Di- and tetracarboxylate ligands for highly luminescent terbium(III) complexes on the basis of sulfonylcalix[4]arene scaffold

Article abstract of DOI:10.1016/j.tetlet.2006.11.153

New di- (2) and tetracarboxylate ligands (4) were prepared on a sulfonylcalix[4]arene platform by O-alkylation of thiacalix[4]arene with ethyl bromoacetate, followed by hydrolysis of the ester function and oxidation of the sulfide bridges. The sulfonyl-based ligands 2 and 4 formed luminescent 1:1 complexes with terbium(III) ion having higher luminescent quantum yield (Φ = 0.29₁ and 0.28₇, respectively) than 1:1 complexes of the corresponding thiacalix[4]arene-based di- (1) and tetracarboxylate ligands (3) (Φ = 0.03₈ and 0.00₃, respectively), implying higher efficiency of sulfonyl ligands (2 and 4) than those of thia ligands (1 and 3) in the energy transfer process.

Full text of DOI:10.1016/j.tetlet.2006.11.153

Tetrahedron Letters 48 (2007) 821–825
Di- and tetracarboxylate ligands for highly luminescent
terbium(III) complexes on the basis of sulfonylcalix[4]arene scaffold
Takayuki Horiuchi,a Nobuhiko Iki,a,* Hitoshi Hoshino,a
Chizuko Kabutob and Sotaro Miyanoa
aDepartment of Environmental Studies, Graduate School of Environmental Studies, Tohoku University,
Aoba-ku, Sendai 980-8579, Japan
bDepartment of Chemistry, Graduate School of Science, Tohoku University, Aoba-ku, Sendai 980-8578, Japan
Received 1 September 2006; revised 24 November 2006; accepted 27 November 2006
Available online 14 December 2006
Abstract—New di- (2) and tetracarboxylate ligands (4) were prepared on a sulfonylcalix[4]arene platform by O-alkylation of thiaca-
lix[4]arene with ethyl bromoacetate, followed by hydrolysis of the ester function and oxidation of the sulfide bridges. The sulfonyl-
based ligands 2 and 4 formed luminescent 1:1 complexes with terbium(III) ion having higher luminescent quantum yield (U = 0.291
and 0.287, respectively) than 1:1 complexes of the corresponding thiacalix[4]arene-based di- (1) and tetracarboxylate ligands (3)
(U = 0.038 and 0.003, respectively), implying higher efficiency of sulfonyl ligands (2 and 4) than those of thia ligands (1 and 3) in
the energy transfer process.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Energy-transfer luminescence is the emission from the
metal center of a lanthanide(III) complex, the excitation
energy for which is transferred from the coordinated
ligand in T1 state. Owing to the mechanism, the energy-
transfer luminescence has characteristic features such
as long lifetime (ls–ms order), large Stokes shift (200–
300 nm), and sharp emission band, which are attractive
from the viewpoint of applications to luminescent label-
ing and sensing.1 For such purposes, design of the ligand
plays the key role, where the prerequisites are (1) ther-
modynamic and kinetic stability of the formed complex;
(2) ability to shield LnIII ion from solvent molecule; and
(3) availability of light absorbing antenna having suit-
able T1 level responsible for the energy transfer. So
far, calixarenes have served as a hopeful molecular scaf-
fold to construct ligands for such use owing to the fea-
sibility of chemical modifications at the phenol moiety
with functional groups to fulfill the above-mentioned
requirements.2 On the contrary, Shinkai and co-workers
reported energy transfer luminescence of a TbIII com-
plex with calix[4]arene-p-tetrasulfonate (CAS), which,
notably, does not have any auxiliary ligating and antenna
groups.3 Recently, we found that sulfur-bridged calix[4]-
arenes, that is, thia- and sulfonylcalix[4]arene-p-tetra-
sulfonates (TCAS and SO2CAS, respectively) can also
be used as the ligand to give TbIII complexes, giving
higher luminescent intensity and longer lifetime (s) than
CAS did.4 The high efficiency of the sulfur-containing
ligands is ascribed to the high efficiency in light absorp-
tion, namely large molar absorptivity (e) as well as the
coordination ability of the bridging sulfur-group to
reduce the number of the coordinated water molecules.
