Preparation of fluoroionophores based on diamine-salicylaldehyde derivatives

Article abstract of DOI:10.1016/j.dyepig.2011.10.007

We have developed a series of fluoroionophores from diamine-salicylaldehyde (DS) derivatives and studied their spectra diversity when interacting with a variety of metal cations. Based on our results, 2-((E)-(2-(dimethylamino) ethylimino)methyl)-4-(4-methoxystyryl)phenol (I) can specifically chelate Zn² in both organic and semi-aqueous solution. Furthermore, 2-((E)-(2-(dimethylamino)ethylimino)methyl)-4-((E)-2-(pyridin-4-yl)vinyl) -phenol (II) is an excellent metal recognizer, successfully identifying Zn ² from acetates, chloride and perchlorate salts. In addition, control compound (E)-N-(2-methoxy-5-((E)-2-(pyridin-4-yl)vinyl)benzylidene)-N′, N′-dimethyl-ethane-1,2-diamine (IV) also showed specific metal response to Cd² from Cd(OAc)₂ over other anions. This anionic-dependent metal binding behavior was imparted by the pyridine moiety, which supports extra intermolecular metal-fluorophore complexation. Fluoroionophores I and II were preliminarily tested for cellular metal staining in general fluorescence and ratio fluorescence methods.

Full text of DOI:10.1016/j.dyepig.2011.10.007

Dyes and Pigments 94 (2012) 371e379
Contents lists available at SciVerse ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Preparation of fluoroionophores based on diamine-salicylaldehyde derivatives
,
Wang-Hua Hong a, Chu-Chieh Lin a, Tung-Sheng Hsieh b, Cheng-Chung Chang b
*
a Department of Chemistry, National Chung Hsing University, Taichung 250, Kuo Kuang Road, Taichung 402, Taiwan
b Graduate Institute of Biomedical Engineering, National Chung Hsing University 250, Kuo Kuang Road, Taichung 402, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history:
We have developed a series of fluoroionophores from diamine-salicylaldehyde (DS) derivatives and studied
their spectra diversity when interacting with a variety of metal cations. Based on our results, 2-((E)-
(2-(dimethylamino)ethylimino)methyl)-4-(4-methoxystyryl)phenol (I) can specifically chelate Znþ2 in
both organic and semi-aqueous solution. Furthermore, 2-((E)-(2-(dimethylamino)ethylimino)methyl)-4-
((E)-2-(pyridin-4-yl)vinyl)-phenol (II) is an excellent metal recognizer, successfully identifying Znþ2 from
acetates, chloride and perchlorate salts. In addition, control compound (E)-N-(2-methoxy-5-((E)-2-(pyr-
idin-4-yl)vinyl)benzylidene)-N0,N0-dimethyl-ethane-1,2-diamine (IV) also showed specific metal response
to Cdþ2 from Cd(OAc)2 over other anions. This anionic-dependent metal binding behavior was imparted by
the pyridine moiety, which supports extra intermolecular metal-fluorophore complexation. Fluo-
roionophores I and II were preliminarily tested for cellular metal staining in general fluorescence and ratio
fluorescence methods.
Received 28 August 2011
Accepted 9 October 2011
Available online 20 October 2011
Keywords:
Fluoroionophore
DS (diamine-salicylaldehyde)
Metal recognizer
Anionic-dependent metal binding
Intermolecular complexation
Cellular metal staining
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
[11e13]. Among them, the sensors with turn-on or ratiometric
fluorescence signaling are particularly valuable [14e24]. However,
Among the chemosensors, fluorescent receptors have been
actively investigated because of their high sensitivity [1]. In
particular, fluorescent chemosensors that show a shift in emission
upon binding with analytes are particularly attractive because they
are not only capable of ratiometric sensing of analytes, but also offer
advantages over conventional monitoring of fluorescence intensity
at a single wavelength [2]. A dual emission system (one comes from
the free fluorophore and another from complex) can minimize
measurement errors from photo-transformation, receptor concen-
trations and environmental effects [3e5].
Metal ions are required for proper function of all cells within
every living organism. For example, their careful regulation is
important for aging and disease [6]. Thus, considering both the
environmental and biological important, the development of
materials and methods for the detection of heavy- and transition-
metal cations is a synergistic advancement [7,8]. Sensors based
on metal-induced changes in fluorescence appear to be particularly
attractive as they offer the potential for high sensitivity at low
analyte concentration coupled with their obvious ease of use [9,10].
Numerous fluorescent sensors for cations are known from the
literature. These include chelators and even commercial materials
to the best of our knowledge, there are no definite rules or plat-
forms for screening these metal chelators based on their fluo-
rophore. These considerations include: (i) why do the fluorescent
enhancement phenomena happen in certain solvents or condi-
tions? (ii) Once the metal chelator is sensitive to specific metal
cations, what about its relative anionic effect? (i.e., most authors
only care about whether the fluorophore chelator can bind to Znþ2
;
few authors can illustrate the binding diversities between ZnCl2,
Zn(OAc)2 or Zn(ClO4)2). (iii) Once the metal screening result
from organic solutions is different from semi-aqueous or aqueous
solution, which can be applied to predict cellular staining? Hence,
establishing
a research cascade including screening of fluo-
rophores, determination of detection environments and the
possibility of application to physiological diagnosis is critically
important.
