Fluorescent photoinduced electron transfer (PET) sensing of anions using charge neutral chemosensors

Article abstract of DOI:10.1039/b107608f

We demonstrate for the first time that the charge neutral anthracene based fluorescent sensors 1a-c, having an aromatic or aliphatic thiourea moiety as an anion receptor, show ideal PET sensor behaviour where the anthracene fluorescence emission is select

Full text of DOI:10.1039/b107608f

Fluorescent photoinduced electron transfer (PET) sensing of anions
using charge neutral chemosensors†
Thorfinnur Gunnlaugsson,*a Anthony P. Davisab and Mark Glynna
a Department of Chemistry, Trinity College Dublin, Dublin 2, Ireland. E-mail: gunnlaut@tcd.ie
b School of Chemistry, University of Bristol, CantockAs Close, Bristol, UK BS8 1TS
Received (in Cambridge, UK) 23rd August 2001, Accepted 23rd October 2001
First published as an Advance Article on the web 9th November 2001
We demonstrate for the first time that the charge neutral
anthracene based fluorescent sensors 1a–c, having an
aromatic or aliphatic thiourea moiety as an anion receptor,
show ideal PET sensor behaviour where the anthracene
fluorescence emission is selectively quenched upon titration
modulate or tune the acidity of the thiourea receptor moiety,
which would lead to different receptor–analyte complex
stability and hence different binding constants. Of the three
chemosensors, 1a was expected to show the strongest binding
due to this effect, and 1c the least. We initially investigated the
binding of 1a using (C4H9)4N(O2CCH3), since AcO2 is known
to form strong directional hydrogen bonding with thiourea, as
well as having a functional group of great biological rele-
with AcO2, H2PO4 and F2 but not by Cl2 and Br2 in
2
DMSO.
1
There is great interest in the design and synthesis of luminescent
based chemosensors for on-line and real time detection of
physiologically important ions and molecules,1 and for environ-
mental monitoring of harmful pollutants.2 While numerous
fluorescent and metal based delayed luminescent sensors for
cations and organic molecules have emerged from the fields of
supramolecular and coordination chemistry,3 sensors for se-
lective detection of anions are still relatively rare, despite the
fact that several elegant examples of anion receptors have been
reported over the years.4 These, however, often involve the
synthesis of complex and challenging organic hosts from
scaffolds such as cholic acid,5 calixarenes and peptides.6
Luminescent anion sensing has recently been achieved4 through
the use of anion receptors composed, for example, from metal
based Lewis acid centres,7 calix[4]pyrroles,8 thiouronium9 and
protonated quinoxaline,10 amine11 or polyamine moieties,12 but
the use of simple and easily synthesised electroneutral anion
receptors for such sensing has been less investigated.13
Intrigued by this fact, we set out to develop the charge neutral
chemosensors 1a–c, employing the criteria of PET sensing
using the fluorophore–spacer–receptor model developed by de
Silva for the detection of cations.14 A few research groups have
attempted to develop PET anion sensors.11–13,15 But, to the best
of our knowledge no such systems, employing neutral anion
receptors, have yet been reported that show ideal PET
behaviour, i.e. (i) only the quantum yield (intensity) and lifetime
of the fluorescence emission should be modulated upon ion
recognition due to (ii) changes in the free energy of electron
transfer (DGPET) between the excited state of the fluorophore
and the receptor upon ion recognition, and (iii) no changes
should be observed in the absorption spectra of the fluor-
ophore.14
vance.16 The H NMR of 1a in DMSO-d6, showed two sharp
signals at 9.62 ppm and 8.36 ppm for the thiourea hydrogens.
These were substantially shifted downfield upon addition of
0.1 ? 2 equivalents of (C4H9)4N(O2CCH3) (Dd = 1.92 and
1.66 ppm respectively after 1 eq.) signifying the formation of a
1+1 binding through hydrogen bonding with a log b = 3.2.†
The fluorescence emission spectra of 1a when titrated with
AcO2 in DMSO displayed typical PET behaviour. In the
absence of AcO2 the fluorescence emission spectra consisted of
three sharp bands at 443, 419 and 397 nm, with a shoulder at 473
nm, when excited at 370 nm with FF = 0.1037.‡ Upon addition
of the AcO2 (0 ? 32 mM), the intensity of these bands
gradually decreased with no other spectral changes being
observed (i.e. no spectral shifts or formation of new emission
bands), Fig. 1a. Using PET nomenclature, the emission can be
said to being ca. 70% (at 443 nm) ‘switched off’, with FF
=
0.0070. Concurrently, the absorption spectra of 1a, consisting of
bands at 390, 370, 352 and 336 nm, was hardly affected by the
addition of AcO2.† This confirms the insulating role of the
methylene spacer, which minimises any ground state inter-
actions between the fluorophore and the anion receptor. Similar
emission and absorption effects were observed for 1b and 1c.
When the fluorescence titrations of 1a–c were carried out in
CH3CN, CH3CO2Et or THF, the emission was also quenched
upon addition of AcO2 but the degree of quenching was
somewhat smaller. In EtOH, which is a highly competitive
hydrogen bonding solvent, no binding was observed between 1a
and AcO2. Furthermore, no exciplex emission was observed in
any of these solvents; in contrast, Teramae et al. have recently
shown that a pyrene analogue of 1c, is a ratiometric anion
The three PET chemosensors 1a–c, were easily made in good
yield from readily available starting materials, Scheme 1. The
9-aminomethyl anthracene 2, synthesised by reducing 9-cya-
noanthracene using B2H6 in THF, was reacted in dry CH2Cl2 at
room temperature, under an inert atmosphere with an equimolar
amount of 4-(trifluoromethyl)phenyl-, phenyl- and methyl
isothiocyanate respectively, 3a–c, yielding 1a–c as off-white
solids that were purified by crystallisation from CH2Cl2. For
comparable UV-Vis binding studies, the thiourea receptor 4 was
prepared in an analogous way from ethylamine. All products
were analysed by conventional methods.† The three different
isothiocyanates 3a–c, were chosen with the aim of being able to
† Electronic supplementary data (ESI) available: 1H, 13C NMR for 1a–c and
UV-Vis and NMR titration results for 1a are available as electronic
supplementary information (ESI). See http://www.rsc.org/suppdata/cc/b1/
b107608f/
Scheme 1 The synthesis of 1a–c. 4 was made in a similar way.
2556
Chem. Commun., 2001, 2556–2557
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b107608f