In quest of superior ligands for the luminescence, we
designed ligands 2 and 4 based on sulfonylcalix[4]arene
(SO2CA) as a scaffold because of the larger e than that
of thiacalix[4]arene (TCA), where two and four carb-
oxymethyl groups are introduced to afford more stable
complexes with better shielding ability of the metal cen-
ter from water. Here, we report the synthesis of the
ligands, their complexation behavior toward TbIII ion,
and the luminescence properties.
Although ligands 2 and 4 were designed on a SO2CA
scaffold, we were reluctant to take the direct etherifica-
tion route by reacting with ethyl bromoacetate consider-
ing the low nucleophilicity of the phenoxide Oꢀ flanked
by two o,o0-sulfonyl groups via delocalization of the
charge through resonance.5,6 Instead, we took the route
of the initial O-functionalization of TCA followed by
oxidation of the bridging sulfur to sulfone: Heating a
Keywords: Calixarenes; Ligand; Luminescence; Terbium.
*
Corresponding author. Tel.: +81 22 795 7222; fax: +81 22 795
7293; e-mail: iki@orgsynth.che.tohoku.ac.jp
0040-4039/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.11.153

822
T. Horiuchi et al. / Tetrahedron Letters 48 (2007) 821–825
OR1
OR2
X
Y
X
2
Y
Y
X
R1
R2
–
–
–
CH2
S
S
SO2
SO2
S
SO2
S
SO3
But
SO3
But
SO3
But
But
But
But
But
But
But
H
H
H
H
H
H
H
H
H
H
CAS
TCA
TCAS
SO2CA
SO2CAS
CH2CO2H
CH2CO2H
CH2CO2H
CH2CO2H
CH2CO2Et
CH2CO2Et
CH2CO2Et
H
H
1
2
3
4
5
6
7
R1
R1
H
SO2
S
S
R1
R1
SO2
mixture of TCA, 8-fold excess of ethyl bromoacetate,
and Na2CO3 as a base catalyst in acetone gave cone-
shaped tetra-O-alkylated 6.7 According to the method
reported by Lhotak8 oxidation of compound 6 by use
of large excess of 3-chloroperoxybenzoic acid (MCPBA)
afforded sulfonyl 7 in a good yield (74%). However, at-
tempted hydrolysis of the ester function of 7 by heating
with K2CO3 as the base in DMSO–water gave a mixture
of bare SO2CA and mono-, di-, and tricarboxylates of
one. This suggests that the o,o0-disulfonyl-substituted
phenoxy moiety of 7 is highly susceptible to nucleophilic
attack by OHꢀ to release the (carbonyl)methoxy moiety
to give the phenols rather than the desired ester-hydro-
lysis products. In other words, SO2CA behaved as a
‘good’ leaving group as indicated above.
Figure 1. ORTEP drawing of ligand 4 with thermal ellipsoids drawn at
30% probability. For clarity, parts of methyl groups found in
disordered But moiety are shown.
Hence, as an alternative route to 4, cone-shaped 6 was
first subjected to hydrolysis to give tetracarboxylic acid
3 as reported elsewhere,9 which was in turn oxidized
with MCPBA to afford 4 in a good yield (85%).10 X-
ray analysis of 4 (Fig. 1)11 showed high disorder mainly
in tert-butyl groups to result in a large R value, but the
resolution is sufficient to establish that the cone-shaped
structure was maintained during the initial hydrolysis
and subsequent oxidation processes as well as to show
that one 1,4-dioxane molecule (disordered) was included
Figure 2. Aromatic and methylene part of 1H NMR spectrum for
ligand 4 (298 K, CD3OD).
atoms. Then the ester function of 5 was hydrolysed by
heating with KOH in ethanol–water mixture to give
TCA-dicarboxylate 1,9 which was further treated with
MCPBA to afford 2 in a yield of 84% from 1.15 The
1
in the cavity. H NMR of 4 in methanol-d4 exhibited
three singlet but broad peaks for But, –CH2O–
(Fig. 2), and ArH protons, suggesting certain molecular
motion reducing molecular symmetry from ideal C4
symmetry of cone conformation. This can be rational-
ized by assuming C2v–C2v conformational interconver-
sion observed in tetra-O-alkylated calix-,12 thiacalix-,13,14
and sulfonylcalix[4]arenes in solutions.8
1
syn conformation of 2 was evidenced by the H NMR
spectrum showing two sets of sharp singlet peaks for
But and ArH, which was maintained during the trans-
formations from syn-di-O-alkylated 5.
