Structural studies of biarylpyridines fluorophores lead to the identification of promising long wavelength emitters for use in fluorescent chemosensors

Article abstract of DOI:10.1016/j.tet.2004.08.049

Fluorescent chemosensors - molecules whose fluorescence emission changes in response to a reversible binding event - require both a substrate binding domain and a reporting fluorophore. Our approach to chemosensor development is based on a combination of a new signaling mechanism and a modular fluorophore synthesis. The latter feature has facilitated detailed study of the properties of polyarylpyridine fluorophores, and has led to the identification of a visibly-emissive pyridine as a promising lead structure for chemosensor development. The results of this study are described herein.

Full text of DOI:10.1016/j.tet.2004.08.049

Tetrahedron 60 (2004) 11075–11087
Structural studies of biarylpyridines fluorophores lead to the
identification of promising long wavelength emitters for use in
fluorescent chemosensors
*
A. G. Fang, J. V. Mello and N. S. Finney
Department of Chemistry and Biochemistry, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA 92093-0358, USA
Received 15 April 2004; revised 24 June 2004; accepted 19 August 2004
Available online 22 September 2004
Abstract—Fluorescent chemosensors—molecules whose fluorescence emission changes in response to a reversible binding event—require
both a substrate binding domain and a reporting fluorophore. Our approach to chemosensor development is based on a combination of a new
signaling mechanism and a modular fluorophore synthesis. The latter feature has facilitated detailed study of the properties of
polyarylpyridine fluorophores, and has led to the identification of a visibly-emissive pyridine as a promising lead structure for chemosensor
development. The results of this study are described herein.
q 2004 Elsevier Ltd. All rights reserved.
1
. Introduction
(4) Tuning of the emission wavelength by remote sub-
stituents on the pyridine ring.
The simple architecture of these systems is self-evident
Fluorescent chemosensors—molecules whose fluorescence
emission changes in response to a reversible binding
event—require a substrate binding domain, a reporting
fluorophore and a signaling mechanism that allows the two
to communicate. We have previously described the develop-
ment of fluorescent chemosensors that rely on confor-
2
(Fig. 1). The generality of the signaling mechanism is
attested to by the perfect correlation between metal–ion
1
binding detected by H NMR and observed fluorescence
response—in all cases to date, ion binding produces a
detectable change in fluorescence emission.
1
–4
mational restriction as signaling mechanism. This effort
was driven by the hypothesis that it would combine two
important and often mutually-exclusive features of other
systems: simplicity of molecular architecture and generality
of signaling mechanism. While we have conducted
extensive photophysical studies on chemosensors based on
biphenyl and biarylacetylene fluorophores, the most promis-
ing of our systems are the biarylpyridine, which exhibit
several of desirable properties:
In parallel with efforts to widen the variation of ligand
structure, we have synthesized numerous new arylpyridine
fluorophores with the objective of shifting both LE and CT
2
emission further into the visible region. We provide here a
detailed description of these efforts, including identification
of important structure–emission relationships, discovery of
a potentially serious limitation to our fluorophore design,
and a solution to this problem guided by an understanding of
structure–emission relationships.
(
(
(
1) Visible emission from a locally excited (LE) state that is
highly responsive to ion-binding induced conformational
restriction.
2) A second, longer-wavelength visible emission band
arising from a charge transfer (CT) state induced by
coordination of an ion to the pyridine nitrogen.
3) Modular synthetic assembly.
2. Background
Fluorescent chemosensors allow fluorescence detection,
with all the associated benefits, of non-fluorescent analytes.
1
The importance of fluorescent chemosensors in applications
as diverse as environmental monitoring, cellular imaging
and biomedical device construction is increasingly widely
appreciated. The majority of established chemosensors fall
into one of 3 categories:
Keywords: Fluorescence; Chemosensor; Pyridine; Synthesis; Spectroscopy.
*
Corresponding author. Present address: Organisch-chemisches Institut,
Universit a¨ t Z u¨ rich, Winterthurer-strasse 190, CH-8057 Z u¨ rich,
Switzerland. Tel.: C41 (01) 635 4283; e-mail: finney@unizh.ch
(1) Systems based on direct interaction between the analyte
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.08.049

11076
A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
Figure 1. An arylpyridine fluorescent chemosensor with dual visible emission.
lone-pair coordination. This generally constrains the
substrate scope to metal ions and dictates that the
fluorophore/receptor hybrid contain benzylic or anilinic
nitrogen atoms, which in turn restricts the design of new
fluorophores and/or binding domains.
(almost always a metal ion) and a nitrogen atom lone
pair that is ‘wired into’ the fluorophore (Fig. 2-1).
1
,5
(
(
2) Systems in which analyte binding displaces a fluor-
ophore or changes its microenvironment (Fig. 2-2).
1
,6
3) Systems in which analyte binding alters the energy
transfer between a fluorophore/acceptor pair, where the
acceptor is either another fluorophore or a quencher
In contrast, the third category, often based on fluorescence
resonant energy transfer (FRET) between donor and
acceptor fluorophores, typically relies on substantially
complex molecular architecture-polypeptides, oligonucleo-
1
,7,8
(Fig. 2-3).
The first category, which represents the statistical
majority of reported fluorescent chemosensors, has the
significant advantage of relying on relatively simple
molecular architecture. Important examples of this
7
,8
tides or fusion proteins. Important embodiments of this
strategy include fusions of green fluorescent protein (GFP)
analogs and calmodulin that can be expressed in vivo to
2
C
2C
2C
7a
approach include Ca - and Zn -responsive chemo-
5
facilitate Ca imaging in cells. Offsetting the limitation
of complexity is the generality of the signaling mechanism,
which places no constraint on the structure of the binding
c,d
sensors with applicability for cellular imaging. The
central limitation of this approach is the requirement for
Figure 2. Broad overview of fluorescent chemosensor strategies.

A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
11077
restriction represent a fourth category addition to the
above list. As mentioned above, the modular nature of
biarylpyridine fluorophores (our preferred member of this
2
fourth category) simplifies structure–property studies. The
goals of the present work were gaining a greater under-
standing of substituent effects and red-shifting the LE and
CT emission wavelengths. We initially focused on the
nature of the aryl substituents at the 2- and 6-positions
(
Fig. 3), and subsequently investigated fluorophores with
varied 4-substituents in greater detail (Fig. 4). With regard
to LE and CT emission, the default emission of biarylpyri-
dines is that from the LE state. However, CT emission can
be induced by simple protonation of the pyridine (Fig. 5).
For both states, the longest wavelength absorption maxi-
mum correlates with the excitation maximum for fluor-
escence emission. In cases where the CT state is non-
emissive (vide infra) the formation of the CT state can still
be verified by characteristic changes in the absorption
spectrum (Fig. 5).
Figure 3. Biarylpyridines with varying substituents at the 2 and 6-positions.
domain beyond a requirement that binding induce a large
change in interchromophore distance.
3.2. Synthesis of arylpyridine fluorophores
The second category, initially based primarily on displace-
ment of covalently-tethered fluorophores from the cavity of
cyclodextrin hosts, are appropriately intermediate between
the first and third categories in terms of complexity and
The syntheses of compounds 3 and 8–11 have been
described previously. Fluorophores 1, 2 and 4–6 were
2
prepared from 2,6-dibromopyridine, while 7 and 12 were
prepared from 2,6-dichloro-4-phenylpyridine and 2,6-
dichloro-4-carboxymethylpyridine, respectively (Scheme 1).
Central to all of these syntheses is the efficiency of palladium
6
generality. Embodiments typically rely on structures no
more complicated than cyclodextrin, and generality is
limited only by the requirement that there be sufficient
similarity between the fluorophore and the analyte that the
analyte can effectively compete with the fluorophore in
binding to the receptor. Important recent examples of this
2
,9,10
catalyzed cross coupling of the required halopyridines.
6
b
4. Discussion
strategy include systems for sensing ATP.
4.1. Influence of the 2,6-substituents on optical
properties
3. Results
Our first generation biarylpyridine chemosensors (cf. Fig. 1)
0 0
all contained as common features o,o -alkyl and p,p -OCH
3
3
.1. Structural studies on ‘fourth category’ fluorophores
substituents on the aryl groups at the 2 and 6 positions. The
0
Fluorescent chemosensors based on conformational
presence of the o,o groups was presumed to be essential
Figure 4. Biarylpyridines with varying substituents at the 4-position.
Figure 5. Absorption and emission spectra of the chemosensor from Figure 1.

