A Fluorescent Self-Amplifying Wavelength-Responsive Sensory Polymer for Fluoride Ions

Article abstract of DOI:10.1002/anie.200352075

A glowing response: Novel fluoride-selective sensors, utilizing fluoride-induced lactonization to highly fluorescent coumarin derivatives, are reported. The incorporation of this sensory system into a conjugated polymer amplifies the fluorescence response

Full text of DOI:10.1002/anie.200352075

Angewandte
Chemie
Ion Sensors
herein a fluoride-sensing system in which a fluoride triggered
À
Si O bond cleavage results in the formation of a highly
A Fluorescent Self-Amplifying Wavelength-
Responsive Sensory Polymer for Fluoride Ions**
fluorescent coumarin [Eq. (1)] and show how the intercon-
nection of this transduction element to a conjugated polymer
produces a 100-fold sensitivity enhancement.
Tae-Hyun Kim and Timothy M. Swager*
Fluoride, the smallest anion, has unique chemical properties
and its recognition and detection are of growing interest
because it is associated with nerve gases and the refinement of
uranium used in nuclear weapons manufacture. As a result,
there is a need to develop new selective and sensitive methods
for fluoride detection in environments that are not easily
served by conventional ion-selective electrodes. Most of the
emerging detection methods have focused on receptors that
differentiate fluoride from other anions, such as chloride,
phosphate, or nitrate and convert the binding into an electro-
chemical or optical response.[1] These efforts have enjoyed
success and in some cases have resulted in the ability to
visually detect fluoride ions.[2–4]
An approach to greater sensitivity in fluorescence detec-
tion schemes is to use the exciton-transporting properties of
semiconducting polymers to amplify perturbations (trans-
duction events) produced by analyte binding. These methods
were originally demonstrated using electron-transfer quen-
ching[5a] and variations on this scheme have been used to
create a number of chemical and biological sensory methods,
most of which involve changes in the emission intensities.[5] In
spite of these successes, it is generally preferable for a
fluorescence sensory signal to involve a new emission at a
different wavelength rather than a modulation of an existing
signal. The advantage of the latter approach is that a new
signal is inherently easier to detect than an intensity change.
Furthermore, nonspecific chemical interactions are more
likely to produce intensity changes than new signals, hence
we can also expect specificity gains in properly designed
sensors that create new emissions. Fluorescence resonance
energy transfer (FRET) is broadly used to produce new
emissive signals and the amplification of these processes
through intramolecular energy transfer from conjugated
polymers has been demonstrated.[6] New analyte-triggered
emissions, if optimally deployed, offer high specificity and low
background. In consideration of how to best construct such a
system for the selective detection of fluoride ions we have
decided to explore schemes using the unique chemical
reactivity of fluoride ions with silicon,[7] rather than elaborate
receptors. In a fusion of these different elements, we report
Our decision to focus on the fluoride-ion-triggered
formation of coumarins, was based upon the fact that
coumarins are laser dyes with high radiative quantum
yields.[8] To this end we synthesized a series of coumarin
derivatives with varying electron-donating groups in the 7-
position of the coumarin to tune the absorption and emission
wavelengths as shown (compounds 1a–3a in Table 1). The
Table 1: Photophysical data in dichloromethane.
Compound
Abs. max.[a] (loge)
Emission max.[a]
F [%][b]
1
2
3
4
1a
2a
3a
4a
322 (3.66)
386(4.31)
–
<0.01
0.2
0.8
12
12
81
450
489
482
415
450
529
517
384 (4.43)
378 (4.76)[c]
334 (3.71)
417 (4.39)
412 (4.64)
396(4.80) [c]
37
30
[a] Given in nanometers. [b] Determined with quinine sulfate as a
standard. [c] Per monomer unit.
synthesis of the indicators involved a two-step sequence,
namely silylation of the corresponding 2-hydroxy benzalde-
hyde, followed by a condensation reaction under the modified
Knoevenagel conditions.[9] The tert-butyldimethylsilyl (TBS)
group was chosen as the protecting group for the phenols to
promote selectivity to fluoride by rendering indicator 2 stable
(unreactive) to likely interferents. Indeed 2 is stable in THF
solutions containing an excess of chloride ions (tetrabutyl-
ammonium salt). To avoid a possible problematic trans–cis
isomerization of the double bond, we used diethyl malonate
to introduce symmetrically substituted double bonds in 1,2 ,
and 3.
