A fluorescent 18-crown-6 based luminescence sensor for lanthanide ions

Article abstract of DOI:10.1016/S0040-4020(00)00528-7

The synthesis and photophysical behavior of a luminescent chemosensor for the early lanthanide elements is described. The sensor is a 1,4-diphenylethynyl-benzene chromophore having 18-crown-6 moieties bound to the outer phenyl rings. The chromophore is lu

Full text of DOI:10.1016/S0040-4020(00)00528-7

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 7045±7049
A Fluorescent 18-Crown-6 Based Luminescence Sensor
for Lanthanide Ions
*
Wen-Sheng Xia, Russell H. Schmehl and Chao-Jun Li
*
Department of Chemistry, Tulane University, New Orleans, LA 70118, USA
Received 15 March 2000; revised 31 March 2000; accepted 2 April 2000
AbstractÐThe synthesis and photophysical behavior of a luminescent chemosensor for the early lanthanide elements is described. The
sensor is a 1,4-diphenylethynyl-benzene chromophore having 18-crown-6 moieties bound to the outer phenyl rings. The chromophore is
luminescent and the emission is quenched by lanthanide (Ln31) ions with larger ionic radii and existing f±f transitions (Ce31, Pr31 and Nd31).
Alkali, alkaline earth ions and lanthanides with smaller radii (Ge31, Tb31, Dy31, and Yb31) do not affect the emission. q 2000 Elsevier
Science Ltd. All rights reserved.
Introduction
particular receptor; (b) tuning the excited state character-
istics of the chromophore; and (c) exploiting adventitious
differences in quenching rate constants for different lantha-
nides. While the ionic radii of the lanthanide ions system-
atically decrease across the lanthanide series, the difference
is relatively small (18 pm) as compared to, for instance, the
alkali ions (.70 pm). However, Umetani and coworkers
found that binding constants of different benzo-crown ethers
with lanthanides vary by a factor of 10 or more across the
series.5 For instance, a sulfonated benzo-18-crown-6
derivative studied by the group has a binding constant of
The use of organic chromophores to sensitize the lumi-
nescence of particular lanthanides (Ln31 ions) has drawn
considerable attention in recent years.1 This is in part due
to the potential for generating systems that exhibit narrow
band, lanthanide localized, luminescence from organic
electroluminescent polymers. Sensitization of lanthanide
excited states is known to occur via the triplet state of the
organic chromophore.2 In addition, the ¯uorescence yield of
the sensitizer is also diminished in some systems because of
either direct energy transfer to the lanthanide or, in
covalently linked sensitizer/lanthanide systems, enhanced
intersystem crossing in the sensitizer induced by a heavy
atom effect of the bound lanthanide. Intramolecular ¯uor-
escence quenching has been shown to occur in supramolec-
ular sensitizer/chelate systems capable of coordinating
lanthanides.1,3,4 Among the systems investigated are various
benzo-crown ether derivatives. Costa and coworkers found
that, for benzo-15-crown-5 in alcohol solvents, Tb31 effec-
tively quenched the benzene ¯uorescence while only a small
amount of quenching was observed with Eu31.3 Others have
found that 15-crown-5 lanthanide complexes of Tb31, Eu31
and Dy31 all exhibit luminescence enhancement when an
organic sensitizer is used as a counterion for the crown/
lanthanide complex in nonaqueous solvents.4
75 M21 for coordination of La31 and only 2 M21 for Yb31
.
Some degree of selectivity can also be achieved by tuning
the energetics of the organic sensitizer used. Nocera and
coworkers showed that Gd31 and Tb31, ions of nearly
identical radius, can be distinguished by a system having
biphenyl bound in the cavity of a cyclodextrin with the
lanthanide coordinated on the periphery of the cyclo-
dextrin.6 In this case the emitting excited state of Gd31 is
higher in energy than the biphenyl triplet excited state and is
not sensitized while Tb31 sensitization is energetically
favorable.