11078
A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
states. The reduction in LE emission parallels that seen
when comparing biphenyl (fZ0.18) to the more twisted
0
11
derivative 2,2 -dimethylbiphenyl (fZ0.01). Based on
previous work with biphenyls, we conclude that this
reduction in quantum yield is in large part due to the
increased efficiency of spin-orbit coupling (SOC) in twisted
1
2
biaryl excited states. While factors independent of biaryl
torsion might be considered, such as C–C twisting and C–H
stretching-mediated deactivation of the excited state, the
0
similarity of the quantum yields of biphenyl and 3,3 -
1
1
dimethylbiphenyl (fZ0.21) argues against this. The ‘gap
rule’—the empirical observation that the rate of IC
decreases as excited state energy increases—also argues
against this possibility, in that the excited state energy is in
3
a,13,14
excess of 100 kcal/mol.
0
An additional consequence of
introducing o,o -CH groups is an undesirable blue shift in
3
both the LE (20 nm) and CT (5 nm) emission. This is also
consistent with reduced planarity and thus reduced conju-
gation in the excited state.
Scheme 1. Synthesis of arylpyridine fluorophores.
based on our previous study of biphenyl systems. However,
the properties of biarylpyridines and biphenyls are suffi-
ciently different that this presumption warranted direct
0
dramatic impact on the efficiency of both absorption and
The addition of p,p -OCH substituents (2 vs 3) has a
3
emission, almost completely compensating for the reduction
0
0
evaluation. The p,p -substituents were initially included
in quantum yield caused by introduction of the o,o -CH
3
based on the ready availability of the precursor aryl
fragments rather than deliberate design, and the inclusion
of these groups even more clearly required evaluation.
groups. In addition, it has a pronounced effect on the
absorption and emission wavelengths. Notably, the emis-
sion maximum for the LE state shifts more than 20 nm to the
red, while that of the CT state shifts by more than 60 nm.
The emission shifts are accompanied by smaller but still
significant red shifts in the LE and CT excitation maxima.
These indicate that both the LE and CT excited states are
highly polarized, with the flanking aromatic rings serving as
electron donors and the pyridine ring, consistent with its
electron deficient nature relative to a hydrocarbon, serving
as an electron acceptor. While this polarization is present in
the ground state, it is clearly more pronounced in the excited
states: the near doubling of 3 reflects a large increase in the
transition dipole for the excitation from ground to excited
The spectroscopic properties of 1–6 (Table 1) reveal that
both the alkyl and OCH substituents are essential to the
3
0
function of our system. The inclusion of the o,o -alkyl
substituents is essential for lowering the LE quantum yield
to the point where it is subject to modulation by
conformational restriction. For instance, the fluorophore in
Figure 1 has an LE quantum yield of w0.02; in contrast,
fluorophores with LE quantum yields O0.10 no longer
respond strongly to binding-induced conformational restric-
tion. (This is one of the ironies inherent in an approach
based on using substrate binding to turn a poor fluorophore
into a good fluorophore—it is essential to start with a poor
fluorophore.) The presence of the polarizing OCH3
substituents provides a desirable red-shift in absorption
and emission but what would be, in the absence of the alkyl
substituents, an undesirable increase in LE quantum yield.
1
5
state. A corresponding increase in the efficiency of
radiative decay of the LE state, shown by the increase in
f, is expected based on the proportionality of the Einstein A
and B coefficients (which govern emission and excitation)
1
6
for fluorophores of very similar structure. The red shifts in
emission wavelength indicate that the p-OCH groups serve
3
to stabilize both the LE and CT excited states, effectively
lowering their energy and reducing the energy gap between
excited and ground state. The effects on absorption and
emission are much more pronounced in the CT state, as
0
The addition of o,o -CH groups to 2,6-biphenylpyridine (1
3
vs 2) leads to a slight reduction in extinction coefficient and
quantum efficiency of emission from both the LE and CT
Table 1. Salient optical properties of 1–6
3
3 (!10 )
a
a,b
F (LE)
c
d
e
f
g
Compound
F (CT)
LEexc
LEem
CTexc
CTem
1
2
3
4
5
6
10.2
8.1
14.8
11.6
9.3
0.037
0.004
0.031
0.560
0.003
0.059
0.951
0.831
0.177
0.544
0.021
0.004
302
280
290
315
272
303
338
320
352
359
342
378
322
307
342
366
324
359
393
388
449
459
450
480
13.5
a
Measured at LE excitation maximum.
Quantum yield of LE emission.
Quantum yield of CT emission.
Excitation maximum for LE state.
Emission maximum for LE state.
Excitation maximum for CT state.
Emission maximum for CT state.
b
c
d
e
f
g

A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
11079
expected based on its greater charge separation relative to
the LE state.
twisted, such that the additional CH groups have relatively
3
little influence on the conformation and thus energetic
separation of the CT state and the corresponding twisted
conformation of the ground state. While this evidence in
indirect, it is consistent with the extensive literature on
0
The importance of the o,o -CH groups is underscored by
3
0
comparison to the corresponding m,m -dimethyl case (3 vs
4
1
7,18
). Relocation of the flanking CH3 groups lowers the
twisted intramolecular charge transfer (TICT) states.
absorptivity of the molecule while simultaneously increas-
ing the LE quantum yield by an order of magnitude and
doubling the CT quantum yield. In addition, the excitation
maxima for the m,m isomer are significantly red shifted,
although there are only very small changes in the emission
There are two significant aspects of the spectroscopic
properties of bis(piperonyl) compound 6. First, the further
increase in polarization induced by two additional alkoxy
substituents (6 vs 3) leads to the longest LE and CT
emission wavelengths in the series (380 and 480 nm,
0
wavelengths. The decreased absorptivity indicates that
0
1
9
respectively, compared to 350 and 450 nm for 3). Second,
and of great practical importance, emission from 6 is a
presage of a trend that is more fully illustrated by 7–12: as
emission from biarylpyridines moves farther to the red, the
quantum yield of the CT state begins to drop off. The first of
these observations may be rationalized much as the
difference between 2 and 3, with electron donating
substituents stabilizing the polarized excited states and
reducing the energetic separation from the ground state. The
second and more problematic observation is again consist-
removal of the o,o -CH groups leads to greatly increased
3
conjugation—and thus polarization—in the ground state,
while the smaller changes in the emission wavelengths
indicate smaller changes in the degree of polarization in the
excited state. The lower value of 3 for 4 is thus explained, as
3
vary with the difference in polarization of the ground and
should scale with the transition dipole which should in turn
1
5
excited states. That is, increasing the polarization of
the ground state without altering the polarization of the
transition state would be expected to decrease 3, in
the absence of other effects. The smaller red shifts in the
1
4
ent with the gap rule. For 6, the 480 nm CT emission is
equivalent to an energy gap of almost exactly 60 kcal/mol,
which correlates with significant decrease in CT quantum
yield (fZ0.06, vs 0.28 for 3). This energy (ca. 60 kcal/mol)
is apparently the upper limit at which IC begins to
contribute to nonradiative decay.
emission wavelengths are consistent with greater excited
0
state planarization, and the effect of removing the o,o -CH
groups is similar in comparing 3 vs 4 and 2 vs 1.
3
The influence of deplanarization is further delineated by the
addition of a second pair of flanking CH groups (3 vs 5).
While deplanarization and polarization offset one another in
3
4.2. Influence of the 4-substituent on optical properties
3
ground state influence of polarization, and the values for 3
, in 5 deplanarization almost fully compensates for the
The influence of the 4-substituent on the emission of
biarylpyridines is delineated by compounds 7–12 (Table 2).
Three prominent trends can be seen. First, increasing the
chromophore surface or polarization by substitution at the 4-
position can lead to a significant red shift in LE emission,
and in one case can alter the identity of the core fluorescing
unit. Second, the quantum yield for LE emission increases
as emission moves to longer wavelengths. Third, CT
emission decreases over the same series, leading quickly
to molecules with non-emissive CT states.
K3
(
reduced by w4!10 ) and the absorption maxima
reduced by w20 nm) of 5 are even lower than those of 2,
(
which lacks both the OCH and additional CH groups.
3
3
Similarly, the quantum yield of both the LE and CT states
descend to values similar to those of 2. The emission
maxima provide valuable structural information. The
wavelength of LE emission shows a w20 nm blue shift
relative to 3, while that the CT emission remains unchanged
within measurement error. This indicates that the ground
and LE states of 5 are less planar and thus less conjugated
than those of 3, as would be expected based on increased
steric congestion. That this congestion does not influence
the CT emission wavelength suggests that the CT state is
By itself, comparison of 2,6-bis(o-tolyl)pyridine and its 4-
phenyl counterpart (7 vs 2) would suggest that the 4-
substituent has a modest influence on most of the properties
Table 2. Salient optical properties of 7–12 (with 2, 3 and 5 for comparison)
g
3
3 (!10 )
a
b
c
d
e
f
Compound
F (LE)
F (CT)
LEexc
LEem
CTexc
CTem
7
8
9
1
1
1
2
3
5
4.9
9.3
38.3
4.4
14.6
4.7
8.1
0.005
0.200
0.210
0.230
0.140
0.070
0.004
0.031
0.003
0.803
0.182
0.016
—
—
—
0.051
0.282
0.021
295
310
310
324
300
300
280
290
272
349
379
400
424
440
457
320
352
342
302
350
370
379
—
360
307
342
324
399
483
505
—
—
—
388
449
450
0
1
2
14.8
9.3
a
b
c
d
e
f
Measured at LE absorption/excitation maximum.
Quantum yield of LE emission.
Quantum yield of CT emission.
Excitation maximum for LE emission.
Emission maximum for LE excitation.
Excitation maximum for LE state.
Emission maximum for CT excitation.
g

