Bis(indolyl)methane derivatives as highly selective colourimetric and ratiometric fluorescent molecular chemosensors for Cu2+ cations

Article abstract of DOI:10.1016/j.tet.2007.12.025

New probes based on bis(indolyl)methane derivatives only sense Cu²⁺ among other heavy and transition metal (HTM) ions through two different channels: the colourless to orange or purple colour change, that is visible to the naked eye, and a rema

Full text of DOI:10.1016/j.tet.2007.12.025

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Available online at www.sciencedirect.com
Tetrahedron 64 (2008) 2184e2191
www.elsevier.com/locate/tet
Bis(indolyl)methane derivatives as highly selective colourimetric
and ratiometric fluorescent molecular chemosensors for Cu2þ cations
*
Rosario Martınez, Arturo Espinosa, Alberto Tarraga , Pedro Molina
*
´
´
Departamento de Qu´ımica Orga´nica, Facultad de Qu´ımica, Universidad de Murcia, Campus de Espinardo, E-30100 Murcia, Spain
Received 6 November 2007; received in revised form 5 December 2007; accepted 10 December 2007
Available online 15 December 2007
Abstract
New probes based on bis(indolyl)methane derivatives only sense Cu2þ among other heavy and transition metal (HTM) ions through two
different channels: the colourless to orange or purple colour change, that is visible to the naked eye, and a remarkable enhancement of the
fluorescence along with a large red-shift in emission, in a water containing medium (CH3CN/H2O).
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
Thus, copper, on one hand, is important for life but, on the
other hand, is highly toxic to organism. For these reasons, the
past few years have witnessed a number of reports on the design
and synthesis of chromogenic9 and fluorescent chemosensors
for the detection of divalent copper ions. For most of the reported
Cu2þ fluorescent chemosensors, the binding of the metal ion
causes a quenching of the fluorescence emission.10 In contrast,
a few chemosensors in which the binding of a Cu2þ ion causes
an increase in the fluorescence have been reported.11 However,
the low sensitivity and high order of interference by chemically
closely related metal ions have necessitated the development of
new highly selective and sensitive Cu2þ fluoroionophores.12
Among these chemosensors, unfortunately, those that can be
applied in aqueous solutions at neutral pH are still rare mainly
because of the strong hydration ability of Cu2þ in water. Actu-
ally, chemosensors of this kind are always considered to be
much more attractive and efficient in the respect that most
copper-containing samples are near-neutral aqueous systems.13
Cu2þ metal ion is a paramagnetic ion with an empty d shell
and can strongly quench the emission of a fluorophore via
a photoinduced metal-to-fluorophore electron or energy trans-
fer mechanism.9c,14 Fluorescence quenching is not only disad-
vantageous for a high signal output upon recognition but also
hampers temporal separation of spectrally similar complexes
with time-resolved fluorometry.15 Thus, it is of interest that
the recognition of Cu2þ by the chemosensor does not quench
the fluorescence.
The development of fluorescent molecular sensors for metal
ions, especially for cations with biological interest, has always
been of particular interest.1 More specifically, molecular sen-
sors directed towards the detection and measurement of diva-
lent copper ions have enjoyed particular attention. The soft
transition metal ion Cu2þ is third in abundance (after Fe2þ
and Zn2þ) amongst the essential heavy metal ions in the hu-
man body and plays a pivotal role in a variety of fundamental
physiological processes in organisms ranging from bacteria to
mammals.2 On the other hand, Cu2þ can be toxic to biological
systems, alterations in its cellular homeostasis are connected
to serious neurodegenerative diseases, including Menkes and
Wilson diseases,3 familiar amyotropic lateral sclerosis,4 Alz-
heimer’s disease5 and prion diseases.6 In this regard, copper
binding studies in b-amyloid peptide,7 human- and bovine-se-
rum albumins8 have been actively reported. The serum iron
transport protein is known to bind Cu2þ and also the residues
135e155 of the cysteine-rich domain of APP (amyloid precur-
sor protein), a protein highly implicated in Alzheimer’s dis-
ease, are participating as Cu2þ binding site.7
* Corresponding authors. Tel.: þ34 968 367 496; fax: þ34 968 364 149
(P.M.); tel.: þ34 968 367 499; fax: þ34 968 364 149 (A.T.).