In methanolic solution, SO2CA-based dicarboxylic acid
2 has an absorption maximum at 290 nm, which origi-
For preparation of SO2CA-based dicarboxylic acid 2,
synthetic route similar to that for 4 was employed: First,
a mixture of TCA, 2-fold excess of ethyl bromoacetate,
and Na2CO3 in acetone were heated to afford syn-di-O-
alkylated TCA (5),9 in which a pair of distal phenol oxy-
gen atoms have two (ethoxycarbonyl)methyl groups
directing syn with respect to a mean plane of four sulfur
nates from p–p transition at the aromatic ring.16 In
*
the presence of an excess of triethylamine, another
absorption band around 350 nm appeared, suggesting
dissociation of the phenolic OH of 2. Upon addition
of terbium(III) nitrate, the absorption at 338 nm linearly
increased up to [TbIII]T/[2]T = 1 (Fig. 3), meaning that 2
formed 1:1 complex. TCA-based 1 similarly behaved to

T. Horiuchi et al. / Tetrahedron Letters 48 (2007) 821–825
823
center in effect. In terms of light absorbing ability, 1:1
TbIII-4 complex has smaller e values than TbIII-3 does
(Table 1), which may result from the splitting of absorp-
tion band of TbIII-4 into two at 290 and 348 nm. In sum-
mary, dicarboxylate ligands 1 and 2 rapidly formed 1:1
complex under a wide range of [TbIII]T/[ligand]T ratio,
whereas tetracarboxylate ligands 3 and 4 slowly did
1:1 in the range of [TbIII]T/[ligand]T 6 1. The slow com-
plex formation of, especially, SO2CA-based tetracarb-
oxylate 4 is indicative of high kinetic stability of the
TbIII complex.
Methanolic solutions containing 1:1 complex of ligands
1–4 with TbIII ion exhibited strong luminescence consist-
ing of four emission bands (Fig. 4a), which correspond
5
7
7
to the transitions from D4 to F6 (around 490), F5
(545), F4 (580), and F3 (620 nm) at TbIII center. Band
shapes slightly differ among complex species, implying
that the coordination environment provided by these
ligands has subtle influences on the f–f electronic transi-
tion in the TbIII center. Excitation spectrum of each
complex shown in Fig. 4b has maxima corresponding to
the absorption maxima, indicating that TbIII center
7
7
Figure 3. Molar ratio plots for TbIII-sulfonyl ligand systems. Closed
circle: diamond:
[2]T = 2.0 · 10ꢀ5 M, [NEt3]T = 2.0 · 10ꢀ4 M,
[4]T = 2.0 · 10ꢀ5 M, [NEt3]T = 1.0 · 10ꢀ4 M.
form 1:1 complex with TbIII, the absorption maximum
of which is at 320 nm originating from p–p transition
*
at the phenoxy moiety. Notably, the e value for TbIII
complex with SO2CA-based 2 at the absorption maxi-
mum is larger than that with TCA-based 1 (Table 1),
meaning that higher efficiency of SO2CA scaffold in light
absorption than that of TCA one is retained after di-O-
alkylation.