View Online
indicator based on the control of intramolecular exciplex
emission.13
the thiourea receptor. Measurements using 1b and 1c and the
same anions showed similar results. For 1b, the same selectivity
trend was observed as for 1a, with smaller binding constants due
to the reduced acidity of the thiourea protons. For 1c the order
of selectivity and the sensitivity was somewhat different with
H2PO42 (log b = 2.05 ( 0.05)) being selectively detected over
AcO2 (log b = 1.75 ( 0.05)). These results show that the anion
sensor’s affinity can be controlled by simple design.16
To investigate the selectivity and the sensitivity of the sensor
towards biologically important anions, we carried out a series of
titrations using N(C4H9)4+ salts of F2, Cl2, Br2 and H2PO42 in
DMSO. In the case of H2PO4 and F2 the fluorescence
2
emission was quenched by ca. 50 (FF = 0.0156) and 90% FF =
0.0011) respectively (at 443 nm), but only minor quenching
( < 7%) was observed when titrated with Cl2 (FF = 0.108) or
Br2 (FF = 0.088), ruling out a quenching by heavy atom effect.
We propose that the quenching is likely to be due to the
modulations of DGPET upon anion sensing. This can be
regarded as an enhancement in the rate of electron transfer from
the HOMO of the thiourea–anion complex to the anthracene
excited state, upon anion recognition i.e, the reduction potential
of the thiourea is increased causing PET to become com-
petitively more viable, which causes the fluorescence emission
to be quenched or ‘switched off’.§ Plotting the fluorescence
intensity changes (at 443 nm) as a function of log [anion] further
supports this view. Fig. 1b, shows several features commonly
In conclusion, the simple fluorescent PET anion chem-
osensors 1a–c show ideal PET sensing behaviour upon ion
recognition, e.g. only the fluorescence emission is ‘switched off‘
in the presence of AcO2, H2PO4 and F2. 1a–c are a very
2
important contribution to the fast growing field of supramo-
lecular anion recognition and sensing.
We thank Enterprise Ireland, Kinerton Ltd, and TCD for
financial support, Dr Hazel M. Moncrieff for helpful discussion
and Dr John E. O’Brien for NMR.
Notes and references
‡ 1a–c FF were measured by comparison with anthracene (FF = 0.27 in
EtOH); D. F. Eaton, Pure Appl. Chem., 1988, 60, 1107. log b was
determined from the equation:
seen for PET cation sensors e.g. the profiles for AcO2, H2PO4
2
and F2 are all sigmoidal, the quenching occurs over two log
concentration units, which is consistent with 1+1 binding and
simple equilibrium. From these changes the binding constant
log b for 1a was measured to be 3.35 ( 0.05) for F2, 2.55
( 0.05) for AcO2 and 2.05 ( 0.05) for H2PO42.‡ Similar
binding constants were found for 4a by measuring the changes
log [(Imax 2 IF)/(IF 2 Imin )] = log [anion] 2 log b.
§ CV measurements on 4 showed two irreversible oxidative waves.
Accurate DGET could not be determined from these measurements. We
investigated the PET dependence of 1a by comparing the FF of 1a with that
of 9-methylanthracene (9MA), which lacks the anion receptor. In the
absence of AcO2 the FF of 9MA was found to be 0.284 in DMSO. This
suggests that PET is active in 1a prior to the anion recognition, but becomes