´
E-mail addresses: atarraga@um.es (A. Tarraga), pmolina@um.es (P.
Molina).
0040-4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.12.025

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
R. Mart´ınez et al. / Tetrahedron 64 (2008) 2184e2191
2185
Current interest in different heterocyclic ring systems con-
taining a pyrrolic NH group is mostly related to their ability to
act as molecular receptors for anions.16 In this context, indole
derivatives have been reported as anion chemosensors.16e,17
Very recently, the first cation receptor making use of the pyr-
rolic NH group has been reported.18
We wish to report here that by using a structurally simple
motif, meso-arylbis(indolyl)methane, highly selective Cu2þ
sensing can be achieved through two different channels, col-
ourimetric and fluorescent, in aqueous CH3CN.
The chemosensor behaviour of receptors 1aee with several
metal cations (Liþ, Naþ, Kþ, Mg2þ, Ca2þ, Ni2þ, Cu2þ, Zn2þ
,
Cd2þ, Hg2þ, Pb2þ, Sm3þ, Eu3þ, Yb3þ and Lu3þ),22 in CH3CN
or CH3CN/H2O (70/30, v/v), was investigated by UVevis and
fluorescence measurements. All titration studies carried out in
CH3CN/H2O (70/30, v/v) were conducted at pH 7 (0.1 M
HEPES), and the titration experiments were analyzed using
a computer program.23
Studies in the presence of the above-mentioned set of metal
ions indicated that only Cu2þ ions promoted a notable
response in their absorption spectra, and all other metal ions
tested induced negligible responses, allowing unmistakable
identification and quantification. Notably, stepwise addition
of Cu2þ ions induced the appearance of a new and strong ab-
sorption band at 484e518 nm in CH3CN and at 485e517 nm
in CH3CN/H2O (70/30, v/v), reaching its maximum in inten-
sity when 1 or 2 equiv of Cu2þ ions, respectively, were added
(Tables 1 and 2, Fig. 1).
2. Results and discussion
The receptors 1aee were prepared in 89e95% yields by
condensation of indole with the appropriate aryl aldehyde in
methanol in the presence of potassium hydrogensulfate19
(Scheme 1). These compounds were isolated as solids and
were stable both in solid state and in solution.
For these receptors 1aee, the low energy (LE) band was
red-shifted, upon addition of Cu2þ ions, which is responsible
for the change of colour from colourless to orange or purple.
This fact can be used for a ‘naked-eye’ detection of Cu2þ
ions. Simultaneously, well-defined isosbestic points were
also observed, which indicate that a neat interconversion
between the uncomplexed and complexed species occurs.
The resulting titrations fitted nicely to a 1:1 binding model,
and the corresponding Kas and detection limits were calculated
both in CH3CN and CH3CN/H2O solutions, respectively
(Tables 1 and 2).
Assessments of the cation affinities also came from
observing the extent to which the fluorescence intensity of
receptors 1aed was affected in the presence of cations.
Upon addition of small amounts of Cu2þ to a solution of re-
ceptors 1aed in CH3CN, a remarkable intensity enhancement
of the emission bands was observed (Fig. 2). It is worth
mentioning that ligand 1d also exhibits an additional band,
corresponding to the pyrene excimer, at l¼478 nm, which
does not increase during the process of addition of the metal
cation (ESI).
Scheme 1.
