New fluorescent sensors based on 1h-pyrazolo[3,4-b]quinoline skeleton

Article abstract of DOI:10.1007/s10895-009-0576-6

Novel fluorescing dyes 1,3,4-triphenyl-6-(1,4,7,10-tetraoxa-13-aza- cyclopentadec-13-ylmethyl)-1H-pyrazolo[3,4-b]quinoline (K1) and 2-[(2-hydroxyethyl)-(1,3,4-triphenyl-1H-pyrazolo[3,4-b]quinolin-6-ylmethyl) -amino]ethanol (L1) have been synthesized and i

Full text of DOI:10.1007/s10895-009-0576-6

J Fluoresc (2010) 20:525–532
DOI 10.1007/s10895-009-0576-6
ORIGINAL PAPER
New Fluorescent Sensors Based on 1H-pyrazolo[3,4-b]
quinoline Skeleton
Marek Mac & Tomasz Uchacz & Tomasz Wróbel &
Andrzej Danel & Ewa Kulig
Received: 15 September 2009 /Accepted: 7 December 2009 /Published online: 19 January 2010
#
Springer Science+Business Media, LLC 2010
. .
Keywords Fluorescence Electron transfer Cation
Abstract Novel fluorescing dyes 1,3,4-triphenyl-6-(1,4,7,
10-tetraoxa-13-aza-cyclopentadec-13-ylmethyl)-1H-pyrazolo
[3,4-b]quinoline (K1) and 2-[(2-hydroxyethyl)-(1,3,
4-triphenyl-1H-pyrazolo[3,4-b]quinolin-6-ylmethyl)-amino]
ethanol (L1) have been synthesized and investigated by the
means of steady state and time-resolved fluorescence
techniques. These compounds act as sensors for the
fluorescence detection of small inorganic cations (lithium,
sodium, barium, magnesium and calcium) in solvents of
different polarities (THF and acetonitrile). The mechanism,
which allows application of these compounds as sensors, is an
electron transfer from the electro-donative part of molecule to
the acceptor part (fluorophore), which is retarded upon
complexation of the electro-donative part by inorganic
cations. We found that crown ether-containing compound is
very sensitive to the addition of any investigated ions but
amino alcohol-containing one exhibits better selectivity to the
addition of two-valued cations. Two kinds of the complexes
(LM+ and L2M+) were found in the investigated systems. In
addition, the dyes may be used as fluorescence indicators in
solvents of lower polarity like tetrahydrofuran.
.
indicators Ground state complexation
Introduction
Fluorescent molecular systems, which undergo strong
signal changes upon addition of the external substances,
are called fluorescent sensors. Such compounds are
attractive for fluorescence detection of nonfluorescent
analytes, such as metal ions in biological systems.
Fluorescent sensors for detection of small inorganic cations
such as sodium, potassium, magnesium or calcium were of
subject of many investigations [1, 2]. Such cations are
involved in many biological processes of fundamental
importance for life. Monitoring of these ions occurring in
small amounts in blood and urine is of major importance in
medicine. Normal concentrations of these cations in blood
and urines (data in parentheses) are 143 (125) mM, 5 (65)
mM, 1 (4) mM, and 1.5 (4) mM for Na+, K+, Mg2+ and
Ca2+, respectively [1]. Concentration of the lithium cation
in plasma is estimated at the level of 0.09–0.15 μmol/l [3].
On the other hand, barium salts are known as toxic agents
(classic signs of barium toxicity are repeated profound
hypokalemia, cardiac arrhythmias, respiratory failure, pro-
longed gastrointestinal dysfunction, paralysis, myoclonus,
hypertension, and profound lactic acidosis) [4], the maxi-
mal concentrations of this cation in serum is in the range of
3–20 μg/dl [5]. Colorimetric determination based on
absorptiometric measurements of the complexes formed
between some dyes and these cations began to be popular a
long time ago, especially using the compounds similar to
EDTA. In the last years much attention has been paid to
developing fluorescent sensors.
:
M. Mac (*) T. Uchacz
Faculty of Chemistry, Jagiellonian University,
Ingardena 3,
30-060 Krakow, Poland
e-mail: mac@chemia.uj.edu.pl
T. Wróbel
Student of Faculty of Chemistry, Jagiellonian University,
Krakow, Poland
:
A. Danel E. Kulig
Department of Chemistry, University of Agriculture,
Balicka 122,
31-149 Kraków, Poland