Schiff base is a well-known component of many chelators due to
its electron-donating ability and adjustable coordination cavity
[25,26]. The chelate center, which is constructed from symmetric or
asymmetric ethylenediamine condensing with benzaldehyde
derivatives, can form complexes with metals and show spectrum
diversity. For example, a Cuþ2 and Niþ2 sensor was constructed
from 1-(2-aminoethylamino) anthracene-9, 10-dione and salicy-
laldehyde [27] as well as a 2-[(2-mercaptophenylimino) methyl]
phenol (MPP) sensor that exhibited specific binding to Mgþ2 [28].
The former showed different visible color changes upon addition of
* Corresponding author. Tel.: þ886 4 22840734x24; fax: þ886 4 228 52422.
E-mail address: ccchang555@dragon.nchu.edu.tw (C.-C. Chang).
0143-7208/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.dyepig.2011.10.007

372
W.-H. Hong et al. / Dyes and Pigments 94 (2012) 371e379
metals, while the latter revealed a unique fluorescent response
applied to quantum dot systems [29]. Furthermore, an ion
exchange induced emission turn-on system was achieved when
Hgþ2 was added to Cuþ2-bis (2-hydroxyl-naphthalene-carbox-
aldehyde) benzyl dihydrazone (Cuþ2eH2NB), a double Schiff base
chelator [30].
In this paper, we report the preparation of diamine-
salicylaldehyde (DS) as the chelator moiety from the condensa-
tion of N-N-dimethylethylenediamine and 2-hydroxybenzaldehyde
(Salicylaldehyde) in advance. Two series of fluorophores were
then constructed by conjugating DS chelators with aryl olefins
(Scheme 1). Meanwhile, the hydroxyl of Salicylaldehyde was
substituted with methyl to study its complexing and photophysical
properties as a candidate for fluorescent chemosensors.
where Fu is the quantum yield of unknown, Af is the integrated
area under the corrected emission spectra, A(lex) is the absorbance
area at the excitation wavelength,
h is the refractive index of the
solution and the subscripts u and s refer to the unknown and the
standard, respectively. For the same lex, we chose Quinine sulfate as
the standard, which has a quantum yield
(excitation 350 m).
F
¼ 0.58 in DMSO
2.3. Cell culture conditions and compound incubation
The MCF-7 human breast cancer cells were grown in Dulbecco’s
modified Eagle’s medium (DMEM) with non-essential amino acid
supplemented with 10% fetal calf serum (FCS). The human lung
adenocarcinoma cell line CL1-0 was grown in RPMI 1640 (Gibco/
InvitrogenÔ, Cat. no. 22400-089) medium containing 10% fetal
bovine serum (FBS). Cell cultures were maintained at 37 ꢁC in
a humidified atmosphere of 95% air and 5% CO2. The cells seeded in
culture plates or dishes were incubated with different concentra-
tions of DS derivatives, whose DMSO stock solutions were diluted
in serum-free medium before use (1/100, v/v).
2. Experimental
2.1. Materials and apparatus
The general chemicals employed in this study were of the best
grade available and were obtained from Acros Organic Co., Merck
Ltd., or Aldrich Chemical Co. and used without further purification.
All solvents were of spectrometric grade. Absorption spectra were
generated using a Thermo Genesys 6 UVevisible spectrophotometer,
and fluorescence spectra were recorded using a HORIBA JOBIN-
YVON Fluoromas-4 spectrofluorometer with a 1-nm band-pass in
a 1-cm cell length at room temperature.
2.4. Cellular imaging
Before the observation of cellular localization, cells were seeded
onto coverslips and incubated for 24 h. On the next day, cells were
incubated with 10
cence images were recorded on a Leica AF6000 confocal fluores-
cence microscopy with a DFC310 FX Digital color camera. The
mM of compound for 4 h, and then the fluores-
l
scanning and fluorescence images were taken under a Leica TCS SP5
confocal fluorescence microscopy. The excitation source was
a 405 nm diode laser. Fluorescence photographs were taken
through related ranges using photomultiplier tubes (PMT).
2.2. Determination of quantum yields
The quantum yields of the DS derivatives were determined
using the following equation [31]:
h
i.h
i
2.5. General procedure for the synthesis of diamine-salicylaldehyde
derivatives (Scheme 1)
Fu
¼
FS ꢀ Afu ꢀ AsðlexsÞ ꢀ hu2
Afs ꢀ AuðlexuÞ ꢀ h2s
Synthesis of the DS derivatives is shown in Scheme 1. First, the
phenol moiety of 5-bromosalicylaldehyde (compound 1a) was
methylated with iodomethane to afford compound 1b. Next,
the samples were subjected to Heck coupling reaction with
4-methoxystyrene or 4-vinylpyridine using Pd(OAc)2 as the cata-
lyst. Finally, Schiff base formation was conveniently performed
with N,N-dimethylethylenediamine in methanol. Details of mate-
rials, synthetic procedures and identifications can be found in the
Supporting Information.