11080
A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
of these fluorophores. The quantum yields are virtually
identical, the extinction coefficient is actually slightly
reduced, and the emission wavelengths for the LE and CT
states shift to longer wavelength by 30 and 15 nm,
respectively. The longer emission wavelength is expected
based on the increase in fluorophore surface area, thus
delocalization and stabilization of the excited state. It is not
immediately obvious why the LE emission wavelength
responds more strongly than the CT state to the increased
delocalization, or why the CT quantum yield responds more
strongly that than of the LE.
absorption spectrum of 8 is similar to that of 3 and the 4-
vinyl derivative (Fig. 3). Immediate differences in peak
shape and the number and relative intensity of the
absorption maxima can be seen for 9. This demonstrates
that the benzofuran substituent is a powerful determinant of
the nature of the excited states, with commensurate
deemphasis of the 2,6-aryl groups.
Installation of a benzoxazole at the 4-position yields a
chromophore with absorption properties that no longer
resemble those of 1–10 (Fig. 7), and the absorption and
emission of which do not change noticeably upon addition
of excess TFA. This structural variation has thus altered the
fluorophore identity to the point that a CT excited state—
radiative or nonradiative—simply no longer exists.
Addition of a 4-phenyl group to the chromophore bearing
both o-CH and p-OCH groups (3) leads to a much more
3
3
dramatic change in spectroscopic properties (8 vs 3). The
extinction coefficient decreases, indicative of a lower
transition dipole that presumably results from an increase
in ground state polarization relative to the LE state. The LE
quantum yield increases substantially (fZ0.20 vs fZ0.03
for 3) while the CT quantum yield decreases (fZ0.09 vs
fZ0.28 for 3). Absorption and emission wavelengths for
both states shift to longer wavelength, the LE emission now
approaching the visible region of the spectrum. The reduced
CT emission efficiency is a consequence of the gap rule, as
described above. The increased LE efficiency must result from
the extended conjugation of the chromophore surface, and we
take this as evidence that the 2,6-substituents—and their
conformation mediated nonradiative decay—contribute less,
proportionally, to the properties of more extended chromo-
phores. (This is substantiated by 9 and 11; vide infra.)
As a counterpart to increasing the chromophoric surface,
placing a methyl carboxylate (which extends conjugation by
only 2 atoms) at the 4-position shows that direct polarization
of the pyridine, like polarization of the 2,6-biaryl groups,
also influences fluorophore behavior (10 vs 3). The
extinction coefficient decreases, consistent with a more
polarized ground state and less difference in polarization
between the ground and excited states. The LE emission
becomes slightly more efficient and moves further into the
visible, the 425 nm maximum representing a 100 nm Stokes
shift. The CT state—which can still be observed by
absorbance—is non-emissive.
The final variant on 4-substitution, 12, is the tetramethyl
analog of 10. It is notable in that while it has a shorter
wavelength LE absorption, as would be predicted by
sterically decreased conjugation, it has a longer emission
wavelength. The red shifted emission, which corresponds to
w62 kcal/mol, now runs afoul of the gap rule and accounts
for the reduced quantum yield, That a longer emission
wavelength results from increased deplanarization suggests
that the optimal conformation of the LE state is no longer as
near planar as was the case for 3. Although this suggests an
optimal geometry more suited to the CT state, which is
nonradiative and expected to be twisted, the CT state can
clearly be identified by the absorption spectrum and is
distinct from the LE state. The apparent change in optimal
LE state geometry remains unexplained, but coupled with
the reduced quantum yield it provides an opportunity for the
development of fluorescent chemosensors with visible
Replacing the phenyl group with a benzofuran provides
further evidence that increased chromophore delocalization
diminishes the impact of nonradiative decay mediated by
the 2,6-aryl substituents (9 vs 3). Extension with benzo-
furan leads to a large increase in extinction coefficient, a
further increase in LE quantum yield and a further reduction
in CT quantum yield. In addition, the LE emission
maximum moves into the visible region (400 nm), and the
CT emission maximum moves from green into the green-
yellow (505 nm). The emission and quantum yield effects
may be rationalized as were those for 8. The near
quintupling of extinction coefficient is surprising but, like
the increased LE quantum yield, shows that the 4-
substituent makes an increased contribution to the character
of the excited state. This can be seen most readily by
comparing the absorption spectra of 8 and 9 (Fig. 6). The
Figure 6. Absorption spectra of 8 and 9.
Figure 7. Comparison of the absorption spectra of 3 and 11.

A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
11081
emission and geometrical constraints different from those
associated with 3.
on the absorbance and emission of these fluorophores. With
0
regard to the 2,6-aryl groups, o,o -alkyl substitution was
shown to be essential for maintaining low initial quantum
0
yields, while p,p -alkoxy substitution was shown to provide
4
.3. The limitations of longer wavelength fluorophores
a beneficial red shift in emission without untoward increase
in emission efficiency. The influence of the 4-substituent on
absorption and emission wavelengths was dramatic, with
LE emission in the green region of the visible spectrum
being possible. An unfortunate side effect of increased LE
and CT emission wavelengths is the accompanying
variation in quantum yields. LE emission efficiency (ideally
low) increases with wavelength, while CT emission
efficiency (ideally high) decreases with wavelength. While
in certain cases we have found that fluorescent chemo-
sensors with relatively high initial quantum yields (e.g., fZ
0.23 for 13) retain useful metal ion response, this is not
generally the case.
The properties of the pyridine fluorophores with visible LE
emission constitute a potential problem for the development
of fluorescent chemosensors based on conformational
control. The approach hinges on being able to start with a
poor fluorophore and induce increased LE emission
efficiency via binding induced conformational restriction.
This, in turn, means the approach is best suited for
fluorophores with low initial quantum yields-unlike the
visibly-emissive 8–11. In addition, the loss of the second
signaling channel—the CT emission—is a setback, in that
dual emission provided an additional level of discrimination
for distinguishing between bound metal ions.
The first of these concerns, reduced sensitivity at longer LE
emission wavelength, is ameliorated to an extent by the
observation that 13, the 4-carboxymethyl analog of the
chemosensor shown in Figure 1, still provides a visible
response to the presence of metal ions (Fig. 8). The w4-fold
increase in emission intensity translates to a fR0.90, which
is a record among the systems we have studied, and is easily
observed by eye. Offsetting this promising result are the lack
Two recent observations represent promising leads for the
development of the next generation of pyridine-derived
fluorescent chemosensors. The first of these arose from
juxtaposing the 4-carboxylate substituent found in 10 and 13
with the 4-vinyl group found in the chemosensor from
Figure 1. By preparing vinylogous amide 14 (Fig. 9), we
have found a fluorophore that benefits from a red shift
induced by the carboxylate substituent, but does not suffer
from the increase in quantum yield seen in 10 or 13. The
second, and more belated, observation is that the importance
2
C
2C
of response to Mg
previously induced formation of the radiative CT state.
and Ca
(cf. Fig. 1), which had
C
While treatment with TFA indicates that 13$H possesses a
0
of the p,p -methoxy groups may have been overestimated.
2
C
2C
nonemissive CT state, neither Mg
nor Ca
induce
Podand 15—similar to 13 and the podand of Figure 1—
responds strongly to presence of metal ions (Fig. 10). While
the absence of the methoxy groups leads to emission
wavelengths that are much shorter than desirable, the lower
initial quantum yield (f%0.01) allows for a much greater
dynamic range for the fluorescence response.
formation of this state, indicating that the pyridine ring is
now sufficiently electron deficient that the nitrogen atom
serves as a poor Lewis base. Related studies with the podand
derived from the 4-phenyl fluorophore (9) showed only
minimal metal ion response, an observation which suggests
C
that the response of 13 to Li is the exception rather than
the rule. That said, it is worth noting that 13 still provides a
Our ongoing efforts focus on gaining a greater under-
standing the properties of 14, exploiting this fluorophore
for the development of fluorescent chemosensors, and
turning the luminescence of long wavelength CT
emission back on. In addition, based on evaluation of
C
visible emissive readout of the presence of Li .
5
. Conclusion and recent progress
1
of the 4-substituent in the series of fluorophores lacking
the p,p -methoxy groups.
5, it is clearly necessary to reinvestigate the influence
In summary, variation of the substituents at the 2, 4 and 6-
positions of polyaryl pyridines can have a profound effect
0
K5
C
3
Figure 8. Titration of 13 (10 M in CH CN) with Li .