526
J Fluoresc (2010) 20:525–532
Much of them work on the basis of a simple electron
transfer mechanism as follows:
fluorescent sensor. The YES logic gate has a single
input and a single output—when the input (cation) is 0
the output (fluorescence) is 0, and when the input is 1
the output is 1. Such an operation is so trivial that it is
not considered in an electronic context, where it simply
represents an electronic conductor. However, it is not
trivial in more general context of input and output. On a
molecular basis, a lot of examples of YES-type behavior
are known and are used for detection of small particles
in biological systems, but have not been recognized
from a logical viewpoint [12].
The molecules which contain electron donor and
acceptor moieties (EDA systems) strongly interact
with each other in the excited state. Thus, in the
absence of inorganic cations the fluorescence quench-
ing of the fluorescing part of the system via electron
transfer mechanism is observed. However, the recog-
nition moiety (electron donor part of the molecule)
can act as a Lewis base in the ground state. Thus,
upon addition of the Lewis acid to the EDA system
the complex is formed (the electrodonative property
of this part of molecule is strongly diminished). In
consequence, the quenching of the fluorescence is
weaker making the intensity of the fluorescence
depending on the concentration of the inorganic
cation. This situation is depicted in Scheme 1.
It should be noted that working principle of the
indicators might be different from this presented
above. There are known sensors with two pendant
hydrocarbons groups (like naphthalene or pyrene)
containing diazacrown ether between these subunits.
Addition of the cations, bounded to the azacrown,
change the ratio of the excimer–monomer fluores-
cence quantum yields [6–8].
In our paper we would like to present the fluorometric
data of two compounds being derivatives of 1,3,4-
triphenyl-1H-pyrazolo[3,4-b]quinoline (PQ).
Experimental
The synthesis of PQ sensors
Both sensors L1 and K1 can be easily prepared from cheap
starting materials in 3–4 steps. Thus, 6-methyl-1,3,4-
triphenyl-1H-pyrazolo[3,4-b]quinoline 1 was prepared in the
three-component reaction by heating a mixture of p-toluidine,
benzaldehyde and 2,5-diphenyl-2,4-dihydropyrazol-3-one
according to the procedure developed in our laboratory
[13]. The resulting compound was brominated with NBS in
the presence of catalytic amount of AIBN. Bromomethyl
derivative 2 reacts with commercially available 1,4,7,10-
tetraoxa-13-aza-cyclopentadecane and diethanolamine yield-
ing 3 (K1) and 4 (L1) respectively (Scheme 2).
It has been established that in many cases the
indicators undergo 1/1 and 2/1 (ligand/metal ion)
complexation [9–11]. Thus, the following equilibria
should be concerned in the description of the overall
complexation process:
K1
Mþ þ L Ð MLþ
Experimental
K2
Mþ þ 2L Ð ML2
þ
Melting points were determined on a Mel-Temp Apparatus
II (capillary). NMR spectra were recorded on a Varian
(Mercury) spectrometer at 25 °C.
and
K
ML2þ þ Mþ Ð 2MLþ
On the other hand, these types of fluorescent molecular
sensors can also be used as YES logic gate. The working
principle of YES logic gate is the same as EDA
6-Methyl-1,3,4-triphenyl-1H-pyrazolo[3,4-b]quinoline 1
Equimolar amounts (0.02 mol) of p-toluidine, benzaldehyde
and 2,5-diphenyl-2,4-dihydropyrazol-3-one were heated to-
gether in diethylene glycol (10 mL) at 190 °C for 2 h. After
cooling the reaction mixture was diluted with ethanol (20 mL)
and boiled. The yellow crystalline precipitate was filtered off,
dried and recrystallized from toluene/hexane (3:1).
Yellow crystals, yield 33%, mp 208–210 °C.
Recognition
moiety
Recognition
moiety
Fluorophore
cation
Fluorophore
strong emission
weak emission
1H NMR(300 MHz, CDCl3, δppm): 8.60(d, J = 8.6 Hz,
2H, 2,6-H1Ph); 8.17(d, J = 8.4 Hz, 1H, 8-H); 7.64–7.56(m,
4H); 7.59(t, J = 7.9 Hz, 2H,); 7.35–7.04(m, 11H); 2.45(s, 3H,
6-Me).
Scheme 1 Principle of work of the EDA cation fluorescence
indicators