The UVevis absorption spectra of receptors 1aee in CH3CN
and CH3CN/H2O (70/30, v/v) are dominated by two strong
absorption bands at 224 and 274e282 nm, respectively, and
a less strong shoulder peak at 290 nm. In addition, receptor 1d
shows the typical pyrene absorption bands20 in the region
241e344 nm. These receptors exhibit a very weak fluorescence
in CH3CN (c¼2.5ꢀ10ꢁ5 M), the excitation spectrum revealing
a lmax¼350 nm as an ideal excitation wavelength (ESI). Their
emission spectra show two structureless bands at 404 and
424 nm, with rather low quantum yield (V¼0.004e0.021).21
The final fluorescence enhancement factors (FEFs) were up
to 1000, as the quantum yields resulted in a notable increase
(60-fold in the case of 1b), and the Stokes shift being evalu-
ated as 3819 cmꢁ1 (Table 3). Such a remarkable shift is useful
Table 1
Data obtained from the UVevis spectra upon titration of ligands 1aee with Cu(OTf)2 in CH3CNa
Ligand
L
L$Cu2þ
Isosbestic points/nm
Kas/Mꢁ1
Detection limit/M
l/nm (3ꢀ10ꢁ3 [Mꢁ1 cmꢁ1])
224 (29.17), 282 (5.55), 290 (sh)
224 (73.30), 282 (14.10), 290 (sh)
l/nm (3ꢀ10ꢁ3 [Mꢁ1 cmꢁ1])
285 (5.58), 398 (2.86), 497 (5.24)
269 (sh), 284 (17.05), 332 (2.89),
438 (sh), 484 (15.36)
1a
1b
218, 232
219, 235
1.3ꢀ105
1.9ꢀ105
4.54ꢀ10ꢁ6
3.54ꢀ10ꢁ6
1c
224 (67.06), 282 (14.99), 290 (sh)
207 (100.08), 275 (18.50), 286 (sh),
488 (21.06), 667 (1.25)
221, 232
7.8ꢀ105
1.2ꢀ105
2.69ꢀ10ꢁ6
4.99ꢀ10ꢁ6
1d
224 (80.66), 241 (65.29), 266 (33.06),
276 (44.81), 329 (25.23), 344 (34.87)
223 (88.30), 240 (sh), 265 (sh),
276 (39.04), 314 (sh), 329 (22.52),
344 (26.17), 512 (10.03), 590 (sh), 730 (1.77)
285 (26.48), 384 (6.16), 518 (17.52)
272, 281, 324, 351
1e
224 (65.40), 274 (27.00), 290 (sh)
219, 229, 264, 281
1.3ꢀ105
3.28ꢀ10ꢁ6
c¼2.5ꢀ10ꢁ5 M.
a

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
2186
R. Mart´ınez et al. / Tetrahedron 64 (2008) 2184e2191
Table 2
Data obtained from the UVevis spectra upon titration of ligands 1aee with Cu(OTf)2 in CH3CN/H2O (7/3)a
Ligand
L
L$Cu2þ
Isosbestic points/nm
Kas/Mꢁ1
Detection
limit/M
l/nm (3ꢀ10ꢁ3 [Mꢁ1 cmꢁ1])
l/nm (3ꢀ10ꢁ3 [Mꢁ1 cmꢁ1])
205 (sh), 295 (sh), 496 (2.27)
1a
1b
1c
1d
224 (29.57), 282 (6.54), 290 (sh)
224 (69.10), 282 (14.54), 290 (sh)
224 (76.53), 282 (19.76), 290 (sh)
224 (74.20), 242 (64.12), 266 (33.72),
277 (43.60), 316 (sh), 329 (24.68),
346 (33.00)
217, 231, 271, 295
219, 232, 277, 294
215, 234, 272, 296
221, 354
3.9ꢀ104
1.7ꢀ104
2.6ꢀ104
4.0ꢀ104
1.13ꢀ10ꢁ5
1.13ꢀ10ꢁ5
7.03ꢀ10ꢁ6
6.11ꢀ10ꢁ6
222 (sh), 262 (sh), 284 (sh), 418 (4.44), 488 (5.07)
204 (81.73), 226 (sh), 261 (19.50), 286 (sh), 485 (7.59)
223 (68.48), 241 (57.44), 266 (25.96), 276 (30.16),
315 (sh), 329 (18.52), 345 (22.52), 509 (5.28)
1e
224 (70.80), 282 (23.44), 290 (sh)
220 (sh), 264 (20.12), 517 (1.32)
213, 231, 265, 302
1.2ꢀ104
1.31ꢀ10ꢁ6
c¼2.5ꢀ10ꢁ5 M.
a
for various studies that require clear-cut differences between
wavelengths for excitation and emission.