The cyclization reaction is irreversible and produces a
time-dependent dosimetric response that is controlled by the
reaction kinetics. The rate of the cyclization of 2 to 2a in THF
solution was monitored by fluorescence and found to be first
order in 2, independent of fluoride (tetrabutylammonium salt,
TBAF), and relatively slow with a rate constant of k = 2 ”
10À4 sÀ1. The rate constants do not change for the ranges from
[*] Prof. Dr. T. M. Swager, Dr. T.-H. Kim
Department of Chemistry
Massachusetts Institute of Technology
Cambridge, MA 02139 (USA)
Fax: (+1)617-253-7929
E-mail: tswager@mit.edu
[**] We are grateful for research funding from the Department of Energy,
Nuclear Nonproliferation Program.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2003, 42, 4803 –4806
DOI: 10.1002/anie.200352075
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4803

Communications
0.5 to 3 molar ratio of fluoride ion relative to 2. Hence it
appears that the silyl cleavage (or fluoride association with
the silyl) is fast and that the cyclization is the slow step. With a
higher 10-fold excess of fluoride ion there is an enhancement
in the rate, and a saturation of the emission intensity is
observed in approximately 2 h (Figure 1a). The rate accel-
eration at high fluoride concentrations suggests additional
ionic associations with the tetrabutylammonium fluoride
(TBAF) salt. The non-innocence of excess fluoride is further
revealed by additional peaks in absorbance spectra of 2a
(Figure 1a inset) that are present at high TBAF concentration
but absent after an aqueous extraction.
To impart greater sensitivity to our fluoride-induced-
lactonization detection scheme, we designed semiconductor-
polymer–indicator conjugates that could amplify the
response. Amplification of a turn-on sensor can in principle
be performed by bringing a chromophore of the proper
spectral characteristics near to the polymer to achieve FRET.
Indeed our purpose in synthesizing 3a which has a longer-
wavelength emission was to enable this type of detection
scheme. However, in spite of our chromophore engineering,
the cyclized coumarins lacked the proper energy matching
with poly(phenylene ethynylene)s (PPEs), whose optical and
electronic properties are well suited for this fluorescent
sensory application.[5]
To overcome the constraints of FRET we pursued a
potentially more powerful approach wherein the electronic
character of the indicator is strongly electronically coupled to
the band structure of the polymer. This approach enables an
additional energy-transfer mechanism (known as Dexter
energy transfer for small molecules) wherein activation of
an indicator locally perturbs the band gap of the polymer.
This approach is further attractive because detailed fluores-
cence-depolarization studies have shown that the strong
electronic coupling (band structure) in PPEs contributes
Figure 1. Emission changes of indicator 2 (a) and sensory polymer
4 (b) upon addition of TBAF. The insets show the respective absorp-
tion spectra of the initial solution (0 min, a) and after exposure to
TBAF (2 h, c). The second inset (top right) is the absorption spec-
trum of 2a after aqueous extraction, which removes absorptions that
appear to be the result of interactions with fluoride ion. Experimental
conditions: 298 K in all cases. a) 2, 3.2”10À6 m in THF and TBAF,
33”10À6 m; b) 4, 1.5”10À6 m in CH2Cl2 (repeating group concentra-
tion) and TBAF, 1.6”10À7 m. For the polymer 4 lmax =378 nm,
e=57600mÀ1 cmÀ1; emission lmax =482 nm and 4a lmax =396nm,
e=63400mÀ1 cmÀ1; emission lmax =517 nm.
more to energy transfer than Förster energy transfer, the basis
[10]
of FRETschemes.