J Fluoresc (2010) 20:525–532
527
Scheme 2 a NBS/CCl4/AIBN;
b K2CO3/AcCN/aza-crown; c
K2CO3/AcCN/diethanoloamine
O
O
N
O
N
O
N
N
3
b
Me
BrCH2
a
N
N
N
N
N
N
c
1
2
HO
N
N
N
N
OH
4
6-Bromomethyl-1,3,4-triphenyl-1H-pyrazolo[3,4-b]
quinoline 2
Yellow crystals, yield 25%, mp133 °C.
1H NMR(300 MHz, CDCl3, δppm): 8.56(d, J = 8.7 Hz,
2H); 8.14(d, J = 8.8 Hz, 1H, 8-H); 7.84(dd, J =
8.8 Hz,1.8 Hz; 1H, 7-H); 7.67(d, J = 1.2 Hz, 1H, 7-H);
7.53(t, J = 7.9 Hz, 2H); 7.30–6.99(m, 11H); 3.71(s, 2H);
3.62–3.53(m, 16H); 2.74(t, 4H).
Pyrazoloquinoline 1 (0.01 mol) was heated with NBS
(0.012 mol) in CCl4 (20 mL) for 12 h. The hot solution was
filtered to remove insoluble residues and evaporated. The
precipitate was crystallized from toluene/hexane (4:1).
Yellow powder, 89%, mp 183–5 °C (dec).
1H NMR(300 MHz, CDCl3, δppm): 8.58(d, J = 8.7 Hz,
2H); 8.23(J = 8.9 Hz,1H,8-H); 7.85(d, J = 1.6 Hz, 1H, 5-H);
7.81(dd, J = 8.9 Hz; 1.6 Hz, 1H, 7-H); 7.58(t, J = 7.1 Hz,
2H); 7.38–7.03(m, 11H); 4.59(s, 2H).
2-[(2-Hydroxyethyl)-(1,3,4-triphenyl-1H-pyrazolo[3,4-b]
quinolin-6-ylmethyl)-amino]etanol 4 (L1)
BrCH2PQ 2 (1 mmol), diethanoloamine (1 mmol) and
anhydrous K2CO3 (1.5 mmol) were heated in acetonitrile
(10 mL) at 80 °C for 5 h. After this time the solid was
filtered off and the solution was cooled. The yellow
crystalline precipitate was filtered off and subjected to
column chromatography (silica gel, Merck 60, 70–230
mesh, using toluene as eluent at the beginning, next a
mixture toluene/ethyl acetate 3:1 and ethyl acetate at the
end of elution).
13C NMR(300 MHz, CDCl3): 150.51; 148.33; 146.89;
144.79; 139.81; 134.18; 133.49; 132.48; 131.55; 130.39;
130.17; 129.11; 129.07; 128.61; 128.08; 127.71; 127.58;
126.67; 125.63; 123.08; 121.04; 115.32; 34.09.
1,3,4-Triphenyl-6-(1,4,7,10-tetraoxa-13-aza-
cyclopentadec-13-ylmethyl)-1H-pyrazolo[3,4-b]
quinoline 3 (K1)
Yellow crystals, mp 143 °C, yield 65%.
1H NMR(300 MHz, CDCl3, δppm): 8.62(d, J = 8.7 Hz,
2H); 8.25(d, J = 8.5 Hz, 1H,8-H); 7.83–7.80(d + s, 2H,
5-H,7-H); 7.61(t, J = 7.9 Hz, 2H); 7.64–7.07(m, 11H); 3.83
(s,2H); 3.62(t,4H); 2.74(t,4H), 2.1(b, 2H, OH).
BrCH2PQ 2 (1 mmol), 1,4,7,10 –tetraoxa-13-aza-cyclo-
pentadecane (1 mmol) and anhydrous K2CO3 (1.5 mmol)
were heated in acetonitrile (10 mL) at 80 °C for 5 h. After
this time the solid was filtered off and the residue was
subjected to column chromatography (silica gel, Merck 60,
70–230 mesh, using toluene as eluent at the beginning, next
a mixture toluene/ethyl acetate 3:1 and ethyl acetate at the
end of elution).
Fluorescence
The solvents: cyclohexane (CHX), dibutyl ether (DBE),
ethyl acetate (EtAc), tetrahydrofuran (THF), glycerol