O
R1 = H
1a
1b
OR1
R1 = CH3
Br
(i)
O
O
OR1
OR1
3. Results and discussion
3.1. Molecular design and the focus of investigation
R1 = H
2a
2b
R
1 = H
3a
3b
R1 = CH3
R1 = CH3
N
Herein, we constructed two fluorescent sensors by conjugating
a DS chelator with 4-vinylanisole (compound I) and 4-vinylpyridine
(compound II) (Scheme 1). The DS acted as the scaffold and the
vinyl aryl acted to extend the electronic resonance. In general,
incorporation of 4-vinylpyridine can support extra intermolecular
binding ability to metal ions. That is, II may exhibit more binding
modes than I when chelating with metal ions. Meanwhile, the
hydroxyl group on DS was protected with methyl to generate
the control compounds III and IV (from I and II, respectively). The
purpose was to investigate if the metal-chelating properties were
affected by the excited state intramolecular proton transfer (ESIPT)
effect [29,32] as well as to estimate the contribution of the phenol
group.
O
(ii)
N
(ii)
N
N
N
OR1
OR1
R1 = H
R1 = CH3
II
IV
R
R
1 = H
I
N
1 = CH3
III
O
Another point of view on the compounds shown in Scheme 1 is
that stilbene acts as the core scaffold and ethylene as the p-linker at
Scheme 1. Preparation of DS derivatives: (i) 4-vinyl aryl, Pd(OAc)2, PPh3, Et3N, MeCN,
reflux, 48 h; (ii) N,N-dimethylethylenediamine, MeOH, reflux, 12 h.

W.-H. Hong et al. / Dyes and Pigments 94 (2012) 371e379
373
the center of the structure. The well-known intramolecular charge
transfer (ICT) phenomenon, which is concerned with the donor-
acceptor dipole moment exchange, should therefore be taken into
account. This means that the dipole moment exchange, before and
after chelating metal, would be the criteria to judge spectral
diversity and in most of cases would be similar to protonation of the
chelator. This is why most literatures reports (in Refs. [14e24])
mimic or predict the spectra diversity of a metal-fluorophore
complex through protonation of the free fluorophore. In
summary, we wish to ascertain whether ESIPT or ICT is the domi-
nant factor in the design of a good metal chelator. Thus, the
establishment of a metal screening platform in this manuscript
included: (i) range determination of neutral structure (chelator)
form [27,33]; (ii) solvent effect determination of free form and
protonated form to check ICT variation; (iii) metal screening in
optimal conditions based on results from (i) and (ii); and (iv) metal
screening in under semi-aqueous or aqueous buffer systems to
study the possibility of cellular application.
protonated forms of compounds I and II in variable organic solvents
are presented in Fig. 2 (results of compounds III and IV are pre-
sented in Fig. S1). Regardless of the neutral or acidic condition, the
absorption changes of both compounds I and II were insensitive to
solvent polarity. These results clearly point toward a largely non-
polar character for the ground state. However, the apparent red
shift in the absorption maximum (la max) of II under acidic condi-
tions could be due to protonation of the pyridine moiety, increasing
donor-accept effects. It is thus necessary and reasonable that the
protonated behavior of pyridine should be taken into consideration
since this functional group may be involved in the mechanism of
metal chelation. On the other hand, considerable solvent effects in
the fluorescence maximum (lf max) were observed in all cases. This
hinted there was more polar character of excited states, which
might have predominantly come from the ESIPT of DS derivatives.
For applicable fluorescent sensors, colorful exchange and/or signal
enhancement are required once they bind to their targets. Here, the
solvent effect results in Fig. 2 were evaluated based on the criteria
of apparent fluorescent enhancements. Compound I in DMF and
MeOH and compound II in DMSO, MeOH and MeCN showed special
emission enhancements, so these solvents were initially selected
for the metal-chelating assay.
3.2. Basic spectroscopic properties and species distribution
diagrams
Further information regarding the dipole moment exchange
between neutral and protonated compounds were previously
collected [34]. In this study, the plots of ET(30) vs. Stokes shifts for
these solvents are presented in Fig. 3 with a focus on the slope
difference between neutral and protonated compounds. In an
overall observation of the plots, unlike control compounds III and
IV, whose ICT behaviors were definite and had no apparent
difference between neutral and acidic environments, the plots of
compounds I and II showed apparent and irregular differences
between their neutral and protonated forms. Here, the deviations
of the linear correlations were inferred due to (i) the phenol group
of the DS moiety causing intramolecular hydrogen bonding
(compounds I and II), demanding that the ESIPT effect should be
counted; (ii) the pyridine group causing intermolecular hydrogen
bonding (compound II). Hence, once the compound chelated with
metals, it was temporally assumed that the spectral changes of I
was the result of ESIPT effects, while II was the result of ICT and
ESIPT effects.
The UVevisible absorption and fluorescence spectra of the
obtained compounds were measured in DMSO and their optical
data are presented in Table 1. Basically, the absorption energies of
these compounds were located in the ultraviolet region with low
quantum yields. On the other hand, the metal binding process is
usually disturbed by protonation of fluorophore, so it is necessary
to consider the pH effect and find optimal sensing conditions. In
other words, protonation spectra of the chelator moiety in a variety
of solvents might predict or mimic the spectra diversity of this
compound with respect to solvent environment once bound to
ions. Thus, we investigated the spectra responses of compounds
toward pH in H2O/MeOH (3:1) in advance to search for their fluo-
rescent “turned on” states. Based on titration, spectrum deconvo-
lution is important to find the distribution diagram of the neutral
species actually playing the role of chelator. Fig. 1 shows the species
distribution diagram for
a system containing a solution of
compounds I and II upon pH titration. L represents the neutral
compound while deprotonation (protonation) of phenol (Schiff
bases, vinylpyridine) is identified with pKa (pKb) value (Table 1).