In this paper we show that complex 1, below, having a
benzo-18-crown-6 receptor and a di(phenylethynyl)benzene
chromophore discriminates between the early and late
lanthanides through quenching of the chromophore
¯uorescence.
Supramolecular chromophore/receptor systems can poten-
tially be used for the selective detection of particular lantha-
nides. Selectivity can be achieved via: (a) control of the
Keywords: chemosensor; lanthanide; luminescence; sensitization; crown
ether.
* Corresponding authors. Tel.: 1504-865-5573; fax: 1504-865-5596;
e-mail: russ@mailhost.tcs.tulane.edu; e-mail: cji@mailhost.tcs.tulane.edu
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00528-7

7046
W.-S. Xia et al. / Tetrahedron 56 (2000) 7045±7049
Table 1. Ionic radii of lanthanide ions and relative ef®ciencies for quench-
ing 1 luminescence in CH3CN
Lanthanide
Ionic radius, pma
Quenched (%)b
La31
Ce31
Pr31
118
114
114
112
104
103
98
,10
70
65
61
,5
15
10
Nd31
Tb31
Dy31
Yb31
a Radii for 8 coordinate Ln31 ions from Handbook of Chemistry and
Physics, 71st ed., Lide D. R., Ed.; CRC: Boca Raton, 1991; pp 4±126.
b Percent quenched based upon decrease in luminescence intensity in
solutions containing a large excess of the lanthanide ion.
390 nm. The absorption is similar to that of a related alkoxy
1,4-di(phenylethynyl)benzene derivative.9
Luminescence from 1 appears as the mirror image of the
absorption with a maximum of 406 nm in CH3CN. The
quantum yield for emission in acetonitrile is high, 0.67
(relative to carbazole reference) and is more like the
emission yield of poly(phenylene ethynylene) than that of
the weakly emitting diphenylacetylene.10
Figure 1. Emission spectra of compound 1 in acetonitrile following addi-
tion of different quantities of La31. [1]/[La31] ratio 1:0 (solid line), 1:1
(±´ ´±), 1:2 (´´´), 1:3 (Ð). [1]ˆ2£1025 M.
The synthesis, spectroscopic properties and luminescence
behavior of acetonitrile solutions of compound 1 in the
absence and presence of various lanthanides is presented.
The luminescence of 1 in the presence of increasing
amounts of La31 is shown in Fig. 1. The integrated lumi-
nescence intensity decreases slightly, the emission maxi-
mum shifts to the blue (395 nm) and the spectrum
becomes more structured with a new band appearing at
368 nm. The luminescence is not signi®cantly quenched,
in part because La31 has no f±f transitions and may also
lack a ligand (phenoxy)-to-metal charge transfer transition
at energies lower than the ligand localized p±pp state. The
distinct emission spectral change without appreciable
Results and Discussion
Ligand 1 was prepared by palladium catalyzed cross
coupling methods similar to those used by others in making
unsymmetric diaryl acetylene derivatives.7,8 Characteri-
1
zation by H and 13C NMR con®rmed the structure of the
quenching of the luminescence intensity is unique to La31
.
bis-crown ligand. The absorption maximum for the ligand
p±pp transition is at 340 nm (log eˆ4.8) with two
shoulders, one at 370 nm and the other at approximately
Addition of Ce31, Nd31 and Pr31 all induce similar emission
spectral changes and, in addition, the luminescence is
signi®cantly quenched. Fig. 2 shows changes in the
emission spectrum that occur upon addition of near stoichio-
metric quantities of Nd31 to an acetonitrile solution of 1. As
with the La31 complex, a new emission band is observed at
368 nm, but the emission intensity drops to approximately
40% that of the unquenched ligand. Very similar results are
observed following addition of Ce31 and Pr31 to 1 in aceto-
nitrile. The limiting degree of quenching is given in Table 1
and the fraction of 1 luminescence quenched as a function of
the lanthanide concentration is shown in Fig. 3. Given the
fact that the luminescence lifetime of the sensitizer 1 is
,10 ns11 and the concentration of the lanthanide is less
than 1024 M in these experiments, the emission spectral
changes cannot be due to diffusional bimolecular reactions
of excited 1 and the lanthanide and must be the result of
formation of some sort of complex between the lanthanide
and 1.