even more efficient after anion recognition. In contrast, the addition of 40
mM of AcO2 to 9MA did not affect the FF.
in its absorption spectra at 286 nm. Importantly, 1a shows good
anion selectivity with AcO2 being recognised over H2PO4
,
2
but both represent families of biological important anions. The
fact that 1a shows higher affinity and more efficient quenching
for F2 than AcO2 is not surprising, since its high charge density
and small size enables it to form strong hydrogen bonding with
1 Chemical Sensors and Biosensors for Medical and Biological Applica-
tions, U. S. Spichiger-Keller, Wiley-VCH, 1998 Weinheim; Ger-
many.
2 C. F. Mason, Biology of Freshwater Pollution, 2nd. edn., Longman,
New York, 1991.
3 A. P. de Silva, D. B. Fox, A. J. M. Huxley and T. S. Moody, Coord.
Chem. Rev., 2000, 205, 41; T. Gunnlaugsson, D. A. Mac Dónaill and D.
Parker, Chem. Commun., 2000, 93.
4 P. A. Gale, Coord. Chem. Rev., 2001, 213, 79; P. A. Gale, Coord. Chem.
Rev., 2000, 199, 181; P. D. Beer and P. A. Gale, Angew. Chem. Int. Ed.,
2001, 40, 486; F. P. Schmidtchen and M. Berger, Chem. Rev., 1997, 97,
1609; P. D. Beer, Chem. Commun., 1996, 689.
5 A. P. Davis and L. J. Lawless, Chem. Commun., 1999, 9; A. P. Davis,
J. J. Perry and R. P Williams, J. Am. Chem. Soc., 1997, 119, 1793.
6 J. L. Atwood, K. T. Holman and J. W. Steed, Chem. Commun., 1996,
1401; A. Metzger and E. V. Anslyn, Angew. Chem. Int. Ed., 1998, 37,
649.
7 L. Fabbrizzi, M. Licchelli, G. Rabaioli and A. Taglietti, Coord. Chem.
Rev., 2000, 205, 85; P. D. Beer, V. Timoshenko, M. Maestri, P. Passaniti
and V. Balzani, Chem. Commun., 1999, 1755; R. S. Dickens, T.
Gunnlaugsson, D. Parker and R. D. Peacock, Chem. Commun., 1998,
1643.
8 P. Anzenbacher Jr., K. Jursíková and J. L. Sessler, J. Am. Chem. Soc.,
2000, 122, 9350; H. Miyaji, P. Anzenbacher Jr., J. L. Sessler, E. R.
Bleasdale and P. A. Gale, Chem. Commun., 1999, 1723.
9 Y. Kubo, M. Tsukahara, S. Ishihara and S. Tokita, Chem. Commun.,
2000, 653.
10 P. E. Kruger, P. R. Mackie and M. Nieuwenhuysen, J. Chem. Soc.,
Perkin Trans. 2, 2001, 1079; C. B. Blake, B. Andrioletti, A. C. Try, C.
Ruiperez and J. L. Sessler, J. Am. Chem. Soc., 1999, 121, 10438.
11 H. Xie, S. Yi, X. Yang and S. Wu, New J. Chem., 1999, 23, 1105.
12 D. H. Vance and A. W. Czarnick, J. Am. Chem. Soc., 1994, 116, 9397;
M. E. Huston, E. U. Akkaya and A. W. Czarnick, J. Am. Chem. Soc.,
1989, 111, 8735.
13 S. Nishizawa, H. Kaneda, T. Uchida and N. Teramae, J. Chem. Soc.,
Perkin Trans. 2, 1998, 2325.
14 A. P. de Silva, H. Q. N. Gunaratne, T. Gunnlaugsson, A. J. M. Huxley,
C. P. McCoy, J. T. Rademacher and T. E. Rice, Chem. Rev., 1997, 97,
1515.
15 C. R. Cooper, N. Spencer and T. D. James, Chem. Commun., 1998,
1365.
16 P. Bühlmann, S. Nishizawa, K. P. Xiao and Y. Umezawa, Tetrahedron,
1997, 53, 1647.
Fig. 1 (a) The changes in the fluorescence spectra of 1a in DMSO upon
addition of acetate. From top: [AcO2] = 0, 92 µM, 550 µM, 1.8 mM, 8.9
mM, 26 mM, 32 mM. (b) Titration profile for 1a showing the changes in the
fluorescence emission as a function of added anion: . = F2, 5 = AcO2,
3 = H2PO42, 8 = Cl2, $ = Br2, when measured at 443 nm. All
titrations were repeated two to three times to ensure reproducibility.
Chem. Commun., 2001, 2556–2557
2557