-glucofuranose bound to 3,3′-o-BBV through
the 1,2- and 3,5-hydroxyl groups (Figure 6).23 A reasonable
energy minimum was obtained for this structure, indicating its
viability to form.
Finally, we studied the binding affinities of the BBVs for
other analytes. In addition to cis-diols, boronic acids are also
known to bind R-hydroxy acids (at pH 7.4, R-hydroxycarboxy-
lates).24 The BBVs were tested for their ability to bind lactate,
Conclusion
The signal transduction mechanism of this sensing system is
based on the interaction between the viologen and HPTS. To
(21) Other glucose sensing systems that display binding constants not propor-
tional to changes in fluorescence intensities include: (a) Cao, H.; Diaz,
D. I.; DiCesare, N.; Lakowicz, J. R.; Heagy, M. D. Org. Lett. 2002, 4,
1503-1505. (b) Gray, C. W., Jr.; Houston, T. A. J. Org. Chem. 2002, 67,
5426-5428. (c) Zhao, J.; Davidson, M. G.; Mahon, M. F.; Kociok-Koehn,
G.; James, T. D. J. Am. Chem. Soc. 2004, 126, 16179-16186.
(22) (a) Bielecki, M.; Eggert, H.; Norrild, J. C. J. Chem. Soc., Perkin Trans. 2,
1999, 449-456. (b) Norrild, J. C.; Eggert, H. J. Am. Chem. Soc. 1995,
117, 1479-1484. (c) Eggert, H.; Frederiksen, J.; Morin, C.; Norrild, J. C.
J. Org. Chem. 1999, 64, 3846-3852.
(25) (a) Lavigne, J. J.; Anslyn, E. V. Angew. Chem., Int. Ed. 1999, 38, 3666-
3669. (b) Wiskur, S. L.; Ait-Haddou, H.; Lavigne, J. J.; Anslyn, E. V.
Acc. Chem. Res. 2001, 34, 963-972. (c) Wiskur, S. L.; Floriano, P. N.;
Anslyn, E. V.; McDevitt, J. T. Angew. Chem., Int. Ed. 2003, 42, 2070-
2072. (d) Gray, C. W., Jr.; Houston, T. A. J. Org. Chem. 2002, 67, 5426-
5428. (e) Manimala, J. C.; Wiskur, S. L.; Ellington, A. D.; Anslyn, E. V.
J. Am. Chem. Soc. 2004, 126, 16515-16519. (f) Wiskur, S. L.; Lavigne,
J. J.; Metzger, A.; Tobey, S. L.; Lynch, V.; Anslyn, E. V. Chem. Eur.
J. 2004, 10, 3792-3804.
(26) (a) Zhao, J.; Davidson, M. G.; Mahon, M. F.; Kociok-Koehn, G.; James,
T. D. J. Am. Chem. Soc. 2004, 126, 16179-16186. (b) Zhao, J.; James,
T. D. Chem. Commun. 2005, 1889-1891. (c) Zhu, L.; Anslyn, E. V.
J. Am. Chem. Soc. 2004, 126, 3676-3677. (d) Zhu, L.; Zhong, Z.; Anslyn,
E. V. J. Am. Chem. Soc. 2005, 127, 4260-4269.
(23) For completeness, R-glucopyranose bound to 3,3′-o-BBV and unbound 3,3′-
o-BBV, were also modeled and found to have similar minimized energies
(see Supporting Information).
(24) (a) Katzin, L. I.; Gulyas, E. J. Am. Chem. Soc. 1966, 88, 5209-5212. (b)
Kustin, K.; Pizer, R. J. Am. Chem. Soc. 1969, 91, 317-322. (c) Friedman,
S.; Pace, B.; Pizer, R. J. Am. Chem. Soc. 1974, 96, 5381-5384. (d)
Friedman, S.; Pizer, R. J. Am. Chem. Soc. 1975, 97, 6059-6062.