11082
A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
Figure 9. Vinylogous amide 15 has desirable visible LE emission.
K5
3 3
Figure 10. Titration of podand 15 (10 M in CH CN) with metal ions. Emission maxima in broken box are Raman scattering from CH CN.
6. Experimental
solutions were equilibrated by stirring prior to acquiring
fluorescence spectra. No efforts were made to exclude water
or air. Quantum yields and extinction coefficients were
6
.1. General notes
2
0
determined by standard methods.
Melting points were obtained in open capillary tubes with a
Thomas Scientific Uni-Melt melting point apparatus and
are uncorrected. H NMR spectra were obtained on Varian
HG-400 (400 MHz) spectrometers. Chemical shifts (d) are
reported in parts per million (ppm) relative to residual
Chromatographic purifications were performed by flash
chromatography with silica gel (Fisher Scientific, 32–
63 mm) packed in glass columns; eluting solvent for each
purification was determined by thin layer chromatography
(TLC). Analytical TLC was performed on glass plates
coated with 0.25 mm silica gel 60 F254 (EM Science) using
UV light for visualization.
1
solvent (CHCl ), s, d, 7.26). Multiplicities are given as: s
3
(
doublets), ddd (doublet of doublet of doublets), dt (doublet
singlet), d (doublet), t (triplet), q (quartet), dd (doublet of
1
3
of triplets) and m (multiplet). Proton-decoupled C NMR
spectra were obtained on Varian HG-400 (100 MHz)
spectrometers. C chemical shifts are reported relative to
Synthetic procedures were carried out under an inert
atmosphere, using standard Schlenk line techniques.
Where noted, solutions were degassed by evacuating and
1
3
K1
CDCl (t, d, 77.0). IR stretches are given in cm ; spectra
3
were obtained on a Nicolet 550 Series II Spectrophotometer.
MALDI-FT mass spectroscopic analyses were provided by
the facility at The Scripps Research Institute; all other mass
spectroscopic analyses were provide by the facility at
UCLA.
backfilling with N
benzene were dried by passage through a column of
activated alumina. Pyridine was distilled from CaH . All
2
other reagents and solvents were used as received unless
otherwise specified.
several times. THF, CH Cl , and
2 2 2
2
1
9
,10
Fluorescence measurements were carried out in spectro-
scopic grade CH CN on a PTI Quantamaster 2000, with
6.2. Experimental procedures
3
flash lamp excitation and 2 nm excitation and emission slit
widths. Solutions of fluorophore were prepared by succes-
Compounds 3 and 8–11 were prepared as described
previously. Compound 1 is commercially available but
2
K5
sive dilution and were typically 1!10 M in CH CN.
was prepared as described below for convenience. Com-
pound 14 was prepared in 5 steps and good yield from ester
10 (Scheme 2). Compound 13 and the fluorescent
chemosensor shown in Figure 1 (25) were prepared from
2-bromo-5-methoxytoluene (Scheme 3). Compound 15 was
prepared in analogy to 2 by variation of the boronic acid.
3
Solutions of metal perchlorate salts were prepared by
successive dilution and were typically 1 M. Fluorescence
titrations were carried out by sequentially adding 0.005 or
0
2
.010 mL aliquots of metal solution via micropipette to
.500 mL of fluorophore solution in a quartz cuvette. The

A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
11083
1
m, 6H), 2.43 (s, 6H). C NMR (100 MHz, CDCl3):
3
(
dZ159.24, 140.48, 136.11, 135.62, 130.50, 129.67, 128.01,
1
1
25.67, 121.83, 20.69. IR (KBr): nZ3030, 2951, 1562,
C
457, 767. HRMS (EI): calcd for C H N [M ] 259.1361;
1
9 16
found 259.1349. TLC (R )Z0.4 (5% acetone/hexanes).
f
6
1
.2.3. Compound 4. Procedure as for 2. White solid, mp
1
40–141 8C, 26%. H NMR (400 MHz, CDCl ): dZ7.95
3
(
6
m, 4H), 7.72 (t, JZ7.6 Hz, 1H), 7.56 (d, JZ8.0 Hz, 2H),
.94 (d, JZ8.8 Hz, 2H), 3.91 (s, 6H), 2.34 (s, 6H). C NMR
1
3
(
1
100 MHz, CDCl ): dZ158.42, 156.41, 136.95, 131.74,
29.16, 126.56, 125.50, 117.06, 109.82, 55.47, 16.58. IR
3
Scheme 2. Preparation of vinylogous amide 14. Reagents and conditions:
(
a) DiBAl/THF, K78 8C, 89%; (b) DMSO, (COCl)
2
, NEt
, THF, 0 8C, 100%; (d)
, DEPBT, THF, RT, 94%.
3 2 2
, CH Cl , K78 8C,
t
K C
(KBr): nZ2943, 1615, 1562, 1510, 1449, 1256, 1143, 1029.
94%; (c) (EtO)
LiOH$2H
2 2 2
P(O)CH CO Et, BuO K
C
HRMS (EI): calcd for C H NO [M ] 319.1572; found
2
O, THF, RT 85%; (e) BnNH
2
21 21
2
3
19.1569. TLC (R )Z0.20 (10% EtOAc/hexanes).
f
6.2.4. Compound 5. 4-Bromo-3,5-dimethylanisole (0.25 g,
1.17 mmol, 2.00 equiv) was cooled to K78 8C in THF
(5 mL). n-BuLi (1.22 mL 1.20 M in hexanes, 1.46 mmol,
2
1
.50 equiv) was added dropwise and stirred at K78 8C for
5 min. ZnCl (0.40 g, 2.92 mmol, 5.00 equiv) was trans-
2
ferred via cannula into the reaction mixture as a THF
solution (10 mL), and the reaction was stirred for an
additional 15 min. The reaction mixture was transferred via
cannula into a flask containing 2,6-dibromopyridine (0.14 g,
0
0
.58 mmol, 1 equiv) and Pd(PPh3)4 (0.03 g, 0.03 mmol,
.05 equiv). The reaction mixture was degassed then
refluxed for 20 h. The reaction was allowed to cool to
room temperature, H O (25 mL) was added, and the mixture
2
was extracted with EtOAc (3!25 mL). The combined
organic fractions were washed with brine (1!25 mL), dried
over Na SO , filtered, and concentrated under vacuum.
2
4
Purification by column chromatography (20% EtOAc/
hexanes) produced 0.09 g (0.27 mmol, 46%) of 5 as a pale
yellow solid, mp 167–168 8C. R Z0.24 (20% EtOAc/
f
Scheme 3. Preparation of 13 and the chemosensor from Figure 1 (25).
CH2-
hexanes). IR (KBr): nZ2961, 1602, 1575, 1463, 1315,
2
Reagents and conditions: (a) NBS, AIBN, PhH, 86%; (b) CH
3
OCH
2
K1 1
1
194, 1160, 1074, and 838 cm . H NMR (CDCl ): dZ7.80
i
3
OCH
dichloro-4-carboxy-methylpyridine, Pd
DiBAlH, THF, K78 8C, 99%; (f) DMSO, (COCl)
2
OH, NaH, THF, 80%; (c) BuLi, B(O Pr)
3
, THF; NaOH, 70%; (d) 2,6-
t
2
(t, JZ8 Hz, 1H), 7.17 (d, JZ8 Hz, 2H), 6.64 (s, 4H), 3.80 (s,
2
(dba)
3
, P Bu
3
, CsF, THF, 52%; (e)
, CH Cl , K78 8C,
13
2
, NEt
3
2
2
6H), and 2.06 ppm (s, 12H). C NMR (CDCl ): dZ159.7,
3
C K
90%; (g) Ph
3
PCH
3
I
, BuLi, THF, 0 8C, 60%.
158.7, 137.1, 136.4, 122.7, 112.9, 112.7, 55.2, and
20.7 ppm. HRMS (DEI): calculated for C H NO (MK
H ): 346.180704, found 346.181374.
23 24 2
C
6
0
1
0
0
.2.1. Compound 1. 2,6-Dibromopyridine (0.100 g,
.422 mmol, 1 equiv), phenylboronic acid (0.155 g,
t C
.27 mmol, 3.00 equiv), [HP Bu ] BF
K
(0.012 g,
.040 mmol, 0.10 equiv), Pd (dba) (0.019 g, 0.021 mmol,
3
4
6.2.5. Compound 6. 2-Bromo-4,5-methylenedioxytoluene
(0.25 g, 1.16 mmol, 3.00 equiv) was cooled to K78 8C in
THF (5 mL). n-BuLi (0.67 mL/1.8 M, 1.20 mmol 3.10 equiv)
was added and the reaction stirred for 5 min, ZnCl2
2
3
.050 equiv), and CsCO (0.756 g, 2.320 mmol, 5.50 equiv)
3
was suspended in THF (10 mL). The reaction mixture was
degassed and refluxed for 16 h. After cooling to room
temperature, H O (50 mL) was added and the mixture
(0.181 g, 1.65 mmol, 3.50 equiv) in THF (3.0 mL) was
2
added. The arylzinc chloride so generated was transferred
via cannula into a degassed solution of 2,6-dibromopyridine
(0.0920 g, 0.387 mmol, 1 equiv), Pd (dba) (0.018 g,
extracted with methylene chloride (2!50 mL). The com-
bined organic extracts were dried over Na SO and
2
4
2
3
t
concentrated. Purification by column chromatography
20% toluene/hexanes) produced 0.098 g (0.422 mmol,
0
0
.019 mmol, 0.050 equiv, P Bu (0.56 mL, 0.14 M in THF,
3
(
4
.077 mmol, 0.20 equiv), and refluxed for 16 h. Upon
1
8%) of 1 as a white solid, mp 78–80 8C. H NMR
cooling to room temperature, the reaction mixture was
diluted with EtOAc (1!50 mL), washed with H O (1!
(
400 MHz, CDCl ): dZ8.18 (m, 4H), 7.82 (dd, 6.8, JZ7.0 Hz,
3
2
1
H), 7.71 (d, JZ7.6 Hz, 2H), 7.52 (m, 4H), 7.46 (m, 2H). IR
250 mL) and saturated NaCl (1!50 mL). The organic layer
was dried over Na SO , concentrated, and chromatographed
(
KBr): nZ3056, 2925, 1588, 1449, 758, 697. TLC (R )Z
f
2
4
0
.11 (20% toluene/hexanes).
on silica gel (10% EtOAc/hexanes) to afford a white solid
(
1
0.047 g, 35%), mp 149–151. H NMR (400 MHz, CDCl ):
3
t
.2.2. Compound 2. Procedure as for 1, except P Bu in
6
THF was used as the phosphine source. White solid, mp
dZ7.75 (t, JZ7.6 Hz, 1H), 7.28 (d, JZ7.6 Hz, 2H), 6.96 (s,
2H), 6.74 (s, 2H), 5.95 (s, 4H), 2.33 (s, 6H). C NMR
3
13
1
6
8–70 8C, 65%. H NMR (400 MHz, CDCl ): dZ7.81 (t,
(100 MHz, CDCl ): dZ158.83, 147.20, 145.50, 135.98,
3
3
JZ7.6 Hz, 1H), 7.45 (m, 2H), 7.36 (d, JZ7.6 Hz, 2H), 7.27
133.73, 129.43, 121.75, 110.45, 109.91, 100.85, 20.59. IR