Hence in
strongly electronically coupled
conjugation schemes a few cycli-
zation events triggered by fluoride
ions will produce local minima in
the band gap that trap mobile
excitons to give a new emission
(Figure 2).[11] The facile transport
of excitons in semiconductive
organic polymers allow these pol-
ymers to sample many potential
transduction sites and thereby
dramatically increase the proba-
bility of emission from a lower-
energy-trapping (cyclized) site.
The result is an amplified turn-on
sensory scheme.
In designing PPE-conjugates
we sought to develop a general
scheme that could be exploited for
future indicator systems. In addi-
tion, band-gap engineering by the
Figure 2. Schematic band diagram illustrating the mechanism by which a semiconductive polymer can
produce an enhancement in a new fluorescence chemosensory response. The horizontal dimension
represents the position along the polymer backbone shown schematically at the bottom. Excitons are
created by absorption of a photon (hn) and then migrate along the backbone. Fluoride-induced lactoni-
zation in the side-chains produces a trapping site with a smaller band gap (Eg) and recombination of
interconnection of pendant chro- excitons at that site results in a new amplified emission.
4804
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 4803 –4806

Angewandte
Chemie
mophores to semiconducting polymers is applicable to other
electrooptical technologies. Our semiconductive-sensory-pol-
ymer designs have focused on the PPE-based material 4
(Scheme 1), in which a thiophene-bridge connects the sensory
o-silyloxybenzylidene malonate moiety with the polymer
backbone.[12] Thiophene is the linkage of choice because its
strong p-electron delocalizing nature will relay electronic
perturbations in the side-chains to the conjugated backbone
and it is synthetically versatile.
Knoevenagel reaction with diethyl malonate to produce the
desired sensory polymer 4.
The effect of coumarin formation on the band gap of
polymer 4 is readily apparent by inspection of the absorption
and emission spectra before and after quantitative conversion
into 4a by treatment with an excess of TBAF. As expected,
both absorbance and emission spectra of the coumarin-
containing form 4a in dichloromethane solution were red-
shifted compared to those of its precursor 4 (Figure 1b).
Moreover, the emission change from blue to blue-green is
readily detected visually. Cyclization also caused a significant
increase in the fluorescence quantum yield of the conjugated
polymer (4: F = 0.12; 4a: F = 0.30). These results indicate
that we can trigger a strong change in photophysical proper-
ties of PPEs using an electronic perturbation from the
appended indicator. The gel permeation chromatography
(GPC) analysis of the polymer 4 and 4a revealed molecular
weights of Mn = 1.06 ” 104 and Mn = 1.05 ” 104, respectively.
Fluorescence studies reveal that our amplification scheme
is highly effective. Kinetic studies for the polymer 4 were not
The synthesis of sensory polymer
4 is shown in
Scheme 1.[13] Bis(trimethylsilylethynyl)phenyl thiophene 7,
obtained from 5 by selective Stille cross-coupling and
subsequent Sonogashira cross-coupling reactions, was deriv-
atized to the corresponding dioxaborolane 8. A further cross-
coupling reaction with the appropriate triflate 9[14] produced
the thiophene-bridged diaryl compound 10. Desilylation of
diethynylenes using TBAF, followed by silylation of phenol
with TIPSCl, gave the monomer 12.[15] A Sonogashira cross-
coupling polymerization provided polymer 14.[16] As a final
step we modified polymer 14 using a titanium-mediated
Scheme 1. Reagents and conditions: a) 2-tributylstannylthiophene, [PdCl2(PPh3)2], DMF, 808C, 12 h, 76%; b) TMSA, [Pd(PPh3)4], CuI, toluene/
DIPA 4:1, 808C, 20 h, 76%; c) tBuLi, THF, À708C, 90 min, then 2-isopropoxy-4,4,5,5-tetramethyl-1,3,2-dioxaborolane, RT, 4 h, 59%; d) [Pd(PPh3)4],
triethylamine, DMF, 1008C, 24 h, 64%; e) TBAF, THF, RT, 1 h; f) TIPSCl, imidazole, DMF, RT, 2 h, 75% over two steps; g) [Pd(PPh3)4], CuI, tolu-
ene/DIPA 4:1, 658C, 24 h, 43%; h) TiCl4 in CCl4 to THF, 08C, then 14 and diethylmalonate in THF, then pyridine in THF, 08C, 1 h and RT, 2 h,
88%; i) TBAF, THF, RT, 3 h, 98%. TMSA=trimethylsilylacetylene, DIPA=diisopropylamine.