3.4. Metal screening
3.3. Solvent effect and fluorescent enhancement of the protonated
compound to mimic metal-chelating spectra
To obtain a quantitative insight into the metal affinity of our
fluorophore candidates, the wavelength changes upon complexa-
tion of various metal ions were determined. The solvent systems
used were compound-dependent due to the criteria from solvent
effects (Fig. 2) and the cation recognition behaviors of compounds
were evaluated from changes in fluorescence intensities upon
addition of metal-containing solutions (the final concentration of
As mentioned above, protonation spectra of the chelator moiety
in various solvents might predict or mimic the spectra diversity of
a compound once binding to metal cations is complete. The
UVevisible absorption and fluorescence spectra of the neutral and
metal in the system was 400
mM). It must be pointed out that it is
Table 1
difficult to maintain the pH conditions of organic solvents, since
most of these metal salts are acidic when dissolved in aqueous
solution. Thus, the control assay was necessary to exclude the
protonation effects from metal salts. For example, Fig. 4a shows the
spectra changes of compound I upon addition of various metal ions
in DMF. In a preliminary analysis of the fluorescence spectra results,
most of metal ions were possibly responsive to I. Following the
criterion of the protonated I (IeHD) curve in Fig. 4a, however, it was
reasonable to assume that compound I chelated Zn2þ and Agþ in
the DMF system. In contrast, control compound III, with the
hydroxyl group absent, revealed no such significant change for any
metal in the fluorescence spectrum under the same conditions. This
revealed the importance of the hydroxyl group for chelation and
showed that the phenol form was responsible for the binding of
Photophysical properties of compounds I and II. Absorption (labs, nm) and emission
(
lem, nm) maxima data were measured from DMSO solution; NMR data (chemical
shift, ppm) were measured from DMSO-d6; pKa and pKb values were determined
from H2O/MeOH (3:1) solutions.
labs
(
3 )
lem
(
Ff)
1H NMR
pKa, pKb
pKa1
ArCH]N
AreOH
pKb1
pKb2
I
II
III
IV
318 (3.74)
325 (3.27)
318 (3.90)
321 (2.99)
420 (0.04)
433 (0.02)
425 (0.10)
390 (0.06)
8.549
8.556
8.632
8.639
13.703
13.398
10.20
9.59
5.10
5.33
5.42
5.51
e
4.70
4.52
labs: absorption maximum (nm);
emission maximum (nm); Ff: quantum yield (quinine sulfate, excitation 350 nm,
3
: extinction coefficient (ꢀ104, Mꢂ1 cmꢂ1); lem
:
F
¼ 0.58 as standard); pKa: deprotonation of phenol; pKb1: protonation of Schiff
base; pKb2: protonation of vinylpyridine.

374
W.-H. Hong et al. / Dyes and Pigments 94 (2012) 371e379
II
I
-
LH
LH2
L
100
50
0
LH3
-
L
L
LH2
L
protonation of
dimethyl amine
vinyl pyridine
schiff base
LH
phenolate
0
2
4
6
8
10 12 14 0
2
4
6
8
10 12 14
pH values
Fig. 1. pH-dependent relative species distribution diagrams of compounds I and II. Deprotonation of phenol and protonations of Schiff bases and vinylpyridine are identified with
pKa and pKb values, respectively. L represents the neutral compounds.
metal through ketoeenol tautomerism from the ESIPT mechanism.
These data also established that the keto form was responsible for
the effective binding of metals [28,35].
response to Clꢂ or OAcꢂ. Combined with the Cdþ2 case of Fig. 4b,
it was clear that the anionic effect should also be included in
the evaluation of the metal screening process. It seemed that
compound II responded to the anionic effect, but not compound I.
Of note, binding properties of control compound IV showed
a more selective response to Cd2þ of Cd(OAc)2 than CdCl2 in DMF
(Fig. S3a) and to Zn2þ of ZnCl2 (emission bands at 450 nm) and Cd2þ
of Cd(OAc)2 (emission bands at 442 nm) in MeCN (Fig. S3b). Even
though we found that these fluorescence enhancements of IV were
all lower then compound II, it was reasonable that these metal
binding affinities or fluorescence enhancements of IV were less
apparent than compound II due to the lack of ESIPT in compound IV,
the same as in the case of control compound III. Nevertheless, based
on these results, it was possible to infer that the pyridine group of
compound II was the factor for anion-dependent metal binding
behavior. The complete metal screening results are listed in the
Table 2. From this table, fluoroionophore candidates with chelation
enhanced fluorescence effects can be found. Consequently, we can
screen solvent relative fluoroionophores that cannot only selec-
tively chelate metal but also recognize this metal from other anions.
In an analogous inference in the case of solvent effects,
compound II was tested for cation selectivity in DMSO and MeCN.