Complexation of the lanthanide ions to 1 is further indicated
by changes in the absorption spectra, shown in Fig. 4 for
Nd31. While the observed spectral perturbation is much
smaller than that observed in the emission spectra, a small
blue shift in the absorption maximum is observed. It is clear
Figure 2. Emission spectra of 1 in acetonitrile following the addition of
different amounts of Nd31. [1]/[Nd31] ratio 1:0 (solid line), 1:1 (±´±), 1:2
(´´´), 1:3 (±´±). [1]ˆ2£1025 M.

W.-S. Xia et al. / Tetrahedron 56 (2000) 7045±7049
7047
the luminescence lifetime of 1 is less than 10 ns,11 a lower
limit on the luminescence quenching rate constant can be
estimated. For Nd31, the degree of quenching of the fully
complexed 1 is approximately 65%; if the lifetime of 1 is
assumed to be the upper limit of 10 ns, the quenching rate
constant is approximately 3£108 s21. Similar values are
obtained for Ce31 and Pr31. The data of Fig. 3 were ®t
using only equilibrium 1a, estimating the plateau fractions
quenched to be 80, 70 and 68% for Ce31, Pr31 and Nd31
,
respectively, and using an equilibrium constant for asso-
ciation of 2£105 M21. With an equilibrium constant of
this magnitude and concentrations of 1 and lanthanide ion
in the 20±80 mM range, approximately 90% of the ¯uoro-
phore is coordinated to a lanthanide at the highest lanthanide
concentrations.
The luminescence spectral bandshape and intensity are not
changed signi®cantly upon addition of a ten fold molar
excess of lanthanides with smaller ionic radii including
Dy31, Tb31 and Yb31. For Dy31, the luminescence of 1 is
partially quenched, but the spectral bandshape is
unchanged. Neither Tb31 or Yb31 quench the ¯uorescence
of 1 to any signi®cant extent at near stoichiometric concen-
trations and no absorption spectral changes are observed in
solutions containing low concentrations of the ions and 1
(approx. 1025 M). The results suggest the binding constant
for association of 1 with Tb31 and Yb31 may be signi®cantly
smaller than the binding constants for the other lanthanide
ions.
Figure 3. The percentage of 1 emission quenched as a function of the ratio
of added lanthanide ions to compound 1 in acetonitrile. La31 (£), Ce31 (W),
Pr31 (O), Nd31 (P), Dy31 (1), Tb31 (K), Yb31 (L). [1]ˆ2£1025 M.
from Fig. 4 that no isobestic point is observed.12 This is
expected since 1 has two crowns for coordination of the
lanthanide and three absorbing species (1, Ln131 and
Ln2131) can exist in solution as shown in Eqs. (1a) and
(1b) below.
1 1 Ln31 ˆ Ln131
Ln131 1 Ln31 ˆ Ln2161
The data indicate that the spectral changes appear complete
when the added lanthanide is in approximately four-fold
excess over the 20 mM 1 in acetonitrile. Fig. 3 also shows
that the degree of quenching reaches a plateau and that the
fraction of 1 emission quenched is between 60 and 75% for
the three lanthanides exhibiting signi®cant quenching. Since
ꢀ1a†
ꢀ1b†
The trend that emerges is illustrated in Table 1. All of the
ions with radii greater than 110 pm associate with the
18-crown-6 moieties of 1 with large enough binding
constants to leave relatively little uncomplexed 1 in solution
even in the 40±80 mM range. The results of Umetani indi-
cate that, for a benzo-18-crown-6, the binding constant
difference between the ions with larger radii (La31, Ce31
,
Pr31 and Nd31) and the ions with smaller radii (Dy31, Tb31
and Yb31) is at least a factor of 6.4 The values measured by
Umetani were obtained in aqueous solution and may not
directly relate to acetonitrile solutions where the binding
constants are much larger, however, the trend of decreasing
binding constant with decreasing lanthanide radius is likely
to be observed regardless of solvent. In dilute solutions of 1
and the lanthanide, a factor of 10 difference in binding
constant is enough to result in order of magnitude
differences in the fraction of free chromophore in solution.