Dicarboxylate ligands 1 and 2 formed 1:1 complex with
TbIII ion very rapidly, whereas tetracarboxylates 3 and 4
did very slowly. For instance, complexation of sulfonyl-
based 4 with TbIII ion took 10 h to reach equilibrium
under [TbIII]T/[4]T = 1 at 45 ꢁC. The molar ratio plot
for 4 (Fig. 3) shows that 1:1 complex formed in the
range [TbIII]T/[4]T 6 1, whereas 2:3 (=ligand:metal)
complex formed at [TbIII]T/[4]T P 1.5. Though 4 does
not have protonic OH at the phenol moiety, there ap-
peared a new absorption band at 348 nm upon forma-
tion of 1:1 complex, while intensity of the original
band at 290 nm decreased but remained with significant
intensity (Table 1). This strongly suggests that alkylated
phenolic O groups of 4 are still able to coordinate to
TbIII center to significantly perturb energy levels of fron-
tier orbitals of the phenol moiety. By contrast, the 2:3
complex does not have absorption band at 348 nm, sug-
gesting that the phenolic O of 4 does no longer coordi-
nate in the complex. For thia-based ligand 3, 1:1
complex formation took 3 h under [TbIII]T/[3]T 6 1 at
45 ꢁC. The 1:1 complex of TbIII-3 only has absorption
maximum in a short wavelength region at 285 nm rather
than 348 nm as in the case of TbIII-4, suggesting that
phenolic O of ligand 3 does not coordinate to TbIII
Figure 4. Emission (a) and excitation (b) spectra for TbIII complexes
with ligands 1–4. (a) (1) [1] = 2 · 10ꢀ5, [Tb] = 5 · 10ꢀ6, [NEt3] = 5 ·
10ꢀ4 M, kex = 320 nm; (2) [2] = 6 · 10ꢀ6, [Tb] = 1 · 10ꢀ6, [NEt3] = 1 ·
10ꢀ4 M, kex = 345 nm; (3) [3] = 6 · 10ꢀ5, [Tb] = 2 · 10ꢀ5, [NEt3] =
Table 1. Photophysical properties of 1:1 complexes of TbIII with 1–4
6 · 10ꢀ4 M,
k , ,
ex = 290 nm and (4) [4] = 6 · 10ꢀ6 [Tb] = 1 · 10ꢀ6
Ligands
kmax (nm)
e (Mꢀ1 cmꢀ1
)
U
[NEt3] = 1 · 10ꢀ4 M, kex = 348 nm and (b) (1) [1] = 2 · 10ꢀ6, [Tb] =
1 · 10ꢀ6, [NEt3] = 5 · 10ꢀ5 M, kem = 546 nm; (2) [2] = 2 · 10ꢀ6, [Tb] =
1
2
3
4
320
345
285
294, 348
10,600
13,000
11,500
9300, 6500
0.0379 0.0007
1 · 10ꢀ6
, k ,
[NEt3] = 5 · 10ꢀ5 M, em = 545 nm; (3) [3] = 3 · 10ꢀ6
0.291 0.003
0.00322 0.00012
0.287 0.005
[Tb] = 1 · 10ꢀ6, [NEt3] = 4 · 10ꢀ5 M, kem = 544 s nm and (4) [4] =
2 · 10ꢀ5, [Tb] = 2 · 10ꢀ6, [NEt3] = 2 · 10ꢀ4 M, kem = 544 nm.

824
T. Horiuchi et al. / Tetrahedron Letters 48 (2007) 821–825
Table 2. Emission lifetimes of 1:1 complexes of TbIII with 1–4 in
CH3OH (sH) and CD3OD (sD) and the number of coordinating
methanol (q) (N = 5)
4. Iki, N.; Horiuchi, T.; Oka, H.; Koyama, K.; Morohashi,
N.; Kabuto, C.; Miyano, S. J. Chem. Soc., Perkin Trans. 2
2001, 2219–2225.
5. Oae, S.; Yoshihara, M.; Tagaki, W. Bull. Chem. Soc. Jpn.
1967, 40, 951–958.
6. Matsumiya, H.; Takahashi, Y.; Iki, N.; Miyano, S. J.
Chem. Soc., Perkin Trans. 2 2002, 1166–1172.
7. Iki, N.; Narumi, F.; Fujimoto, T.; Morohashi, N.;
Miyano, S. J. Chem. Soc., Perkin Trans. 2 1998, 2745–
2750.
Ligands
sH (ms)
sD (ms)
q
1
2
3
4
0.183 0.001
1.40 0.01
1.57 0.03
1.47 0.004
0.190 0.001
2.34 0.01
2.71 0.01
2.42 0.01
1.8
2.4
2.2
2.2
8. Lhotak, P. Tetrahedron 2001, 57, 4775–4779.
9. Iki, N.; Morohashi, N.; Narumi, F.; Fujimoto, T.; Suzuki,
T.; Miyano, S. Tetrahedron Lett. 1999, 40, 7337–7341.