The fluorescence spectrum of II showed marked enhancements at
480 nm with the addition of Zn2þ and less apparent fluorescent
enhancement with Cd2þ (from CdCl2 but not Cd(OAc)2) and Hg2þ in
DMSO with respect to the same concentration of HCl (Fig. 4b). The
binding properties of II toward various metal ions in MeCN (Fig. 4c)
were also evaluated and the results showed more complicated
selectivity to Znþ2 (Fig. 4d). Initially, upon the addition of Zn2þ ions,
only Zn(OAc)2 could efficiently enhance the fluorescence intensity
of II over other anions such as Clꢂ and ClO4ꢂ; the emission band at
480 nm blue shifted to 470 nm and the intensity increased
dramatically (w8.6 fold) when exciting the 315 nm lmax absorption.
While further investigation of the anion responses to the detection
systems found that special emission bands at 550 nm grew when
we excited either 380 or 430 nm, which belong to the unique
absorption of II/Zn(ClO4)2 in MeCN (Fig. S2), there was no similar
II
I
DMSO
DMF
DMSO
1.5
DMF
1.0
0.5
MeCN
MeOH
Acetone
THF
MeCN
MeOH
Acetone
1.0
THF
DCM
DCM
0.5
0.0
0.0
1.0
II/[H+]
I/[H+]
1.5
1.0
0.5
0.0
0.5
0.0
300
400
500
600
300
400
500
600
Wavelength (nm)
Wavelength (nm)
Fig. 2. UVevisible absorption and fluorescence spectra of 20 m
M I and II in organic solvents under neutral and acidic conditions (12 mM [Hþ]); absorption lmax is the excited
wavelength.

W.-H. Hong et al. / Dyes and Pigments 94 (2012) 371e379
375
neutral form
acidic form
14
12
10
8
14
12
10
8
II
I
6
6
4
4
14
14
IV
III
12
10
8
12
10
8
6
6
4
4
35
40
45
ET(30)
50
55
60
35
40
45
50
55
60
ET(30)
Fig. 3. The plots of solvent polarity (ET(30)) vs. Stokes shifts using the results in Fig. 2 and Fig. S1.
3.5. Metal screen in a semi-aqueous system
fluorescence measurements were collected in a semi-aqueous
system at 25 ꢁC (H2O/MeOH ¼ 3:1, pH w 7.0 HEPES buffer,
After systematically looking for selective behavior toward
different metal ions for potential biosensor applications,
[I] ¼ 20
m
M and [M] ¼ 400
mM). As shown in Fig. 5a, free compound
I typically showed emission bands at 370 nm and 550 nm with very
Fig. 4. Metal screening fluorescence spectra of 20 mM (a) I in DMF, (b) II in DMSO, (c) II in MeCN and (d) II in MeCN by exciting the relative absorption of compound in Zn(ClO4)2.
Metal salts: MgCl2, CaCl2, MnSO4, FeCl2, FeCl3, Co(OAc)2, Ni(OAc)2, Cu(OAc)2, ZnCl2, Zn(OAc)2, Zn(ClO4)2, AgNO3, CdCl2, Cd(OAc)2, HgCl2 and Pb(NO3)2 were used in this study. The
excited wavelength is absorption maxima.

376
W.-H. Hong et al. / Dyes and Pigments 94 (2012) 371e379
Table 2
competition between solvation of salts and complexation of metals
to the fluorophore. This hypothesis explained that anion-depen-
dent metal-chelating behaviors should only be observed in organic
solvent.
Optical response of fluoroionophore candidates I, II and IV (20
metal salts (400 M). Numbers in the table represent the relative fluorescent
enhancement with respect to the metal-free system.
mM) to different
m
Znþ2
(Clꢂ
Znþ2
Znþ2
Agþ
Cdþ2
(Clꢂ
Cdþ2
Hgþ2
)
(OAcꢂ
)
ðClOꢂ4 Þ
)
(OAcꢂ
)
3.6. Illustration of binding mode
I (DMF)
I (H2O/MeOH)
II (DMSO)
II (H2O/MeOH)
II (MeCN)
6.6
11.7
9.5
6.5
9.6
9.8
3.9
8.6
5.8
8.7
5.3
11.5
3.2
5.36
4.19
5.48
To gain more insight into the chemosensing properties and
mechanisms of these metal-chelating fluorophores toward metal
ions, absorption and fluorescence titrations of compounds with
metal ions in semi-aqueous buffer solution (H2O/MeOH ¼ 3:1 (v/v),
pH w 7.0 HEPES buffer) were carried out. In the fluorescence
3.5
3.54
7.7a
6.0b
IV (DMF)
IV (MeCN)
3.00
3.30
4.3
emission of I at 20 mM when adding Zn(ClO4)2 from 2 to 80 mM, we
a
Excited wavelength ¼ 430 nm.
Excited wavelength ¼ 380 nm.
b
observe that a weak fluorescence peak shifted from 360 and 560 to
510 nm, gradually increasing in intensity. Under similar conditions,
upon the gradual addition of Cd(OAc)2 from 0 to 4 equivalents, the
lem of II underwent a red shift from 510 nm to 520 nm and a 4-fold
fluorescence enhancement was also observed. The titration
conditions and results are shown in Fig. S4. Regardless of absorp-
tion or emission titration spectra, there was more than one iso-
sbestic point indicating that the binding behaviors were not
a single step.
The strong enhancement of the fluorescence upon interaction
with metals allowed us to study the binding using fluorescence
titration at low concentration. In this type of study, the Scatchard
binding model determines the number(s) of equivalent binding
low intensities under these conditions. Fluorescent enhancement
at 514 nm was observed with addition of Zn2þ, however, no
increase in fluorescent emission was seen with other metal ions.