9
J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1283

A R T I C L E S
Gamsey et al.
were in accord with those previously reported.28 1H NMR (CDCl3, 500
MHz) δ 7.15 (ddd, J ) 7.75, 4.5, 1.0 Hz, 1H), 7.24 (dd, J ) 4.5, 1.5
Hz, 2H), 7.66 (dt, J ) 8.0, 2.0 Hz, 1H), 8.41 (dd, J ) 5.0, 1.5 Hz,
1H), 8.44 (d, J ) 4.5, 1.5 Hz, 2H), 8.63 (d, J ) 1.5, 1H); 13C NMR
(CDCl3, 125 MHz) δ 121.4, 123.7, 133.5, 134.2, 144.9, 148.0, 150.0,
150.4.
gain more insight into the nature of this interaction, a charge-
transfer complex between 4,4′-m-BBV and HPTS was formed
in the solid state and studied by X-ray crystallography. This
correlates with solution data, where evidence of a charge-transfer
complex has been observed by UV-vis. Further elucidation of
the viologen-HPTS interaction was accomplished via quench-
ing studies. A relationship between Stern-Volmer quenching
constants and the reduction potentials of benzyl viologens was
established, which helped to confirm a donor-acceptor mech-
anism.
We have demonstrated that good analyte selectivity can be
achieved by synthetically altering the viologen-based receptor
unit to allow for cooperative binding of diboronic acids. The
nitrogen-positioning in the bipyridyl rings, as well as the boronic
acid-positioning around the benzyl rings, were modified in order
to create receptors with unique binding environments relative
to one another. One receptor, 3,3′-o-BBV, displayed a very large
binding constant for glucose and was found to be more selective
for glucose over fructose. Other receptors showed selectivity
for glucose over galactose and displayed enhanced glucose/
fructose binding ratios. The six receptors were also screened
against three different R-hydroxycarboxylates in a neutral
aqueous medium. Tartrate and malate both elicited large
fluorescence increases and moderate to high apparent binding
constants for all of the receptors. The receptor 3,3′-o-BBV,
displayed an 8-fold preference for tartrate over malate. Lactate,
a potential interferent for in vivo glucose detection, was
fortunately found to have very low binding constants with all
of the receptors.
N,N′-Bis-(benzyl-2-boronic acid)-[3,4′]bipyridinium Dibromide
(3,4′-o-BBV). To a solution of 2-bromomethylphenylboronic acid (0.3
g, 1.38 mmol) in DMF (15 mL) was added 3,4′-bipyridyl (0.095 g, 0.6
mmol), and the reaction was stirred at 70 °C for 48 h. After the mixture
cooled to room temperature, acetone (50 mL) was added to the clear
yellow solution. The resulting pale-yellow precipitate was collected
by centrifugation, washed with acetone, and dried under a stream of
1
argon (0.27 g, 77% yield). H NMR (D2O, 500 MHz) δ 6.21 (s, 2H),
6.27, (s, 2H), 7.71 (m, 6H), 7.92 (m, 2H), 8.39 (dd, J ) 8.5, 6.5 Hz,
1H), 8.51 (d, J ) 6.5 Hz, 2H), 9.16 (m, 4H), 9.60 (s, 1H); 13C NMR
(D2O, 62.5 MHz) δ 66.0, 66.7, 127.4, 130.0, 130.9, 131.0, 132.3, 132.3,
132.5, 132.7, 136.2, 136.3, 136.3, 136.7, 136.8, 145.3, 146.2, 146.9,
147.5, 150.9; 11B NMR (80 MHz, D2O) δ 28.3. HRMS-ESI m/z calcd
for C24H24B2BrN2O4 [M - Br]+: 505.11056, found 505.11693.
N,N′-Bis-(benzyl-3-boronic acid)-[3,4′]bipyridinium Dibromide
(3,4′-m-BBV). To a solution of 3-bromomethylphenylboronic acid (0.69
g, 3.2 mmol) in DMF (25 mL) was added 3,4′-bipyridyl (0.2 g, 1.28
mmol), and the reaction was stirred at 80 °C for 48 h. The yellow
precipitate was collected by centrifugation, washed with DMF and then
1
acetone, and dried under a stream of argon (0.58 g, 76% yield). H
NMR (D2O, 500 MHz) δ 6.04 (s, 2H), 6.09 (s, 2H), 7.65 (q, J ) 7.5
Hz, 2H), 7.73 (d, J ) 7.5 Hz, 2H), 7.95 (m, 4H), 8.40 (dd, J ) 8.0, 6.5
Hz, 1H), 8.55 (d, J ) 7.0 Hz, 2H), 9.13 (d, J ) 8.5 Hz, 1H), 9.23 (m,
3H), 9.65 (s, 1H); 13C NMR (D2O, 125 MHz) δ 66.0, 66.6, 127.9,
130.4, 130.5, 132.8, 132.9, 133.1, 133.2, 135.6, 135.6, 136.5, 136.5,
136.6, 145.2, 146.4, 146.7, 147.4, 151.0; 11B NMR (80 MHz, D2O)
δ 24.8. HRMS-ESI m/z calcd for C24H24B2BrN2O4 [M - Br]+:
505.11056, found 505.10820.