11084
A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
(
calcd for C H NO [(MKH) ] 346.1079; found
KBr): nZ2925, 1562, 1510, 1449, 1047, 933. HRMS (EI):
50 mL), and saturated NaCl (1!25 mL). The organic phase
was dried over Na SO and was concentrated. Purification
C
2
1
16
4
2
4
3
46.1084. TLC (R )Z0.20 (12% EtOAc/hexanes).
by flash column chromatography (30% EtOAc/hexanes)
provided 0.063 g (0.131 mmol, 94%) of a yellow solid. H
f
1
6
1
.2.6. Compound 7. Procedure as for 2. White solid, mp
NMR (CDCl ): dZ7.69 (d, 1 h, JZ15.6 Hz), 7.35 (m, 9H),
6.82 (m, 4H), 6.60 (d, 1H, JZ15.6 Hz), 5.98 (t, 1H, JZ5.6 Hz),
4.60 (d, 2H, JZ5.6 Hz), 3.84 (s, 6H), 2.44 (s, 6H). IR (KBr):
3
1
29–130 8C, 75%. H NMR (400 MHz, CDCl ): dZ7.72
3
(
(
dd, JZ1.6, 6.8 Hz, 2H), 7.60 (s, 2H), 7.51 (m, 5H), 7.29
13
m, 6H), 2.48 (s, 3H). C NMR (100 MHz, CDCl3):
nZ3283 (b), 2934, 1676 (s), 1615, 1239, 1047, 741. HRMS
C
dZ159.84, 148.55, 140.51, 138.27, 135.71, 130.52, 129.67,
28.95, 128.85, 128.07, 126.95, 125.68, 119.94, 20.75. IR
KBr): nZ3039, 2925, 1588, 1545, 1396, 767. HRMS (EI):
(EI): calcd for C31
478.2266. TLC (R )Z0.33 (35% EtOAc/hexanes)
f
H N O [M ] 478.2256; found
30 2 3
1
(
C
calcd for C H N [(M–H) ] 334.1596; found 334.1592.
TLC (R )Z0.34 (5% EtOAc/hexanes).
2
5 20
6
.2.10. Compound 15. Procedure as for 1. Pale yellow
1
f
viscous oil, 90%. H NMR (400 MHz, CD CN): dZ7.89 (t,
3
1
4
H, JZ7.6 Hz), 7.57 (s, 2H), 7.54 (m, 4H), 7.42 (m, 4H),
.62 (s, 4H), 3.49–3.39 (m, 8H), 3.23 (s, 6H). C NMR
6
1
.2.7. Compound 12. 4-Bromo-3,5-dimethylanisole (0.25 g,
.17 mmol, 2.00 equiv) was cooled to K78 8C in THF
13
(100 MHz, CDCl ): dZ158.28, 140.10, 136.50, 135.99,
3
(
2
5 mL). n-BuLi (1.22 mL 1.2 M in hexanes, 1.46 mmol,
.5 equiv) was added dropwise and stirred at K78 8C for
1
5
29.92, 129.00, 128.39, 127.59, 122.23, 71.86, 71.11, 69.41,
8.96. IR (neat): nZ2895, 1731, 1579, 1442, 1107. HRMS
1
5 min. ZnCl (0.40 g, 2.92 mmol, 5.00 equiv) was trans-
2
C
(EI): calcd for C H NO [M ] 407.2097; found 407.2086.
25 29 4
ferred via cannula into the reaction mixture as a THF
solution (10 mL), and the reaction was stirred for an
additional 15 min. The reaction mixture was transferred via
cannula into a flask containing 2,6-dichloroisonicotinic acid
methyl ester (0.120 g, 0.58 mmol, 1 equiv) and Pd(PPh3)4
(0.03 g, 0.03 mmol, 0.05 equiv). The reaction mixture was
degassed then refluxed for 16 h. The reaction was allowed to
cool to room temperature, H O (25 mL) was added, and the
2
mixture was extracted with EtOAc (3!25 mL). The
combined organic fractions were washed with brine (1!
TLC (R )Z0.30 (30% acetone/hexanes).
f
6
.2.11. Compound 16. DiBAlH (5.43 mL, 1.0 M in
hexanes, 5.43 mmol, 3.00 equiv) was added dropwise to a
solution of methyl ester 10 (0.656 g, 1.81 mmol, 1 equiv) in
THF (10 mL) at K78 8C. After stirring for 5 min at K78 8C,
the reaction mixture was warmed to room temperature and
quenched by the addition of methanol (w2.5 mL), followed
by H O (40 mL). After stirring for 10 min, the reaction was
2
extracted with EtOAc (3!40 mL) and the combined
organic extracts washed with saturated NaCl (1!25 mL).
The organic phase was dried over Na SO and was
25 mL), dried over Na SO , filtered, and concentrated under
2 4
vacuum. Purification by column chromatography (20%
2
4
EtOAc/hexanes) produced 0.02 g (0.06 mmol, 10%) of 12
as a yellow solid, mp 189 8C. R Z0.20 (20% EtOAc/
concentrated. Purification by flash column chromatography
30% EtOAc/hexanes) provided 0.56 g (1.61 mmol, 89%)
of 16 as a white solid, mp 149–151 8C. H NMR (CDCl ):
f
(
hexanes). IR (KBr): nZ2960, 1732, 1607, 1439, 1396,
1
K1
1
3
1
7
2
1
2
345, 1316, 1245, and 1157 cm . H NMR (CDCl ): dZ
3
dZ7.34 (dd, 2H, JZ0.8, 8.4 Hz), 7.18 (s, 2H), m (6.77, 4H),
.75 (s, 2H), 6.64 (s, 4H), 3.96 (s, 3H), 3.80 (s, 6H), and
.05 ppm (s, 12H). C NMR (CDCl ): dZ165.7, 160.9,
58.9, 138.1, 137.0, 132.7, 122.1, 112.8, 55.3, 52.8, and
0.7 ppm. HRMS (DEI): calculated for C H NO (MK
H ): 404.186184, found 404.187089.
1
3
4.62 (d, 2H, JZ4.4 Hz), 3.84 (s, 6 h), 2.75 (b, 1H), 2.38 (s,
3
13
6
H). C NMR (CDCl ): dZ159.13, 158.92, 150.25,
3
1
37.27, 133.22, 130.95, 119.01, 115.85, 110.99, 63.55,
2
5
26
4
C
55.25, 21.04. IR (KBr): nZ3214 (b), 2934, 2838, 1606,
C
457, 1300. HRMS (EI): calcd for C H NO [M ]
1
22 23 3
3
hexanes).
49.1678; found 349.1665. TLC (R )Z0.15 (30% EtOAc/
f
6.2.8. Compound 13. Procedure as for 1 except 22 was used
as the boronic acid, P Bu in THF as the phosphine and CsF
t
3
1
as the base. Colorless oil, 52%. H NMR (400 MHz,
CDCl ): dZ7.99 (s, 2H), 7.54 (d, 2H, JZ8.4 Hz), 7.20 (d,
6.2.12. Compound 17. DMSO (0.25 mL, 3.50 mmol,
2.40 equiv) was added to oxalyl chloride (0.25 mL,
2.90 mmol, 2.00 equiv) in CH Cl (15 mL) at K78 8C.
3
2
H, JZ2.8 Hz), 6.93 (dd, 2H, JZ2.8, 8.4 Hz), 4.72 (s, 4H),
3
.98 (s, 3H), 3.87 (s, 6H), 3.57–3.54 (m, 8H), 3.35 (s, 6H).
C NMR (100 MHz, CDCl ): dZ165.83, 160.12, 159.22,
42.50, 137.73, 131.55, 131.42, 120.07, 114.26, 113.45,
2
2
1
3
3
After stirring for 5 min, alcohol 16 (0.51 g, 1.45 mmol,
1 equiv) in CH Cl (2.5 mL) was added via syringe and
1
7
1
2
2
1.85, 71.40, 69.61, 59.17, 55.42, 52.73. IR (neat): nZ2890,
737 (s), 1606, 1248, 1099. HRMS (MALDI): calcd for
the reaction stirred for 5 min at K78 8C. Triethylamine
(1.01 mL, 7.25 mmol, 5.00 equiv) was then added and the
reaction warmed to room temperature. Upon warming, the
solution was diluted with CH Cl (50 mL) and washed with
C
C H NO [(MKH) ] 548.2255; found 548.2271. TLC
8
2
9
36
(
Rf)Z0.20 (45% EtOAc/hexanes).
2
2
H O (2!50 mL) and saturated NaCl (1!25 mL). The
2
6
0
(
.2.9. Compound 14. Carboxylic acid 19, (0.054 g,
.139 mmol, 1 equiv), N,N -diisopropyl-N-ethylamine
organic phase was dried over Na SO and was concentrated.
4
2
0
0.048 mL, 0.278 mmol, 2.00 equiv), and 3-(diethoxyphos-
Purification by flash column chromatography (25% EtOAc/
hexanes) provided 0.474 g (1.37 mmol, 94%) of a viscous
22
1
phoryloxy)-1,2,3-benzotriazin-4(3H)-one(DEPBT) (0.083 g,
.278 mmol, 2.00 equiv) was suspended in THF (2.5 mL).
After stirring for 5 min, benzylamine (0.018 mL,
.167 mmol, 1.20 equiv) was added and the resulting
yellow oil. H NMR (CDCl ): dZ10.15 (s, 1H), 7.71 (s,
3
0
2H), 7.46 (d, 2H, JZ9.2 Hz), 6.85 (m, 4H), 3.86 (s, 6H),
1
3
2.45 (s, 6H). C NMR (CDCl ): dZ191.77, 160.52, 159.61,
3
0
142.10, 137.47, 132.21, 131.18, 119.91, 116.18, 111.32,
55.31, 21.20. IR (neat): nZ2969, 2838, 2733, 1711 (s),
1606, 1239, 1178, 1047. HRMS (EI): calcd for C H NO
3
solution was stirred for 3 h at 23 8C. The solution was
diluted with CH Cl (50 mL) and washed with H O (2!
2
2
2
22 20