Thus for these lanthanides, examined in dilute solutions at
near stoichiometric ratios of the lanthanide to sensitizer,
relatively small binding constant differences can lead to
large differences in the fraction of lanthanide bound to the
sensitizer and the degree of quenching observed.
The mechanism of ¯uorescence quenching is not entirely
clear in these systems. One mechanism involves energy
transfer to create excited states of the lanthanide by either
coulomb or exchange processes. The rate constant for
energy transfer by coulombic interactions is determined
by the radiative decay rate constant of the energy donor
and the magnitude of overlap of the normalized donor
emission and the acceptor absorption (weighted by the oscil-
lator strength of the acceptor). Because the molar absorp-
tivities of all the lanthanides are very small (,10 M21
Figure 4. Absorption spectral changes of 1 in acetonitrile with the addition
of different amounts of Nd31. [1]/[Nd31]: 1:0 (solid line), 1:1 (Ð), 1:2 (´´´).
[1]ˆ2£1025 M.

7048
W.-S. Xia et al. / Tetrahedron 56 (2000) 7045±7049
cm21), the overlap integral will be small even if the donor
emission and acceptor absorption spectra overlap com-
pletely. As a result, calculated rate constants for coulombic
energy transfer are several orders of magnitude lower than
the radiative decay rate constant of 1. Energy transfer from
the singlet state of 1 to the lanthanides could also occur via
an electron exchange process. The crown is covalently
attached directly to the chromophore of 1 and, provided
electronic coupling between the lanthanide and the chromo-
phore is adequate, the energy transfer could occur by this
mechanism. However, it is known that the contraction of the
lanthanide f orbitals tends to result in very weak electronic
coupling in energy transfer reactions.13 An alternate
mechanism for ¯uorescence quenching involves increasing
the ef®ciency for intersystem crossing in 1 as a result of
spin-orbit coupling interactions of 1 with the lanthanide
following complexation. This heavy atom effect depends
strongly on the separation of chromophore and the lantha-
nide. Given the fact that La31, an ion having no low energy
f±f states available for energy transfer, does not signi®-
cantly alter the intensity of the luminescence of 1, quench-
ing by intersystem crossing enhancement seems less likely
than quenching by energy transfer. We plan to examine this
in greater detail using transient absorption spectroscopy to
look at formation and decay of triplet 1 in the presence and
absence of various lanthanide ions.
nide salts were obtained from Fischer Chemical as nitrates
and were used without additional puri®cation.
Synthesis of 1,4-bis(30-ethynylbenzo-18-crown-6)-2,3,
5,6-tetramethylbenzene, 1. In a 25 ml rb ¯ask Pd(OAc)2
(44 mg, 0.05 equiv.), CuI (37 mg, 0.05 equiv.) and tri-
phenylphosphine (102 mg, 0.1 equiv.) were suspended in
6:1 CH3CN/H2O (3 ml) solution under an N2 blanket. The
mixture was stirred for 20 min. In a separate 250 ml rb ¯ask
1,4-diiodo-2,3,5,6-tetramethylbenzene (1.5 g, 1 equiv.),
trimethylsilylacetylene (1.5 g, 4 equiv.) and triethylamine
(2.7 ml) were dissolved in 6:1: THF/H2O and purged with
N2. The Pd containing solution was then transferred to the
tetramethylbenzene solution by cannulation and the mixture
was stirred under N2 for 2 h. Following removal of the THF
under reduced pressure, methanol (100 ml) was added and a
brown precipitate formed. The precipitate was collected by
®ltration and puri®ed by ¯ash column chromatography on
silica using hexane as eluent. The principal chromato-
graphic band was evaporated to a small volume, precipitated
by addition of methanol (50 ml) and ®ltered to yield 1,4-
di(trimethylsilylethynyl)-2,3,5,6-tetramethylbenzene (0.5 g,
46%). This was dissolved in THF (50 ml) and tetrabutyl-
ammonium ¯uoride (1.8 ml, 1 M THF solution) was added.