Experimental Section
Synthesis. The syntheses of 4,4′-o-BBV, 4,4′-m-BBV, and 4,4′-BV
have been reported.9g
3,3′-Bipyridyl (2). To a 100-mL oven-dried round-bottomed flask
with a sidearm and condenser, was added 3-bromopyridine (1.16 mL,
12 mmol), 3-pyridineboronic acid (1.23 g, 10 mmol), and anhydrous
1,4-dioxane (20 mL) under argon. A degassed aqueous solution of Na2-
CO3 (2 M, 10 mL) was then added via syringe to the vigorously stirred
reaction mixture, followed by the addition of Pd(OAc)2 (0.11 g, 0.5
mmol) and PPh3 (0.65 g, 2.5 mmol). The reaction flask was then purged
using five argon/vacuum back-fill cycles, then stirred for 2 h at 95 °C.
After the mixture was cooled to room temperature, water was added
(50 mL), and the reaction was extracted with ethyl acetate (2 × 100
mL). The combined organics were washed with brine (2 × 75 mL),
dried with MgSO4, and evaporated to an oil under reduced pressure.
The residue was purified by chromatography on silica gel (pretreated
with 10% triethylamine in dichloromethane) using 20% ethyl acetate
in dichloromethane to give 1.0 g (64% yield) of clear oil. Spectroscopic
data were in accord with those previously reported.29 1H NMR (CDCl3,
500 MHz) δ 7.38 (dd, J ) 8.0, 5.0 Hz, 2H), 7.86 (dt, J ) 7.5, 2.0 Hz,
2H), 8.62 (dd, J ) 5.0, 1.5 Hz, 2H), 8.81 (d, J ) 2.5 Hz, 2H); 13C
NMR (CDCl3, 125 MHz) δ 123.9, 133.6, 134.5, 148.3, 149.4.
General. Reactions were performed using standard syringe tech-
niques and were carried out in oven-dried glassware under an argon
atmosphere. The 2- and 3-bromomethylphenylboronic acids were
purchased from Lancaster. All other reagents were purchased from
Aldrich. Dimethylformamide (DMF) was dried over CaH2 prior to use.
1H NMR spectra were recorded on a Varian spectrometer at 500 MHz
and are reported in ppm with respect to TMS (δ ) 0). Proton decoupled
13C NMR spectra were recorded on a Varian at 125 MHz and are
reported in ppm.27 11B NMR spectra were recorded on a Bruker at 80.25
MHz and are reported in ppm with respect to BF3·OEt2 (δ ) 0). High-
resolution mass measurements were obtained on a benchtop Mariner
ESITOF mass spectrometer or a JEOL JMS-AX505HA mass spec-
trometer from a matrix of p-nitrobenzyl alcohol for FAB.