Angew. Chem. Int. Ed. 2003, 42, 4803 –4806
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4805

Communications
[2] a) H. Miyaji, J. L. Sessler, Angew. Chem. 2001, 113, 158; Angew.
Chem. Int. Ed. 2001, 40, 154.
[3] a) J. J. Lavigne, E. V. Anslyn, Angew. Chem. 1999, 111, 3903;
Angew. Chem. Int. Ed. 1999, 38, 3666.
[4] S. Yamaguchi, S. Akiyama, K. Tamao, J. Am. Chem. Soc. 2001,
123, 11372.
[5] a) Q. Zhou, T. M. Swager, J. Am. Chem. Soc. 1995, 117, 12593;
b) J.-S. Yang, T. M. Swager, J. Am. Chem. Soc. 1998, 120, 5321;
c) J.-S. Yang, T. M. Swager, J. Am. Chem. Soc. 1998, 120, 11864;
d) T. M. Swager, Acc. Chem. Res. 1998, 31, 201; e) D. T.
McQuade, A. E. Pullen, T. M. Swager, Chem. Rev. 2000, 100,
2537; f) for a recent review, see: T. M. Swager, J. H. Wosnick,
MRS Bull. 2002, 27, 446.
[6] D. T. McQuade, A. H. Hegedus, T. M. Swager,J. Am. Chem. Soc.
2000, 122, 12389.
[7] For another sensor approach that used the affinity of fluoride to
silicon see: S. Yamaguchi, S. Akiyama, K. Tamao, J. Am. Chem.
Soc. 2000, 122, 6793.
possible in the same way as for the small molecule 2, because
the fluorescence quantum yield of the polymer varies as a
function of conversion and at low conversion we observe less
efficient emission. Nevertheless, it is clear that the polymer 4
has a faster response to fluoride ion (Figure 1b). The
enhanced fluoride sensitivity of 4 relative to 2 is clearly
evident when one considers that exposure of 4 (1.5 ” 10À6
m
based upon the repeating group) to fluoride (1.6 ” 10À7 m,
TBAF in THF) results in a complete transformation of the
fluorescence intensity maximum of 4 in 2 h, whereas 2
required a 100-fold higher fluoride concentration to achieve
a complete fluorescence transformation over the same time
with all other conditions the same (Figure 1). Hence we
conclude that 4 is approximately 100-times more sensitive
than its small-molecule-based counterpart 2. This amplifica-
tion factor is in accord with our previous determination of
exciton diffusion in isolated polymers.[5a] We further note that
energy migration within isolated polymers (i.e. dilute solu-
tions) is inherently inefficient owing to the one-dimensional
exciton diffusion.[17] Thin films that display two- and three-
dimensional transport characteristics exhibit more favorable
transport statistics and also benefit from improved Förster
energy-transfer process between polymer chains. Unfortu-
nately 4 is strongly quenched in thin films, however, it should
be possible to generate related indicator materials with
desirable solid-state properties and much larger amplification
factors in the future.
[8] K. D. Drexhage in Dye Lasers, Vol. 1 (Ed.: F. P. Schäfer),
Springer, New York, 1977.