Moreover, the binding selectivity was anion independent since
the detection systems all in response to ZnCl2, Zn(OAc)2 and
Zn(ClO4)2. Meanwhile, absorption spectra of these metal-
containing solutions showed no apparent differences and stayed
in the UV region with no distinct color change in the solution.
Compound II showed anion independent selectivity to Zn2þ and
Cd2þ under the same conditions with fluorescent enhancement
from 510 nm to 500 nm and 520 nm, respectively. In particular,
there were apparent red shifts in the absorption bands (Fig. 5b) and
this case was similar to the results from solvent effects for
protonated II (Fig. 2).
In this semi-aqueous buffer solution, protonation of pyridine
should be excluded. Hence, the phenomenon of absorption red
shifts was likely the result of the pyridine of compound II
becoming involved in the metal-chelating mode. That is, an extra
intermolecular binding mode was possible for compound II to
chelate metal ions and support extra coordinate binding to anion
to form metal cation-anion-fluorophore complexes. Moreover,
metal-chelating behaviors were proposed to arise from the
sites and the affinities of ligands for those sites (
[36,37]. These titration data were then applied to construct the
binding plots of vs. Cf. The binding ratio is defined as Cb/Cm,
g
/Cf ¼ nK ꢂ
gK)
g
g
where Cf, Cb, and Cm are the molar concentrations of free ligand,
bound ligand, and metal, respectively. In our Scatchard analysis, the
difference between Ct (total concentration of ligand) and Cb gave
the magnitude of Cf. The slightly non-linear Scatchard plots in
Fig. 6a used for obtaining both K and n values became irrelevant
[38] and indicated again that the binding was not a single step.
Here, the K values for I to Znþ2 were 2.91 ꢀ106 and 3.41 ꢀ105 molꢂ1
with n w 1.04 and w1.88. It was thus reasonable that a larger value
of
g
at low concentrations of Znþ2 was probably due to a 2:1 I: Znþ2
Fig. 5. Metal screening absorption (left) and fluorescence (right) spectra of 20
[M] ¼ 400 M). The excited wavelength is absorption maxima.
m
M (a) I and (b) II in a semi-aqueous system at 25 ꢁC (H2O/MeOH ¼ 3:1, pH w7.0 HEPES buffer,
m

W.-H. Hong et al. / Dyes and Pigments 94 (2012) 371e379
377
pyridine moiety of II took part in the chelating process. Conse-
quently, the binding modes of II to metal ions involve the binding
modes seen in Fig. 6b and c (i.e., head-to-tail mode).
3.7. Application as biosensor of cellular imaging
Following the results from semi-aqueous systems, the cell
permeability of compounds I and II was determined. Breast cancer
cell line MCF-7 was incubated with compounds (10
increase of fluorescence intensity in living cells was observed upon
addition of metal salts (Znþ2 or Cdþ2, 50
M) into the medium with
incubation for 1 h at 37 ꢁC. Similar cellular fluorescent enhance-
ment was also observed when cells were exposed to 50 M of metal
salts for 4 h and then further incubated with 10 M compounds for
mM) and the
m
m
m
2 h at 37 ꢁC. The images were obtained on a fluorescence micros-
copy with a digital color camera and excited by a blue light cube
(Fig. 7). Regardless of metal pre-incubation or post-incubation, the
penetration and intracellular metal sensing of I and II were very
clear.
The fluorescence images of intracellular metal ions were
further observed under a confocal microscope. In order to apply
405 nm diode laser irradiation, the cell staining concentrations of
compounds I and II were increased to 20
mM because their
absorption wavelengths were lower then 400 nm. Fig. 8a shows
negligible fluorescence when lung adenocarcinoma cells CL1-
0 were incubated with I for 2 h at 37 ꢁC, but fluorescence
enhancement was observed from the addition of ZnCl2 (50
into the medium with incubation for another 1 h at 37 ꢁC (Fig. 8b).
In addition, evaluation of the scanning spectra suggested that
the lmax of the emission wavelength (w510 nm, Fig. 8c) was very
close to the spectrum of a Znþ2-fluorophore complex in semi-
aqueous solution (Fig. 5a). Similar results were also observed
when II-incubated cells were treated with ZnCl2 (Fig. 8e) and
CdCl2 (Fig. 8f), but the lmax (w530 nm) of the emission wave-
mM)
Fig. 6. (a) Binding constant and binding ratio illustrations for I to zinc and II to zinc
and cadmium. The Scatchard plots were based on the titration results shown in Fig. S4
with elimination of unreasonable data. (b) Possible head-to-head binding mode for
l
compound:metal (blue ball)
compound:metal ¼ 2:1. (For interpretation of the references to color in this figure
¼
2:1. (c) Possible head-to-tail binding mode for
legend, the reader is referred to the web version of this article.)
binding ratio and binding modes as proposed in Fig. 6b (i.e., head-
to-head mode).
length from the
l scanning spectra was longer than that in semi-
aqueous condition (Fig. 5b). These differences were proposed to
be the result of cellular environment of II stay being more
hydrophobic or that aggregation occurred. Nevertheless, the
results suggested that I and II could be used to image intracellular
Znþ2 or Cd2þ in living cells using both general fluorescence and
ratio fluorescence methods. These compounds are thus potentially
useful for the study of the toxicity or bioactivity of metal ions in
living cells.