The mixture was stirred at room temperature overnight. To
this 0.1N HCl (100 ml) was added and the resulting mixture
was extracted twice with CH2Cl2 (100 ml). The organic
layers were mixed, dried with MgSO4, puri®ed by ¯ash
chromatography on silica using hexane eluent and the
second fraction was collected to yield a yellow solid on
evaporation. The solid was reprecipitated from CH2Cl2/
ethanol to yield 1,4-diethynyl-2,3,5,6-tetramethylbenzene
(0.3 g, 92%).
An additional point relating to these systems is that no
lanthanide luminescence is observed. In the experiments
involving addition of lanthanide ions to acetonitrile
solutions of 1, the lanthanide was dissolved in an aqueous
solution and the solutions were aerated. Others have noted
that lanthanide ion quenching of aromatic hydrocarbon
chromophores yields no lanthanide luminescence and have
attributed the observation to the existence of low lying
charge transfer states.14 Given this, the lack of lanthanide
luminescence in these systems is not unexpected.
Compound 1 was prepared by mixing the diethynylbenzene
derivative (58 mg, 0.32 mmol) and 30-bromobenzo-18-
crown-6 (250 mg, 0.64 mmol) in piperidine (dried by
P2O5, 50 ml) and adding a N2 purged piperidine solution
of Pd(OAc)2 (7 mg), CuI (6 mg) and triphenylphosphine
(17 mg). The resulting solution was heated to 808C under
an N2 blanket for 2 h. The solution was cooled to room
temperature and poured into 0.1N HCl (100 ml) and then
extracted with CH2Cl2 (50 ml) twice. The organic layers
were collected, dried with MgSO4, concentrated by
evaporation and the product was precipitated by addition
of methanol. The product was reprecipitated from CH2Cl2/
methanol twice and the resulting pale yellow solid of 1 was
obtained by ®ltration. Yield 70 mg (71%). 1H NMR (CDCl3,
d ppm): 7.14 (1H, d, Jˆ8 Hz), 7.04 (1H, d, Jˆ3 Hz), 6.84
(1H, d, Jˆ12 Hz), 4.19 (4H, s), 3.94 (4H, s), 3.73 (12H, m),
2.49 (6H, s). 13C NMR (CDCl3, d ppm) 150, 149, 136, 120.6,
124.3, 117, 114, 98.6, 87.8, 71.3, 71.2 (m), 71.1(m), 69.9
(d), 69.5 and 69.4. IR (KBr, cm21): 2914, 2858, 1506, 1456,
1400, 1250, 1123, 1056. Elemental analysis for C46H58O12
(%), C: 68.83; H: 7.23. Found: C: 69.32; H, 7.28.
Conclusion
This work illustrates that, through a combination of ion
selectivity and differences in energy transfer ef®ciencies,
some degree of selectivity is obtained in discriminating
lanthanides in solution. The chromophore 1, having two
10-crown-6 moieties, coordinates the early lanthanides
more effectively than those at the end of the series. The
association of the lanthanide can be detected spectrophoto-
metrically, but much larger changes are observed in the
chromophore ¯uorescence upon lanthanide coordination to
1. The ¯uorescence of 1 is quenched upon complexation of
particular lanthanides and preliminary results suggest
quenching is via an intramolecular energy transfer process.
The results also suggest that tuning of the crown ring size
may allow some degree of selectivity in sensing lanthanides
with very similar ionic radii from those with either larger or
smaller radii.