A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
11085
C
(MKH) ] 346.1443; found 346.1444. TLC (R )Z0.47
[
3.00 equiv) in THF (50 mL) at 0 8C. After stirring for
10 min, benzyl bromide 20 (2.00 g, 7.14 mmol, 1 equiv) in
THF (3.0 mL) was added via syringe. The reaction mixture
was warmed to RT and stirred for 18 h. The solvent was
removed on the rotary evaporator and the residue dissolved
in diethyl ether (50 mL), which was subsequently washed
f
(
30% EtOAc/hexanes).
6
1
0
.2.13. Compound 18. Aldehyde 19 (0.046 g, 0.132 mmol,
equiv) and triethylphosphonoacetate (0.033 mL,
.165 mmol, 1.25 equiv) was suspended in THF (5 mL)
t
and cooled to 0 8C. KO Bu (0.165 mL, 1.0 M in THF,
with H
2
O (2!50 mL) and saturated NaCl (1!25 mL). The
SO and was concentrated.
Purification by flash column chromatography (25% EtOAc/
0.165 mmol, 1.25 equiv) was added via syringe and the
reaction stirred for 10 min at 0 8C. Upon warming to room
organic phase was dried over Na
2
4
temperature, the solution was diluted with EtOAc (30 mL)
and washed with H O (2!25 mL) and saturated NaCl (1!
hexanes) provided 1.56 g (5.67 mmol, 80%) of a clear oil.
1
H NMR (400 MHz, CDCl ): dZ7.40 (d, 1H, JZ8.8 Hz),
2
3
1
5 mL). The organic phase was dried over Na SO and was
2
7.10 (d, 1H, JZ3.2 Hz), 6.70 (dd, 1H, JZ8.8, 3.2 Hz), 4.60
(s, 2H), 3.80 (s, 3H), 3.71 (m, 2H), 3.62 (m, 2H), 3.42 (s,
4
concentrated. Purification by flash column chromatography
25% EtOAc/hexanes) provided 0.055 g (0.132 mmol,
1
3
(
1
3H). C NMR (100 MHz, CDCl ): dZ158.79, 138.30,
3
1
00%) of a highly viscous green/yellow oil. H NMR
132.72, 114.56, 113.92, 112.40, 72.24, 71.79, 69.92, 59.08,
55.41. IR (neat): nZ2890, 1580, 1475, 1291, 1108. HRMS
(CDCl ): dZ7.69 (d, 1H, JZ15.6 Hz), 7.41 (d, 2H, JZ9.2 Hz),
3
C
7
.38 (s, 2H), 6.83 (m, 4H), 6.64 (d, 1H, JZ16.0 Hz), 4.30
(MALDI): calcd for C11
found 297.0094. TLC (R )Z0.25 (10% EtOAc/hexanes).
f
H O [(MKNa )] 297.0097;
15 3
(
q, 2H, JZ6.8 Hz), 3.85 (s, 6H), 2.46 (s, 6H), 1.37 (t, 3H,
1
3
JZ6.8 Hz). C NMR (CDCl ): dZ165.97, 159.82, 159.36,
3
1
1
42.22, 141.73, 137.33, 132.80, 131.00, 122.27, 119.61,
16.02, 111.16, 60.92, 55.28, 21.11, 14.39. IR (neat):
6.2.17. Compound 22. n-BuLi (5.40 mL, 1.6 M solution in
hexanes, 8.60 mmol, 1.20 equiv) was added dropwise to a
solution of aryl bromide 21 (1.96 g, 7.14 mmol, 1 equiv) in
THF (50 mL) at K78 8C, and the resulting solution stirred
for 5 min. Triisopropyl borate (2.16 mL, 9.30 mmol,
nZ2960, 2838, 1711 (s), 1606, 1239, 1178, 1047. HRMS
C
MALDI): calcd for C H NO [(MKH) ] 418.2018;
(
26 28 4
found 418.2024. TLC (R )Z0.37 (25% EtOAc/hexanes).
f
1.30 equiv) was added via syringe and the flask warmed
6
1
.2.14. Compound 19. Ester 18 (0.115 g, 0.275 mmol,
equiv) was suspended in THF (4 mL), LiOH$2H O
to RT. The majority (w75%) of the solvent was removed
with the rotary evaporator and 1 M HCl (50 mL) was added
and the resulting solution was stirred for 4 h. The mixture
was extracted with ether (3!33 mL) and the combined
organic fractions were extracted with 1 M NaOH (3!
15 mL). The aqueous extracts were combined and neutral-
ized with HCl (conc.) until acidic to litmus. The aqueous
layer was then extracted with EtOAc (3!33 mL) and the
2
(0.017 g, in 1.0 mL of H O, 0.413 mmol, 1.50 equiv) was
2
added and the reaction stirred for 30 min at 23 8C. The
reaction mixture was neutralized by the addition of 1 M HCl
(0.413 mL, 0.413 mmol, 1.50 equiv) and extracted with
CH Cl (3!25 mL). The organic phase was dried over
2
2
Na SO and was concentrated to afford 0.091 g
2
4
1
0.234 mmol, 85%) of a pure yellow solid. H NMR
(
combined organic phases were dried over Na SO and
2 4
(
CDCl ): dZ7.78 (d, 1H, JZ16.4 Hz), 7.42 (d, 2H, JZ9.2 Hz),
concentrated. This produced a colorless oil (1.20 g,
5.00 mmol, 70%) that hardened upon standing. This
material was used in subsequent reactions without further
3
7
6
1
1
2
.39 (s, 2H), 6.83 (m, 4H), 6.65 (d, 1H, JZ16.0 Hz), 3.85 (s,
1
3
H), 2.46 (s, 6H). C NMR (CDCl ): dZ170.58, 159.96,
3
1
59.48, 144.43, 141.42, 137.39, 132.61, 131.03, 121.51,
19.85, 116.08, 111.23, 55.32, 21.10. IR (KBr): nZ2934,
838, 2593, 1711, 1615, 1291, 1248, 1047. HRMS
purification. H NMR (400 MHz, CDCl
3
): dZ7.88 (d, 1H,
JZ8.4 Hz), 6.89 (dd, 1H, JZ8.4, 2.4 Hz), 6.81 (d, 1H,
JZ2.4 Hz), 6.27 (b, 2H), 4.63 (s, 2H), 3.83 (s, 3H), 3.65 (m,
2H), 3.55 (m, 2H), 3.37 (s, 3H). IR (KBr): nZ3362 (b),
2917, 1606, 1370, 1265, 1073.
C
(MALDI): calcd for C H NO [(MKH) ] 390.1705;
24 24 4
found 390.1684.
6
2
2
0
.2.15. Compound 20. 2-Bromo-5-methoxytoluene (5.00 g,
6.2.18. Compound 23. Procedure as for 1 except 22 was
used as the boronic acid, P Bu in THF as the phosphine and
t
4.00 mmol, 1 equiv), N-bromosuccinimide (5.13 g,
8.8 mmol, 1.20 equiv), and AIBN (0.08 g, 0.48 mmol,
.02 equiv) was suspended in benzene (100 mL). The
3
1
CsF as the base. Colorless oil, 52%. H NMR (400 MHz,
CDCl ): dZ7.99 (s, 2H), 7.54 (d, 2H, JZ8.4 Hz), 7.20 (d,
3
reaction mixture was heated to reflux for 1 h and then
cooled to 23 8C. The resulting mixture was filtered and the
filtrate was diluted with diethyl ether (75 mL) and washed
2H, JZ2.8 Hz), 6.93 (dd, 2H, JZ2.8, 8.4 Hz), 4.72 (s, 4H),
3.98 (s, 3H), 3.87 (s, 6H), 3.57–3.54 (m, 8H), 3.35 (s, 6H).
1
3
C NMR (100 MHz, CDCl ): dZ165.83, 160.12, 159.22,
3
with H O (2!100 mL) and saturated NaCl (1!50 mL).
2
142.50, 137.73, 131.55, 131.42, 120.07, 114.26, 113.45,
71.85, 71.40, 69.61, 59.17, 55.42, 52.73. IR (neat): nZ2890,
1737 (s), 1606, 1248, 1099. HRMS (MALDI): calcd for
The organic phase was dried over Na SO and was
2
4
concentrated. The crude solid was triturated with diethyl
ether/hexanes (1:4) and filtered to yield 5.50 g (19.60 mmol,
C
C H NO [(MCH) ] 548.2255; found 548.2271. TLC
2
9
36
8
1
6%) of a white solid, mp 80–82 8C. H NMR (400 MHz,
8
(R )Z0.20 (45% EtOAc/hexanes).
f
CDCl ): dZ7.45 (d, 1H, JZ8.8 Hz), 6.99 (d, 1H, JZ3.2 Hz),
3
6
.74 (dd, 1H JZ2.8, 8.8 Hz), 4.55 (s, 2H), 3.80 (s, 3H). IR
6.2.19. Compound 23. Procedure as for 16. Colorless oil,
1
(
(
KBr): nZ2943, 2847, 1483, 1291, 1256, 1012, 811. TLC
Rf)Z0.58 (10% EtOAc/hexanes).
99%. H NMR (400 MHz, CDCl ):dZ7.57 (d, 2H, JZ8.4 Hz),
3
7.55 (s, 2H), 7.12 (d, 2H, JZ2.4 Hz), 6.93 (dd, 2H, JZ2.4,
8
.4 Hz), 4.72 (d, 2H, JZ6.8 Hz), 4.61 (s, 4H), 3.91 (b, 1H),
13
3.86 (s, 6H), 3.65–3.60 (m, 8H), 3.40 (s, 6H). C NMR
6
.2.16. Compound 21. 2-Methoxyethanol (1.13 mL,
4.30 mmol, 2.00 equiv) was added dropwise to a solution
1
of NaH (0.8, 60% w/w in mineral oil, 21.40 mmol,
(100 MHz, CDCl ): dZ159.53, 157.65, 150.80, 136.88,
3
133.21, 131.36, 119.64, 114.47, 113.59, 72.08, 71.60, 69.60,