10. Compound 4: A mixture of cone 39 (95.3 mg, 0.100 mmol)
in acetonitrile (9 ml) and MCPBA (0.863 g, 5.00 mmol) in
chloroform (8 ml) was heated at reflux for 12 h. After
evaporation to dryness, the solid residue was washed with
chloroform (10 ml) to remove m-chlorobenzoic acid.
Recrystallization from methanol–water mixture afforded
essentially pure sample of 4 (92.3 mg, Yield. 85.4%). Mp
345 ꢁC (decomp.). IR (KBr, cmꢀ1): 3435 (OH), 2966 (CH),
1742 (C@O), 1331 (SO2). 1H NMR (400 MHz, acetone-d6,
300 K) d 1.22 (36H, br s, C(CH3)3), 5.54 (8H, br s, OCH2),
7.96 (8H, br s, ArH). FAB MS (NBA): m/z 1081.4
[M+H]+.
is excited by energy absorbed by and transferred from
the phenol moiety. Luminescence quantum yields (U)
for 1:1 complexes of TbIII with ligands 1–4 were deter-
mined (Table 1). Surprisingly, TCA-based di- (1) and
tetracarboxylate ligands (3) gave complexes with sub-
stantially smaller U values than un-O-functionalized
TCAS did (U = 0.15),4 implying that energy-transfer
efficiency decreased by introducing substituents at the
phenolic O to affect the T1 level responsible for excita-
tion of TbIII center. On the other hand, SO2CA-based
ligands 2 and 4 have much higher quantum yield as com-
pared to TCA-based 1 and 3. Detailed analysis of the
energy levels of the complexes should wait further study,
but at least one can say that SO2CA-based 2 and 4 are
essentially more efficient ligands to transfer the absorbed
energy from its T1 state to 5D4 level of TbIII center than
thia-based 1 and 3. In order to evaluate the ability of
carboxylate ligands 1–4 to shield TbIII center from sol-
vent molecules, we estimated the number of coordinat-
ing methanol (q) for the complex by use of Horrocks
equation, q ¼ 8:4ðsꢀH1 – sDꢀ1Þ,17 where sH and sD are lumi-
nescent lifetimes of the complex in CH3OH and
CD3OD, respectively (Table 2). As can be seen, the
values of sH as well as sD are in the scale of ms order
except for TbIII-1. Irrespective of ligand species, the
number q is ca. 2, which is smaller than the number of
coordinating water to TbIII-TCAS (q = 4.5) and SO2-
CAS (q = 4.2) complexes.4 This implies that introducing
carboxylate groups to TCA and SO2CA scaffolds suc-
cessfully enhances the ability to shield the TbIII center.
11. Crystallographic analysis: A 20-ml glass vial containing
10 ml of a solution of 4 (30 mg) in 1,4-dioxane was left open
in a dark place to allow slow evaporation of the solvent.
After 5 days, colorless crystal suitable for X-ray analysis
was obtained. X-ray data: C64H104O32S4, FW = 1513.75,
monoclinic, C2 (#5), crystal dimensions 0.20 · 0.20 ·
0.02 mm, T = 223 K, Mo-Ka radiation, a = 34.886(5),
˚
b = 16.021(2), c = 15.972(2) A, b = 103.7313(7)ꢁ, V =
8671.7(1) A , Z = 4, DCalcd = 1.159 g/cm3, Full-matrix
3
˚
least-squares, R = 0.103 and Rw = 0.317 for observed
10,814 reflections, GOF = 1.24. Crystallographic data
(excluding structure factors) for the structures in this paper
have been deposited with the Cambridge Crystallographic
Data Centre as supplementary publication numbers CCDC
612397. Copies of the data can be obtained, free of charge,
on application to CCDC, 12 Union Road, Cambridge CB2
1EZ, UK [fax: +44 1223 336033 or e-mail: deposit@
ccdc.cam.ac.uk].
12. Arduini, A.; Fabbi, M.; Mantovani, M.; Mirone, L.;
Pochini, A.; Secchi, A.; Ungaro, R. J. Org. Chem. 1995,
60, 1454–1457.
13. Lang, J.; Dvorakova, H.; Bartosova, I.; Lhotak, P.;
Stibor, I.; Hrabal, R. Tetrahedron Lett. 1999, 40, 373–
376.