[9] a) W. Lehnert, Tetrahedron Lett. 1970, 4723; b) M. Black, J. I. G.
Cadogan, H. McNab, A. D. MacPherson, V. P. Roddam, C.
Smith, H. R. Swenson, J. Chem. Soc. Perkin Trans. 1 1997, 2483.
[10] A. Rose, C. G. Lugmair, T. M. Swager, J. Am. Chem. Soc. 2001,
123, 11298.
[11] T. M. Swager, C. J. Gil, M. S. Wrighton, J. Phys. Chem. 1995, 99,
4886.
[12] The alternative approach to develop PPE based on the 3,6-
diethynyl-2-hydroxy-benzaldehyde was not pursued, as o-phe-
noxyacetylenes could undergo tandem cyclization to the corre-
sponding benzofuran by use of TBAF or CsF, for example, see:
A. Sakai, T. Aoyama, T. Shioiri,Heterocycles 2000, 52, 643.
[13] See Supporting Information for details.
[14] 2-Hydroxy-4-(trifluoromethanesulfonyl) benzaldehyde (9) was
obtained by treating 4-hydroxy-2-methoxybenzaldehyde with
triflic anhydride in pyridine (95%), followed by demethylation
involving BBr3 in CH2Cl2 at À408C (79%). See also Supporting
Information.
[15] Triisopropylsilyl group (TIPS) was used instead oftert-butyldi-
methylsilyl (TBS) as the thermal polymerization conditions
required a more stable protecting group for phenol.
[16] 1,4-Diiodobenzene with ortho-dialkoxy side-chains 13 was
chosen as a comonomer to produce a sensory polymer with
higher frequencies (blue-shifted). Z. Zhu, T. M. Swager, Org.
Lett. 2001, 3, 3471.
In summary, we have introduced a new system for the
detection of fluoride ion and successfully amplified the
response using exciton migration in a semiconducting organic
polymer. In contrast to most semiconductive-polymer sensor
schemes that rely on changes in emission intensity, this
sensory system utilizes a new fluorescence signal. The
approach of directly interconnecting the indicator electroni-
cally with the polymer's band structure is a promising
alternative to FRETschemes and we intend to apply this
approach to other analytes of interest.
Received: June 6, 2003 [Z52075]
Published Online: September 23, 2003
[17] I. A. Levitsky, J. Kim, T. M. Swager,J. Am. Chem. Soc. 1999, 121,
1466.
Keywords: fluorescence · fluorides · materials science ·
.
polymers · sensors
[1] a) C. Dusemund, K. R. A. S. Sandanayake, S. Shinkai, J. Chem.
Soc. Chem. Commun. 1995, 333; b) H. Yamamoto, A. Ori, K.
Ueda, C. Dusemund, S. Shinkai, Chem. Commun. 1996, 407;
c) C. R. Cooper, N. Spencer, T. D. James, Chem. Commun. 1998,
1365; d) M. Nicolas, B. Fabre, J. Simonet, Chem. Commun. 1999,
1881; e) C. B. Black, B. Andrioletti, A. C. Try, C. Ruiperez, J. L.
Sessler, J. Am. Chem. Soc. 1999, 121, 10438, and references
therein; f) H. Miyaji, W. Sato, J. L. Sessler, Angew. Chem. 2000,
112, 1847; Angew. Chem. Int. Ed. 2000, 39, 1777; g) S. Mason,
J. M. Llinares, M. Morton, T. Clifford, K. Bowman-James, J. Am.
Chem. Soc. 2000, 122, 1814; h) M. Takeuchi, T. Shioya, T. M.
Swager, Angew. Chem. 2001, 113, 3476; Angew. Chem. Int. Ed.
2001, 40, 3372; i) For a recent review on anion recognition, see:
P. D. Beer, P. A. Gale, Angew. Chem. 2001, 113, 502; Angew.
Chem. Int. Ed. 2001, 40, 486.
4806 ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 4803 –4806