Spectral analysis
NMR spectra were obtained on a GE-400 spectrometer
(400 MHz). Absorption spectra were obtained on an HP
8452 diode array spectrophotometer. All luminescence
measurements were made using a SPEX Fluorolog spectro-
¯uorimeter equipped with a 450 W Xe arc lamp for excitation,
Experimental
All solvents used were reagent grade and were used without
further puri®cation unless speci®ed otherwise. The lantha-

W.-S. Xia et al. / Tetrahedron 56 (2000) 7045±7049
7049
1995, 67, 135±140. (e) Sabbatini, N.; Guardigli, M.; Lehn, J. M.
Coord. Chem. Rev. 1993, 123, 201±228.
single grating monochromators and both CCD and thermo-
electrically cooled PMT detectors. Spectra were not
corrected for detector response.
2. (a) Latva, M.; Takalo, H.; Mukkala, V. M.; Matachescu, C.;
Rodriguez Ubis, J. C.; Kankare, J. J. Lumin. 1997, 75, 149±169.
(b) Sabbatini, N.; Guardigli, M.; Manet, I.; Bolletta, F.; Ziessel, R.
Inorg. Chem. 1994, 33, 955±959.
1
UV±Vis spectra, ¯uorescence spectra H NMR measure-
ments were obtained by adding the concentrated solution
of metal ions to a solution of ligands in a 1 cm path length
cuvette by aliquot. No more than 10 ml of metal ion solution
was added in 3 ml of the measuring solution for ¯uor-
escence and UV±Vis measurements. All absorption and
luminescence measurements were made using aerated
solutions.
3. Costa, S. M.; Queimado, M. M.; Frausto da Silva, J. J. R.
J. Photochem. 1980, 12, 31±39.
4. Tran, C. D.; Zhang, W. Anal. Chem. 1990, 62, 835±840.
5. Sasaki, T.; Umetani, S.; Matsui, M.; Tsurubou, S.; Kimura, T.;
Yoshida, Z. Bull. Chem. Soc. Jpn 1998, 71, 371±377.
6. Rudzinski, C. M.; Engebretson, D. S.; Hartmann, W. K.;
Nocera, D. G. J. Phys. Chem. A 1998, 102, 7442±7446.
7. (a) El-Ghayoury, A.; Ziessel, R. Tetrahedron Lett. 1997, 38,
2471±2474. (b) Grosshenny, V.; Romero, F. M.; Ziessel, R.
J. Org. Chem. 1997, 62, 1491±1500.
Acknowledgements
8. (a) Weijin, T.; Nesbitt, S.; Heck, R. F. J. Org. Chem. 1990, 55,
63. (b) Heck, R. F. Org. React. 1981, 27, 345.
The authors wish to thank the Tulane/Xavier Center for
Bioenvironmental Research for support of this work.
9. Nguyen, P.; Lesley, G.; Marder, T. B.; Ledoux, I.; Zyss J. Chem.
Mater. 1997, 9, 406±408.
References
10. Davey, A. P.; Elliott, S.; O'Connor, O.; Blau, W. Chem.
Commun. 1995, 1433.
1. (a) Steemers, F. J.; Verboom, W.; Reinhoudt, D. N.; van der
Tol, E. B.; Verhoeven, J. W. J. Am. Chem. Soc. 1995, 117, 9408±
9414. (b) Sabbatini, N.; Guardigli, M.; Manet, I. Adv. Photochem.
1997, 23, 213±278. (c) Magennis, S. W.; Parsons, S.; Corval, A.;
Woollins, J. D.; Pikramenou, Z. Chem. Commun. 1999, 61±62. (d)
Sabbatini, N.; Guardigli, M.; Manet, I.; Ungaro, R.; Casnati, A.;
Ziessel, R.; Ulrich, G.; Asfari, Z.; Lehn, J.-M. Pure Appl. Chem.
11. The lifetime is based upon preliminary time correlated single
photon counting experiments in our laboratory.
12. The absorption spectral changes for Ce31 and La31 are larger
and the lack of an isobestic point is clearly apparent.
13. Porter, G.; Wright, M. R. Discuss. Faraday Soc. 1959, 27, 18.
14. Sohna, J.-E. S.; Fages, F. Tetrahedron Lett. 1997, 38, 1381±
1384.