14. Lhotak, P.; Stastny, V.; Zlatuskova, P.; Stibor, I.; Michl-
ova, V.; Tkadlecova, M.; Havlicek, J.; Sykora, J. Collect.
Czech. Chem. Commun. 2000, 65, 757–771.
In conclusion, here we prepared new di- and tetracarb-
oxylate ligands (2 and 4, respectively) on a SO2CA scaf-
fold to form 1:1 complex having high luminescence
quantum yields resulting from high efficiency in the en-
ergy transfer process. Considering additional advanta-
ges of TbIII complexes derived from SO2CA-based
ligands such as longer wavelength of absorption maxima
of TbIII-2 and -4 complexes suitable for N2 laser excita-
tion, larger molar absorptivity of TbIII-2 complex, and
high kinetic stability of TbIII-4 complex, 2 and 4 are
promising ligands for analytical applications.
15. Compound 2: A mixture of syn-19 (100 mg, 0.119 mmol)
and MCPBA (616 mg, 3.57 mmol) in chloroform (10 ml)
was heated at reflux for 6 h. After evaporating the mixture
to dryness, solid residue was triturated twice with warm
hexane (10 ml) followed by filtration to remove m-chloro-
benzoic acid. The residue was washed with warm 1,2-
dichloroethane (10 ml) and recrystallized from acetone–
hexane mixture to afford essentially pure sample of 2
(96.5 mg, yield 84.0%). Mp 323 ꢁC (decomp.). IR (KBr,
cmꢀ1): 3378 (OH), 2967 (CH), 1735 (C@O), 1318 (SO2).
1H NMR (400 MHz, acetone-d6, 300 K) d 1.33 (18 H, s,
C(CH3)3), 1.40 (18H, s, C(CH3)3), 4.90 (4H, s, OCH2),
8.18 (4H, s, ArH), 8.43 (4H, s, ArH). FAB MS (NBA):
m/z 965.3 [M+H]+.
References and notes
1. Lanthanide Probes in Life, Chemical and Earth Sciences:
Theory and Practice; Bunzli, J. C. G., Choppin, G. R.,
¨
Eds.; Elsevier, 1989.
2. Sabbatini, N.; Guardigli, M.; Manet, I.; Ziessel, R.
Calixarenes 2001 2001, 583–597.
3. Sato, N.; Yoshida, I.; Shinkai, S. Chem. Lett. 1993, 1261–
1264.
16. Spectroscopic measurement: Stock solutions were prepared
by dissolving ligands (1–4), Tb(NO3)3Æ6H2O, and NEt3 in
methanol. For preparation of TbIII complex, appropriate

T. Horiuchi et al. / Tetrahedron Letters 48 (2007) 821–825
825
amount of solutions of a ligand, NEt3, and TbIII ion were
pipetted into a 25-ml volumetric flask, then the mixture
was made up with methanol and allowed to stand for 1–4 h
at room temperature or 45ꢁC. The sample was subjected to
UV-absorption measurement with a Shimadzu UV-2500-
PC and/or luminescence measurement with a Hitachi
F-4500 spectrometers. Luminescence quantum yields were
estimated by use of the equation U = Us(A/As)N(es/e) (Cs/
C), where A: luminescence intensity obtained by integra-
tion of luminescence spectra based on wavenumber, C:
concentration of luminescent substances, subscript ‘s’ is
standard, quinine sulfate. Us = 0.546 (kex = 366 nm) in
0.5 M H2SO4.18 N = (n/ns)2, n, ns: refractive indices of
methanol (1.3284) and water (1.3330). kex: Excitation
wavelength for estimation of U, which does not necessarily
correspond to absorption maximum. e: Molar absorptivity
of the complex at kex. (kex/ nm, e/Mꢀ1 cmꢀ1) = (340, 5910)
for 1, (360, 9224) for 2, (290, 13,100) for 3, (360, 8010)
for 4.
17. Sabbatini, N.; Guardigli, M.; Lehn, J.-M. Coord. Chem.
Rev. 1993, 123, 201–228.
18. Melhuish, W. H. J. Phys. Chem. 1961, 65, 229–235.