The values of the association constants and detection
limits24 were calculated and are summarized in Table 3. It is
noticeable that only Cu2þ from all other metal ions tested is
able to produce a drastic change in the emission spectrum, in-
dicating that this enhancement of the fluorescence is highly
specific for Cu2þ cations.
The response of the fluorescence of receptors 1aed towards
a set of metal cations (see above) was also studied in CH3CN/
H2O (70/30, v/v). Under these conditions, 1aed display two
weak emission bands (lexc¼350 nm) at 391e412 and
436 nm, respectively, with a low quantum yield (V¼0.009e
0.011). Titration experiments demonstrate that only Cu2þ
metal cations yielded progressively a significant decrease in
the 391e412 and 436 nm emission bands and at the same
time, a new emission band centred at 454e516 nm is devel-
oped. In compound 1c, a clear isoemissive point at 454 nm
indicates that only one type of complexing mechanism is in-
volved (Fig. 3). It is noticeable that compound 1d also pres-
ents an additional band at 494 nm, which upon addition of
Cu2þ is blue-shifted at 454 nm. The final fluorescence en-
hancement factors (FEFs) were up to 60 (Table 4), and the
quantum yields resulted in a two- to three-fold increase com-
pared to those of the free receptors. The Stokes shift was eval-
uated to be 6545e9191 cmꢁ1 (Table 4). As Figure 3 shows,
the ratio of the long-wavelength emission at 516 nm to the
short wavelength at 436 nm (I516/I435) in compound 1c in-
Figure 1. Variation of the UVevis spectra of ligand 1b (c¼2.5ꢀ10ꢁ5 M) upon
addition of increasing amounts of Cu(OTf)2 in (a) CH3CN (0, 0.2, 0.4, 0.6, 0.8
and 1 equiv of Cu2þ) and (b) CH3CN/H2O (7/3), (0, 0.5, 1.0, 1.5 and 2.0 equiv
of Cu2þ); arrows indicate the absorptions that increase (up) and decrease
(down) during the experiments.
creases with the addition of increasing amounts of Cu2þ
,
achieving its maximum when 2 equiv of Cu2þ was added.
The red-shifts make the ratiometric measurement possible.
The ratio of the fluorescence intensities of two appropriate
Table 3
Data obtained from the fluorescence spectra upon titration of ligands 1aed
with Cu(OTf)2 in CH3CNa
Ligand
L
L$Cu2þ DV (FEF) Kas/Mꢁ1 Detection Stokes
limit/M
shift/cmꢁ1
lem/nm lem/nm
1a
1b
1c
1d
a
404, 424 404, 424 18 (263)
404, 424 404, 424 60 (935)
2.5ꢀ106 3.86ꢀ10ꢁ6 3819
3.1ꢀ106 2.91ꢀ10ꢁ6 3819
Figure 2. Variation of the fluorescence spectra of ligand 1c (c¼2.5ꢀ10ꢁ5 M in
CH3CN) upon addition of increasing amounts (0, 0.25, 0.5, 0.75 and 1 equiv)
of Cu(OTf)2; arrows indicate the emissions that increase during the
experiments.
404, 424 404, 424 30 (1000) 3.5ꢀ106 4.42ꢀ10ꢁ6 3819
391, 478 404, 424 4 (14)
1.3ꢀ106 3.36ꢀ10ꢁ6 3819
c¼2.5ꢀ10ꢁ5 M.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
R. Mart´ınez et al. / Tetrahedron 64 (2008) 2184e2191
2187
with water. After filtration the remaining solid was found to
be identical to the receptor 1b. Furthermore, the addition of
Cuþ or Cu2þ to a CH3CN solution of compound 2a or 2b,
readily prepared from 1b or 1c, respectively, by oxidation
with DDQ, does not induce neither fluorescence nor optical
changes at all.
A general principle in designing fluorogenic and chromo-
genic chemosensors is based on analyte coordination events,
therefore, both the interaction with the analyte and the change
in colour or fluorescence should be reversible. Then, we have
proved the reversibility of the recognition process by carrying
out an experimental test: to a solution of 1b, in CH2Cl2,
1 equiv of copper(II) triflate was added to obtain the com-
plexed species, and the UVevis spectrum was recorded.