In another case, the K values for II to Znþ2 were 1.84 ꢀ 106 and
4.52 ꢀ 105 molꢂ1 with n w 1.04 and w2.11; the K values for II to
Cdþ2 were 1.97 ꢀ 106 and 5.08 ꢀ 105 molꢂ1 with n w 1.08 and
w1.96. These results for II revealed that the more than one binding
ratio was dominant due to its larger region of titration and non-
linear Scatchard curves. Together with its apparent red shift of
absorption within the titration process, we inferred that the
Fig. 7. The bright field images (left) and fluorescence images (right) from fluorescence microscopy for (a) IeZnþ2, (b) IIeZnþ2 and (c) IIeCdþ2 systems. MCF-7 cancer cells were
incubated with 10 M compounds (upper) and then 50 M metal
M metal salts were added into the medium and incubation for 1 h at 37 ꢁC (middle). Cells were exposed to 50
salts for 4 h and then further incubated with 10
M compounds for 2 h at 37 ꢁC (bottom). The images were excited by a blue light cube light path through a 370/20 nm band-pass
m
m
m
m
filter and emission was collected and filtered through a 450-nm long-pass filter. (For interpretation of the references to color in this figure legend, the reader is referred to the web
version of this article.)

378
W.-H. Hong et al. / Dyes and Pigments 94 (2012) 371e379
Fig. 8. Confocal fluorescence images of I and II in CL1-0 cells. The excitation light is 405 nm, and the collected wavelength ranged from 450 to 650 nm. Fluorescence images of CL1-0
cells were incubated with I (a) or II (d) for 2 h, washed, and then further incubated with 50 mM ZnCl2 (b), (e) or CdCl2 (f) for 1 h. (c) Normalized l scanning spectra from 420 to
600 nm with 10 nm integration of (b), (e) and (f).
[3] Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2þ indicators with
greatly improved fluorescence properties. J Biol Chem 1985;260:3440e50.
[4] Banthia S, Samanta A. A new strategy for ratiometric fluorescence detection of
transition metal ions. J Phys Chem B 2006;110:6437e40.
4. Conclusions
To develop a set of available fluorescence sensors, optimized
chelating conditions for the DS derivatives studied here were
evaluated by observing their ESIPT and ICT behaviors through pKa
measurement and solvent effect assays. Together with data from
the control compounds, ESIPT was found to be necessary for
a compound to become a metal chelator and ICT was necessary for
a metal chelator to have a color shift and/or signal enhancement.
[5] Jung HJ, Singh N, Jang DO. Highly Fe3þ selective ratiometric fluorescent probe
based on imine-linked benzimidazole. Tetrahedron Lett 2008;49:2960e4.
[6] Lippard SJ, Berg JM. Principles of bioinorganic chemistry. Mill Valley,
California, USA: University Science Books; 1994. 139e175.
[7] Denk W, Strickler JH, Webb WW. 2-photon laser scanning fluorescence
microscopy. Science 1990;248:73e6.
[8] Lichtman JW, Conchello JA. Fluorescence microscopy. Nat Methods 2005;2:
910e9.
[9] Lakowicz JR. Principles of fluorescence spectroscopy. 3rd ed. New York:
Springer; 2006.
[10] Domaille DW, Que EL, Chang CJ. Synthetic fluorescent sensors for studying the
cell biology of metals. Nat Chem Biol 2008;4:168e75.
[11] Palacios MA, Wang Z, Montes VA, Zyryanov GV, Hausch BJ, Jursıkova K, et al.
Hydroxyquinolines with extended fluorophores: arrays for turn-on and
ratiometric sensing of cations. Chem Commun; 2007:3708e10.
[12] Callan JF, de Silva AP, Magri DC. Luminescent sensors and switches in the 21st
century. Tetrahedron 2005;61:8551e88.
Consequently, fluoroionophore I displayed specific binding to Znþ2
,
and fluoroionophore II could not only chelate Znþ2 or Cdþ2 but also
had anion-dependent metal binding behavior in organic solvents.
Finally, I and II were applied to the identification of Znþ2 and Cdþ2
in cells after evaluating their metal binding selectivity in semi-
aqueous systems.
[13] Valeur B, Leray I. Design principles of fluorescent molecular sensors for cation
recognition. Coord Chem Rev 2000;205:3e40;
Acknowledgments
Lee JW, Lee JS, Kang M, Su AI, Chang YT. Visual artificial tongue for quanti-
tative metal cation analysis by an off-the-shelf dye array. Chem Eur J 2006;12:
5691e6.
This work was supported financially by the National Science
Council (NSC 99-2113-M-005-011-MY2) of Taiwan.
[14] Woodroofe CC, Lippard SJ. A novel two-fluorophore approach to ratiometric
sensing of Zn2þ. J Am Chem Soc 2003;125:11458e9.
[15] Royzen M, Durandin A, Young Jr VG, Geacintov NE, Canary JW. A sensitive
probe for the detection of Zn(II) by time-resolved fluorescence. J Am Chem Soc
2006;128:3854e5.
Appendix. Supplementary material
[16] Maruyama S, Kikuchi K, Hirano T, Urano Y, Nagano T. A novel, cell-permeable,
fluorescent probe for ratiometric imaging of zinc ion. J Am Chem Soc 2002;
124:10650e1.