3,4′-Bipyridyl (3). To a 100-mL oven-dried round-bottomed flask
with a sidearm and condenser, was added 3-bromopyridine (1.16 mL,
12 mmol), 4-pyridineboronic acid (1.23 g, 10 mmol), and anhydrous
1,4-dioxane (20 mL) under argon. A degassed aqueous solution of Na2-
CO3 (2 M, 10 mL) was then added via syringe to the vigorously stirred
reaction mixture, followed by the addition of Pd(OAc)2 (0.11 g, 0.5
mmol) and PPh3 (0.65 g, 2.5 mmol). The reaction flask was then purged
by using five argon/vacuum back-fill cycles and then stirred for 4 h at
95 °C. After cooling to room temperature, water was added (50 mL),
and the reaction was extracted with ethyl acetate (2 × 100 mL). The
combined organics were washed with brine (2 × 75 mL), dried with
MgSO4, and evaporated to an oil under reduced pressure. The residue
was purified by chromatography on silica gel (pretreated with 10%
triethylamine in dichloromethane) using 1:1 ethyl acetate/dichlo-
romethane to give 1.1 g (70% yield) of clear oil. Spectroscopic data
N,N′-Bis-(benzyl-2-boronic acid)-[3,3′]bipyridinium Dibromide
(3,3′-o-BBV). To a solution of 2-bromomethylphenylboronic acid (0.59
g, 2.76 mmol) in DMF (20 mL) was added 3,3′-bipyridyl (0.187 g, 1.2
mmol), and the reaction was stirred at 70 °C for 48 h. After the mixture
was cooled to room temperature, acetone (50 mL) was added to the
clear yellow solution. The resulting white precipitate was collected by
centrifugation, washed with acetone, and dried under a stream of argon
1
(0.56 g, 80% yield). H NMR (D2O, 500 MHz) δ 6.20 (s, 4H), 7.68
(m, 6H), 7.90 (d, J ) 7.0 Hz, 2H), 8.34 (dd, J ) 8.5, 6.5 Hz, 2H), 8.95
(d, J ) 8.0 Hz, 2H), 9.14 (d, J ) 6.5 Hz, 2H), 9.27 (s, 2H); 13C NMR
(27) Due to relaxed 13C-11B spin-spin coupling, signals for carbons directly
(28) Shiao, M. J.; Shieh, P.; Lai, J. S. Synth. Commun. 1988, 18, 1397-1402.
(29) Fort, Y.; Becker, S.; Caubere, P. Tetrahedron 1994, 50, 11893-11902.
attached to boron are not observed.
9
1284 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Boronic Acid-Based Glucose Sensors
A R T I C L E S
(D2O, 62.5 MHz) δ 65.2, 128.5, 129.7, 131.0, 131.3, 134.2, 135.0,
135.3, 143.2, 144.6, 145.4; 11B NMR (80 MHz, D2O) δ 28.7. HRMS-
ESI m/z calcd for C24H24B2BrN2O4 [M - Br]+: 505.11056, found
505.10291.
collected, of which 10432 were unique [R(int) ) 0.1020], and 5162
were observed [I > 2σ(I)]. The structure was solved by direct methods
(SHELXS-97) and refined by full-matrix least-squares on F2 (SHELXL97)
using 761 parameters. All non-hydrogen atoms were refined with
anisotropic thermal parameters. The H atoms on C atoms were generated
geometrically and refined as riding atoms with C-H ) 0.95-0.99 Å
and Uiso(H)) 1.2 times Ueq(C) for CH, CH2 and OH groups, Uiso-
(H)) 1.5 times Ueq(C) for CH3 groups. The H atoms on O atoms were
found in the electron difference map and refined as riding atoms with
O-H ) 0.86 Å and Uiso(H) ) 1.2 times Ueq(O). The maximum and
minimum peaks in the final difference Fourier map were 0.805 and
-0.429 e Å-3. Crystal Data: C108H102B6N6O36S6, MW ) 2317.18,
triclinic, P1h, a ) 9.5930(12) Å, b ) 13.9110(17) Å, c ) 20.342(2) Å,
R ) 102.126(2)°, â ) 90.323(2)°, γ ) 105.605(2)°. V ) 2550.6(5)Å,3
T ) 93(2) K, Z ) 1, R1 [I > 2σ(I)] ) 0.0629, wR2 (all data) ) 0.1810,
GOF (on F2) ) 0.981.
Fluorescence Measurements. General. All reagents and analytes
were purchased from Aldrich. All solutions were prepared with water
that was purified via a Barnstead NANOpure system (17.7 MΩ/cm).