The CH2Cl2 solution of the complex was washed several
times with water until the colour of the solution changed
from orange to colourless. The organic layer was dried and
the optical and 1H NMR spectra were recorded, and they
were found to be the same than those of the free receptor
1b. This experiment was carried out over several cycles, the
optical spectrum was recorded after each step and found to
be recovered on completion of the step, thus demonstrating
the high degree of reversibility of the sensing process
(Fig. 4), inset.
The other metal ions studied exhibited basically no discern-
ible changes at all. So, this fluorescence change can be possi-
bly utilized in devices for the measurement of Cu2þ
concentrations even in the presence of other cations.
The interference in the selective response of 1aed in the
presence of Cu2þ, from the other metal cations tested, was
also studied by using cross-selectivity experiments based on
the comparison of the fluorescence results obtained by addi-
tion to the free ligand or to the complex 1aed$Cu of the other
metals tested (Fig. 4).
Figure 3. Variation of the fluorescence spectra of ligand 1c (c¼2.5ꢀ10ꢁ5 M)
in CH3CN/H2O (7/3) upon addition of increasing amounts (0, 0.5, 1.0, 1.5
and 2 equiv) of Cu(OTf)2; arrows indicate the emissions that increase (up)
and decrease (down) during the experiments.
emission wavelengths provides a more reliable measurement
of the concentration.
The fluorescence titration data indicate an empirical 1:1
stoichiometry for the complexes formed, the estimated associ-
ation constants being 1.0ꢀ105e1.9ꢀ105 Mꢁ1 and the calcu-
lated detection limit being 2.43ꢀ10ꢁ6e9.17ꢀ10ꢁ6 M. The
difference in the association constants obtained from absor-
bance and emission titration is not surprising. While the for-
mer shows binding of the receptor in the ground state, the
latter reflects binding in the excited electronic state.25
The Stokes shift value in acetonitrile is consistent with fluo-
rescent arising from a charge transfer state, which is confirmed
by solvatochromic measurements. The unresolved vibronic
structure of the red-shifted emission bands and the large
Stokes shift in acetonitrile/water of the complexes suggest
that the excited states possess charge transfer character, al-
though the formation of an excimer species in the presence
of water must not be ruled out.26
The stoichiometry of the complex was determined by the
changes in the fluorogenic response of 1aed in the presence
In order to demonstrate that the sensing event is not based
on the formation of 2 by oxidation of the receptor 1b with
Cu2þ 27 we have carried out two experimental tests. In the first
one, a CH3CN solution of the complex 1b$Cu2þ was concen-
trated to dryness and the residue was washed several times
Table 4
Data obtained from the fluorescence spectra upon titration of ligands 1aed
with Cu(OTf)2 in CH3CN/H2O (7/3)a
Ligand
L
L$Cu2þ
DV
(FEF)
Kas/Mꢁ1 Detection
limit/M
Stokes
shift/cmꢁ1
lem/nm
lem/nm
1a
1b
1c
391, 436 400, 503 3 (34) 1.3ꢀ105 7.49ꢀ10ꢁ6 8691
391, 436 391, 513 3 (60) 1.0ꢀ105 9.17ꢀ10ꢁ6 9078
412, 436 412, 436, 2 (46) 1.9ꢀ105 6.06ꢀ10ꢁ6 9191
516
Figure 4. Changes in the fluorescence intensity of 1b$Cu2þ even when there
are other different cations. The term ‘all’ include Liþ, Naþ, Kþ, Mg2þ
,
Ca2þ, Ni2þ, Zn2þ, Cd2þ, Hg2þ, Pb2þ, Sm3þ, Eu3þ, Yb3þ and Lu3þ. Inset:
stepwise complexation/decomplexation cycles, using water as decomplexation
agent, carried out by UVevis analysis.
1d
391, 494 391, 454 2 (3)
1.0ꢀ105 2.43ꢀ10ꢁ6 6545
a
c¼2.5ꢀ10ꢁ5 M.