11086
A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
6
1
3.99, 59.12, 55.41. IR (neat): nZ3441 (b), 2908, 1606,
431, 1239, 1082, 872. HRMS (MALDI): calcd for
2. For our previous work on pyridine derived fluorescent
chemosensors, see: (a) Mello, J. V.; Finney, N. S. Angew.
Chem. Int. Ed. 2001, 40, 1536–1538. (b) Mello, J. V.; Finney,
N. S. Org. Lett. 2001, 3, 4263–4265. (c) Fang, A. G.; Mello,
J. V.; Finney, N. S. Org. Lett. 2003, 5, 967–970.
C
C H NO [(MKH) ] 498.2486; found 498.2475. TLC
2
8
36
7
(
Rf)Z0.19 (60% THF/hexanes).
2C
3
. For our previous work on hydrocarbon and Ru(II)bpy
3
6
.2.20. Compound 24. Procedure as for 17. Colorless oil,
1
derived chemosensors, see: (a) McFarland, S. A.; Finney, N. S.
J. Am. Chem. Soc. 2001, 123, 1260–1261. (b) McFarland,
S. A.; Finney, N. S. J. Am. Chem. Soc. 2002, 124, 1178–1179.
(c) McFarland, S. A.; Finney, N. S. Chem. Commun. 2003,
388–389.
9
0%. H NMR (400 MHz, CDCl ): dZ10.18 (s, 1H), 7.93
3
(
s, 2H), 7.58 (d, 2H, JZ8.4 Hz), 7.18 (d, 2H, JZ2.8 Hz),
6
3
.95 (dd, 2H, JZ2.8, 8.4 Hz), 4.69 (s, 4H), 3.88 (s, 6H),
13
.61–3.54 (m, 8H), 3.36 (s, 6H). C NMR (100 MHz,
CDCl ): dZ192.30, 159.98, 159.32, 142.51, 137.62,
3
1
6
1
31.64, 131.39, 120.06, 114.26, 113.48, 71.89, 71.36,
9.63, 59.08, 55.45. IR (neat): nZ2908, 1711 (s), 1615,
553, 1248, 1099, 889. HRMS (MALDI): calcd for
4. For an early example of the use of conformational restriction
of a binaphthyl to provide a fluorescent response to binding,
see: Takeuchi, M.; Yoda, S.; Imada, T.; Shinkai, S.
Tetrahedron 1997, 53, 8335–8348. Note that in this case the
C
C H NO [(MKH) ] 496.2330; found 496.2318. TLC
2
8
34
7
(
Rf)Z0.33 (35% EtOAc/hexanes).
maximum reported I/I was !1.3.
0
5
. For recent representative examples, see: (a) Guo, X.; Qian, X.;
Jia, L. J. Am. Chem. Soc. 2004, 126, 2272–2273. (b) Nolan,
E. M.; Lippard, S. J. J. Am. Chem. Soc. 2003, 125,
14270–14271. (c) Walkup, G. K.; Burdette, S. C.; Lippard,
S. J.; Tsien, R. Y. J. Am. Chem. Soc. 2000, 122, 5644–5645.
6
.2.21. Compound 25. A suspension of methyl triphenyl-
phosphonium iodide (0.175 g, 0.433 mmol, 2.90 equiv) in
THF (5 mL) was cooled to 0 8C. n-BuLi (0.271 mL, 1.60 M
in hexanes, 0.433 mmol, 2.90 equiv) was added via syringe
and the mixture stirred for 30 min. Aldehyde 24 (0.075 g,
6. For representative examples, see: (a) Ueno, A.; Ikeda, A.;
Ikeda, H.; Ikeda, T.; Toda, F. J. Org. Chem. 1999, 64,
382–387. (b) Metzger, A.; Anslyn, E. V. Angew. Chem. Int.
Ed. 1998, 37, 649–652. (c) Corradini, R.; Dossena, A.;
Galaverna, G.; Marchelli, R.; Panagia, A.; Sartor, G. J. Org.
Chem. 1997, 62, 6283–6289. (d) Hamasaki, K.; Usui, S.;
Ikeda, H.; Ikeda, T.; Ueno, A. Supramol. Chem. 1997, 8,
125–135.
0
.151 mmol, 1 equiv) in THF (1 mL) was added via syringe
and the reaction stirred for an additional 30 min. The
reaction was warmed to RT and stirred for 5 h. The solution
was diluted with CH Cl (20 mL) and washed with H O
2
2
2
(
2!25 mL) and saturated NaCl (1!10 mL). The organic
phase was dried over Na SO and was concentrated.
2
4
Purification by flash column chromatography (45%
EtOAc/hexanes) afforded 0.045 g (0.091 mmol, 60%) of
7. For representative examples of protein/macromolecule FRET
based chemosensors, see: (a) Truong, K.; Sawano, A.; Mizuno,
H.; Hama, H.; Tong, K. I.; Mal, T. K.; Miyawaki, A.; Ikura, M.
Nat. Struct. Biol. 2001, 8, 1069–1073. (b) Miyawaki, A.;
Llopis, J.; Heim, R.; McCaffery, J. M.; Adams, J. A.; Ikura,
M.; Tsien, R. Y. Nature 1997, 388, 882. (c) Adams, S. R.;
Harootunian, A. T.; Buechler, Y. J.; Taylor, S. S.; Tsien,
R. Y. Nature 1991, 349, 694. Note that there are numerous
FRET based assays for enzymatic activity, etc. that are of
great value but do not meet the definition of a chemosensor
1
5 as a pale yellow viscous oil. H NMR (400 MHz,
2
CDCl ): dZ7.50 (d, 2H, JZ8.4 Hz), 7.44 (s, 2H), 7.18 (d,
3
2
H, JZ2.4 Hz), 6.91 (dd, 2H, JZ2.4, 8.4 Hz), 6.77 (dd, 1H,
JZ11.2, 17.6 Hz), 6.06 (d, 1H, JZ17.6 Hz), 5.51 (d, 1H,
JZ11.2 Hz), 4.71 (s, 4H), 3.86 (s, 6H), 3.60–3.52 (m, 8H),
1
3
3
1
1
.36 (s, 6H). C NMR (100 MHz, CDCl ): dZ159.55,
3
58.17, 145.12, 137.51, 135.06, 132.58, 131.10, 118.79,
18.31, 113.62, 113.21, 71.89, 71.19, 69.50, 59.05, 55.39.
IR (neat): nZ2882, 1597, 1501, 1248, 1099, 1038, 881, 811.
C
HRMS (MALDI): calcd for C H NO [(MKH) ]
2
9
36
6
(
see Ref. 1h) .
4
hexanes).
94.2537; found 494.2560. TLC (R )Z0.20 (35% EtOAc/
f
8
. For representative examples of small molecule FRET based
chemosensors, see: (a) Arduini, M.; Felluga, F.; Mancin, F.;
Rossi, P.; Tecilla, P.; Tonellato, U.; Valentinuzzi, N. Chem.
Commun. 2003, 1606–1607. (b) Schneider, S. E.; O’Neil,
S. N.; Anslyn, E. V. J. Am. Chem. Soc. 2000, 122, 542–543.
(
c) Pearce, D. A.; Walkup, G. K.; Imperiali, B. Biorg. Med.
References and notes
Chem. Lett. 1998, 8, 1963–1968. (d) Godwin, H. A.; Berg, J. M.
J. Am. Chem. Soc. 1996, 118, 6514–6515.
1
. For representative reviews of the chemosensor field, see:
a) Rurack, K.; Resch-Genger, U. Chem. Soc. Rev. 2002, 31,
16–127. (b) Lavigne, J. J.; Anslyn, E. V. Angew. Chem. Int.
9
. For relevant reviews of Pd-catalyzed cross coupling reactions,
see: (a) Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95,
(
1
2
457–2483. (b) Negishi, E.-i. In Metal-Catalyzed Cross
Ed. 2001, 40, 3118–3130. (c) Valeur, B.; Leray, I. Coord.
Chem. Rev. 2000, 205, 3–40. (d) de Silva, A. P.; Eilers, J.;
Zlokarnik, G. Proc. Natl. Acad. Sci. U.S.A. 1999, 96,
Coupling Reactions. Diederich, F. Stang, P. J., Eds.; Wiley:
New York, 1998. Chapter 1. (c) Suzuki, A. J. Organomet.
Chem. 1999, 576, 147–168.
8336–8337. (e) Snowden, T. S.; Anslyn, E. V. Curr. Opin.
1
0. While not always required, we find the use of P(t-Bu)
3
Chem. Biol. 1999, 3, 740–746. (f) de Silva, A. P.; Gunaratne,
H. Q. N.; Gunnlaugsson, T.; Huxley, A. J. M.; McCoy, C. P.;
Rademacher, J. T.; Rice, T. E. Chem. Rev. 1997, 97,
typically provides the best yields for Suzuki coupling to 2,6-
dichloropyridine derivatives. For early reports of the utility of
this phosphine, see: (a) Old, D. W.; Wolfe, J. P.; Buchwald,
S. L. J. Am. Chem. Soc. 1998, 120, 9722–9723. (b) Littke,
A. F.; Fu, G. C. Angew. Chem. Int. Ed. 1997, 37, 3387–3388.
1
515–1566. (g) Fabbrizzi, L.; Licchelli, M.; Pallavicini, P.;
Sacchi, D.; Taglietti, A. Analyst 1996, 121, 1763–1768.
h) Fluorosensors for Ion and Molecule Recognition. Czarnik,
A. W., Ed.; American Chemical Society: Washington, DC,
994.
(
For the more recent use of this phosphine as its HBF
4
salt, see:
1
Netherton, M. R.; Fu, G. C. Org. Lett. 2001, 3, 4295–4298.