Buffers (pH 7.4, 0.1 ionic strength) were freshly prepared from KH2-
PO4 and Na2HPO4 before use. pH measurements were taken on a
Denver Instrument UB-10 pH/mV meter and calibrated with standard
buffer solutions (pH 4, 7, and 10 from Fisher). Fluorescence spectra
were taken on a Perkin-Elmer LS50-B luminescence spectrometer, and
were carried out at 25 °C. Standard quartz fluorescence cuvettes were
used in all studies. HPTS was excited at 460 nm, and the emission
(λmax ) 510 nm) was collected from 480 to 650 nm. All experiments
were carried out in triplicate, and the errors in the reported binding
constants are based on the standard deviation of three independent
determinations. All data were analyzed using the Solver (nonlinear least-
squares curve fitting) in Microsoft Excel.
N,N′-Bis-(benzyl-3-boronic acid)-[3,3′]bipyridinium Dibromide
(3,3′-m-BBV). To a solution of 3-bromomethylphenylboronic acid (0.69
g, 3.2 mmol) in DMF (25 mL) was added 3,3′-bipyridyl (0.2 g, 1.28
mmol), and the reaction was stirred at 80 °C for 48 h. The white
precipitate was collected by centrifugation, washed with DMF and then
1
acetone, and dried under a stream of argon (0.57 g, 76% yield). H
NMR (D2O, 500 MHz) δ 6.04 (s, 4H), 7.59 (t, J ) 7.5 Hz, 2H), 7.69
(d, J ) 8.5 Hz, 2H), 7.87 (d, J ) 7.5 Hz, 2H), 7.90 (s, 2H), 8.37 (dd,
J ) 7.5, 6.0 Hz, 2H), 9.0 (d, J ) 8.0 Hz, 2H), 9.2 (d, J ) 6.0 Hz, 2H),
9.45 (s, 2H); 13C NMR (D2O, 125 MHz) δ 66.5, 130.4, 130.5, 133.0,
135.1, 135.7, 135.8, 136.6, 144.5, 146.2, 146.6; 11B NMR (80 MHz,
D2O) δ 26.0. HRMS-ESI m/z calcd for C24H24B2BrN2O4 [M - Br]+:
505.11056, found 505.10969.
N,N′-3,4′-Bis(benzyl)bipyridinium Dibromide (3,4′-BV). To a
solution of benzyl bromide (0.45 mL, 3.75 mmol) in DMF (5 mL)
was added 3,4′-bipyridyl (0.23 g, 1.5 mmol), and the reaction was stirred
at 70 °C for 16 h. The yellow precipitate was collected by centrifugation,
washed with acetone, and dried under a stream of argon (0.67 g, 90%
1
yield). H NMR (D2O, 500 MHz) δ 6.08 (s, 2H), 6.14 (s, 2H), 7.69
(m, 10H), 8.47 (dd, J ) 8.0, 6.0 Hz, 1H), 8.62 (d, J ) 7.0 Hz, 2H),
9.21 (dt, J ) 8.0, 2.0 Hz, 1H), 9.28 (d, J ) 7.0 Hz, 2H), 9.29 (m, 1H),
9.71 (s, 1H); 13C NMR (D2O, 62.5 MHz) δ 66.0, 66.6, 128.0, 130.5,
130.7, 130.8, 131.1, 131.5, 131.6, 133.5, 133.6, 136.5, 145.2, 146.5,
146.7, 147.4, 150.9. HRMS-FAB m/z calcd for C24H22BrN2 [M - Br]+:
417.0966, found 417.1004.
N,N′-3,3′-Bis(benzyl)bipyridinium Dibromide (3,3′-BV). To a
solution of benzyl bromide (0.45 mL, 3.75 mmol) in DMF (5 mL)
was added 3,3′-bipyridyl (0.23 g, 1.5 mmol), and the reaction was stirred
at 70 °C for 16 h. The yellow precipitate was collected by centrifugation,
washed with acetone, and dried under a stream of argon (0.67 g, 90%
yield). 1H NMR (D2O, 500 MHz) δ 6.12 (s, 4H), 7.70 (m, 10H), 8.45
(dd, J ) 8.0, 6.0 Hz, 2H), 9.07 (dt, J ) 8.0, 1.5 Hz, 2H), 9.27 (d, J )
6.0 Hz, 2H), 9.55 (s, 2H); 13C NMR (D2O, 62.5 MHz) δ 66.6, 130.4,
130.8, 131.1, 131.6, 133.5, 135.9, 144.6, 146.2, 146.6. HRMS-FAB
m/z calcd for C24H22BrN2 [M - Br]+: 417.0966, found 417.0956.