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.dyepig.2011.10.007.
[17] Qi X, Jun EJ, Xu L, Kim SJ, Hong JSJ, Yoon YJ, et al. New BODIPY derivatives as
off-on fluorescent chemosensor and fluorescent chemodosimeter for Cu2þ
:
cooperative selectivity enhancement toward Cu2þ
. J Org Chem 2006;71:
References
2881e4.
[18] Nolan EM, Lippard SJ.
A “turn-on” fluorescent sensor for the selective
[1] de Silva AP, Gunaratne HQN, Gunnlaugsson T, Huxley AJM, McCoy CP,
Rademacher JT, et al. Signaling recognition events with fluorescent sensors
and switches. Chem Rev 1997;97:1515e66.
detection of mercuric ion in aqueous media. J Am Chem Soc 2003;125:
14270e1.
[19] Yoon S, Albers AE, Wong AP, Chang CJ. Screening mercury levels in fish with
a selective fluorescent chemosensor. J Am Chem Soc 2005;127:16030e1.
[2] Lakowicz JR. Topics in fluorescence spectroscopy. In: Probe design and
chemical sensing, vol. 4. New York: Plenum Press; 1994.

W.-H. Hong et al. / Dyes and Pigments 94 (2012) 371e379
379
[20] Hirano T, Kikuchi K, Urano Y, Nagano T. Improved fluorescent probes for
zinc, ZnAFs, suitable for biological applications. J Am Chem Soc 2002;124:
6555e62.
[21] Wu Y, Peng X, Guo B, Fan J, Zhang Z, Wang J, et al. Boron dipyrromethene
fluorophore based fluorescence sensor for the selective imaging of Zn(II) in
living cells. Org Biomol Chem 2005;3:1387e92.
[22] Peng X, Du J, Fan J, Wang J, Wu Y, Zhao J, et al. A selective fluorescent sensor
for imaging Cd2þ in living cells. J Am Chem Soc 2007;129:1500e1.
[23] Ha-Thi MH, Penhoat M, Michelet V, Leray I. Highly selective and sensitive
phosphane sulfide derivative for the detection of Hg2þ in an organoaqueous
medium. Org Lett 2007;9:1133e6.
[24] Taki M, Desaki M, Ojida A, Iyoshi S, Hirayama T, Hamachi I, et al. Fluorescence
imaging of intracellular cadmium using a dual-excitation ratiometric che-
mosensor. J Am Chem Soc 2008;130:12564e5.
[25] Lewinski J, Zachara J, Justyniak I, Dranka M. Hydrogen-bond supramolecular
structure of group 13 Schiff base complexes. Coord Chem Rev 2005;249:
1185e99.
[26] Chen HY, Tang HY, Lin CC. Ring-opening polymerization of lactides initiated
by zinc alkoxides derived from NNO-tridentate ligands. Macromolecules
2006;39:3745e52.
[27] Kaur N, Kumar S. Single molecular colorimetric probe for simultaneous
estimation of Cu2þ and Ni2þ. Chem Commun 2007;29:3069e70.
[28] Singh N, Kaur N, Mulrooney RC, Callan JF. A ratiometric fluorescent probe for
magnesium employing excited state intramolecular proton transfer. Tetra-
hedron Lett 2008;49:6690e2.
[29] Singh N, Mulrooney RC, Kaur N, Callan JF. A nanoparticle based chromogenic
chemosensor for the simultaneous detection of multiple analytes. Chem
Commun; 2008:4900e2.
[30] He G, Zhao Y, He C, Liu Y, Duan C. “Turn-on” fluorescent sensor for Hg2þ via
displacement approach. Inorg Chem 2008;47:5169e76.
[31] Vieira Ferreira LF, Costa SMB. Fluorescence quantum yield evaluation of
strongly absorbing dye solutions as a function of the dye concentration.
J Lumin 1991;48e49:395e9.
[32] Hadjoudis E, Mavridis IM. Photochromism and thermochromism of Schiff
bases in the solid state: structural aspects. Chem Soc Rev 2004;33:579e88.
[33] Tang XL, Peng XH, Dou W, Mao J, Zheng JR, Qin WW, et al. Design of a semi-
rigid molecule as a selective fluorescent chemosensor for recognition of Cd(II).
Org Lett 2008;10:3653e6.
[34] Loutfy RO, Arnold BA. Effect of viscosity and temperature on torsional relax-
ation of molecular rotors. J Phys Chem 1982;86:4205e11.
[35] Bhardwaj VK, Singh N, Hundal MS, Hundal G. Mesitylene based azo-coupled
chromogenic tripodal receptors-a visual detection of Ag(I) in aqueous
medium. Tetrahedron 2006;62:7878e86.
[36] Hollemans HJG, Bertlna RM. Scatchard plot and heterogeneity in binding
affinity of labeled and unlabeled ligand. Clin Chem 1975;21:1769e73.
[37] Chang CC, Wu JY, Chien CW, Wu WS, Liu H, Kang CC, et al. A fluorescent
carbazole derivative: high sensitivity for quadruplex DNA. Anal Chem 2003;
75:6177e83.
[38] Chaires JB. Analysis and interpretation of ligand-DNA binding isotherms. Meth
Enzymol 2001;340:3e22.