A. G. Fang et al. / Tetrahedron 60 (2004) 11075–11087
11087
1
1
1. Berlman, I. Handbook of Fluorescence Spectra of Aromatic
Molecules; Academic: New York, 1971.
either state, still has a large 3). However, the classical model
for oscillator strength equates the transition dipole to the
polarizability of a molecule in an applied electric field, which
qualitatively relates the ground and excited state dipole
moments. As we are comparing molecules of very similar
structure, we believe that our argument is at least qualitatively
valid. See: Turro, N. Modern Molecular Photochemistry;
University Science Books: Sausilito, CA, 1991; pp 81–90.
6. The UV spectra suggest that the Einstein coefficient for
absorption, and thus the rate of radiative decay, are unchanged.
See: Calvert, J. G.; Pitts, J. N. Photochemistry; Wiley: New
York, 1996; pp 173–174.
2. For the empirical correlation between biaryl dihedral angle and
nonradiative decay, see: (a) Fujii, T.; Suzuki, S.; Komatsu, S.
Chem. Phys. Lett. 1978, 57, 175–178. (b) Fujii, T.; Suzuki, S.;
Komatsu, S. Bull. Chem. Soc. Jpn 1982, 55, 2516–2520 and
references therein.
1
1
3. For general treatments of nonradiative decay, see: (a) Freed,
K. F. Acc. Chem. Res. 1978, 11, 74–80. (b) Klessinger, M.;
Michl, J. Excited States and Photochemistry of Organic
Molecules; VCH: New York, 1995; pp 252–260.
1
4. For select early derivations of the inverse exponential
relationship between energy gap and relaxation rate, see:
1
1
7. (a) Rettig, W. Angew. Chem. Int. Ed. 1986, 25, 971–988.
(
(
a) Engleman, R.; Jortner, J. Mol. Phys. 1970, 18, 145.
b) Freed, K. F.; Jortner, J. J. Chem. Phys. 1970, 52, 6272–
(
b) Bhattachryya, K.; Chowdhury, M. Chem. Rev. 1993, 93,
07–535.
5
6291. (c) Siebrand, W. J. Chem. Phys. 1967, 46, 440–447. This
can also be regarded as a restatement of Ermolaev’s Rule:
8. For recent work on the TICT states of simple pyridine
derivatives, see: Dobkowski, J.; Wojcik, J.; Kozminski, W.;
Kolos, R.; Waluk, J.; Michl, J. J. Am. Chem. Soc. 2002, 124,
2406–2407 and references therein.
Ermolaev, V. L. Sov. Phys. Usp. 1963, 80, 333–358. Usp. Fiz.
3 KaDE
1
Nauk 1963, 80, 3–40. More quantitatively, kICz10 e
z
1
3
10 f (where a is a proportionality constant and fn is the
n
Franck–Condon factor for the transition in question). For
1
9. While m-alkoxy groups are generally considered electron
withdrawing due to s polarization, this generalization is based
on ground state phenomena such as electrophilic aromatic
energy gaps (DE) of w50 kcal/mol, Franck–Condon factors
5
8 K1
are w10 , leading to kICw(10 s . This is slow relative to
typical rates of fluorescence emission and ISC, leading to the
qualitative DEz50–60 kcal/mol cutoff for ignoring IC. See:
Turro, N. Modern Molecular Photochemistry; University
Science Books: Sausilito, CA, 1991; pp 180–185.
*
substitution. The fact that the excited states are dominantly p
in nature apparently leads to the m-alkoxy groups being
moderately electron donating.
2
2
0. Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 2nd
ed.; Kluwer Academic: New York, 1999.
1
5. The assertion that the transition dipole should correlate with
ground or excited state dipole moment is incorrect at the
quantum mechanical level, and inconsistent with some
observations (for instance, hexatriene, which has centrosym-
metric ground and excited states and thus no net dipole in
1. Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen,
R. K.; Timmers, F. J. Organometallics 1996, 15, 1518–1520.
22. Li, H.; Jiang, X.; Ye, Y. H.; Fan, C.; Romoff, T.; Goodman, M.
Org. Lett. 1999, 1, 91–93.