Cyclic Voltammetry. Cyclic voltammetry was performed on CHI
440 electrochemical workstation. A glassy carbon (GC) working
electrode (polished using METASERV 2000 Polisher/Grinder), a Ag/
AgCl reference electrode, and a platinum wire as counter electrode
were used as a normal three-electrode configuration. Potentials were
reported against the normal hydrogen electrode (NHE). The test
solution, containing 5 mM viologen in pH 7.4 phosphate buffer (0.1
M ionic strength), was purged with argon, and a stream of argon was
maintained above it to keep the solution anaerobic.
X-ray Structure Determination. (BBV2+)3(HPTS3-)2·2CH3-
OH·2H2O. A concentrated aqueous solution of 4,4′-m-BBV (1 mL)
was added to a concentrated aqueous solution of HPTS (1 mL), resulting
in immediate precipitation of a fluffy, pink solid. The solid was then
isolated and dissolved in hot methanol/water (10:1). After standing at
room temperature for 3 days, dark-red colored crystals formed.
A needle of dimensions 0.02 × 0.25 × 0.38 mm3 was mounted in
the 93(2) K nitrogen cold stream provided by a CRYO Industries low
temperature apparatus on the goniometer head of a Bruker SMART
Apex diffractometer equipped with an ApexII CCD detector. Diffraction
data were collected with graphite-monochromated Mo KR radiation
employing a 0.3° ω scans and approximately a full sphere of data to
2θmax of 55°. A multiscan correction for absorption was applied using
the program SADABS 2.10. Upon inspection of the scattering intensities
the data were truncated to 2θ )53°, yielding a total of 26407 reflections
Fluorescence Measurements for Quenching Studies. Fluorescence
measurements were done in situ by taking the emission spectrum of
HPTS at a series of quencher concentrations. The emission spectrum
of HPTS (2 mL of 4 × 10-6 M in buffer) was first obtained, quencher
was added (0.5-10 µL aliquots of 5 mM in buffer), the solution was
shaken for 30 s, and the new emission was measured after each quencher
addition.
Data Analysis. Stern-Volmer quenching constants were calculated
by fitting the data with eq 1:
F0/F ) (1 + Ks[Q]) eV[Q]
(1)
where F0 is the initial fluorescence intensity, F is the fluorescence
intensity after the addition of quencher, V is the dynamic quenching
constant, Ks is the static quenching constant, and [Q] is the quencher
concentration.13d
Fluorescence Measurements for Analyte Sensing Studies. The
emission spectrum of HPTS (2 mL of 4 × 10-6 M in buffer) was taken,
quencher was added (200 µL of 5 mM in buffer) to obtain a quencher/
HPTS ratio of 125:1, the emission measured, and then analyte solution
was added (0.5-10 µL aliquots of 1 M in pH 7.4 buffer); the solution
was shaken for 30 s, and the new emission was measured after each
addition of analyte.
Data Analysis. Apparent binding constants were calculated by fitting
the data with eq 2:
F/F0 ) (1 + (Fmax/F0)Kb[A])/(1 + Kb[A])
(2)
where F0 is the fluorescence intensity of the quenched dye, F is the
fluorescence intensity after the addition of analyte, Fmax is the calculated
intensity at which the fluorescence reaches its maximum, Kb is the
apparent binding constant, and [A] is the concentration of analyte.30
Acknowledgment. The authors thank the BioSTAR Project
and the Industry-University Cooperative Research Program with
(30) Cooper, C. R.; James, T. D. J. Chem. Soc., Perkin Trans. 1 2000, 963-
969.
9
J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1285

A R T I C L E S
Gamsey et al.
GluMetrics, Inc. for their financial support. We also thank
Professor Shaowei Chen for his helpful discussions.
coordinates, 1H and 13C NMR spectra for all compounds
reported; crystallographic information files (CIF). This material
is available free of charge via the Internet at http://pubs.acs.org.
Supporting Information Available: Fluorescence spectra,
binding isotherms, cyclic voltammograms, molecular modeling
JA066567I
9
1286 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007