528
J Fluoresc (2010) 20:525–532
triacetate (GTA), dichloromethane (MeCl2), aliphatic alco-
hols (from methanol to octanol), acetonitrile (ACN),
dimethylformamide (DMF), propylene carbonate (PC) and
dimethyl sulfoxide (DMSO) were of spectroscopic grade
and were used as received (all from Aldrich). All the
solvents did not show any traces of fluorescence. For
fluorescence measurements the solutions of the dyes were
degassed using multiple freeze-pump-thaw cycles. The
sample concentration of the dyes for spectroscopic meas-
urements was ca. 10−5 M (this corresponds to absorbance’s
of ca. 0.2–0.3 at the excitation wavelengths used in the
fluorescence investigations). Lithium, sodium, barium,
magnesium and calcium (tetrahydrate) perchlorate (Aldrich)
were used as received. In an independent measurement it
was found that a small addition of deionised water couldn’t
influence significantly the fluorescence of the acetonitrilic
solution of the dyes (K1 and L1).
Fluorescence measurements were performed on a home-
built spectrofluorimeter and a time-correlated single photon
counting setup. For time-resolved fluorescence measure-
ments a picosecond diode laser (λ = 400 nm, 70 ps pulse
duration) (IBH-UK) was used as the excitation source. For
steady-state fluorescence measurements a 365 or 405-nm
line of medium-pressure mercury lamp was used. The
fluorescence quantum yield measurements were carried out
with quinine sulphate in water (Φfl = 0.55) [14] as an
actinometer. The measurements of the fluorescence inten-
sity upon addition of the perchlorates were done without
degassing of the samples.
The equilibrium constants (K1) for the complexation of
the ligand by alkali cations are calculated by fitting the
Eq. 1 to the relative fluorescence intensities in the presence
of different concentration (cM) of perchlorates [15].
ꢀ
ꢁ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Φð1Þ ꢀ Φð0Þ
2cL
2
ΦðcM Þ ¼ Φð0Þ þ
cL þ cM þ 1=K1 ꢀ ðcL þ cM þ 1=KÞ ꢀ 4cLcM
ð1Þ
where Φ(0), Φ(∞) and Φ(cM) is the fluorescence intensity of
the dye alone, with the large concentration of the salt and
with the cM concentration of the salt, respectively, cL
indicates the analytical concentration of the ligand.
For 2/1 complexation where the binding constant K2 is
defined by the Eq. 2:
Results
Absorption spectra of the dyes are rather insensitive to
solvent polarity. The addition of the perchlorates only
influences the absorption in the UV region of the spectra.
The influence is rather small as pointed out in Fig. 1.
Fluorescence quantum yields of the dyes strongly
depend on solvent polarity. In solvents of low polarities
the fluorescence quantum yields are very high and decrease
with increasing solvent polarity, even though the position of
the maximum is almost not influenced by polarity of the
solvent.
Â
Ã
MLþ2
K2 ¼
ð2Þ
2
½Lꢁ ½Mþꢁ
The Eq. 1 can be converted to a more complex one:
2x
ΦðcM Þ ¼ Φ0 þ ðΦ1 ꢀ Φ0Þ
ð3Þ
cL
50000
40000
30000
20000
10000
0
Where x denotes the concentration of the complex MLþ2 .
This quantity can be obtained by numerical solving of the
equation, using own Fortran procedures:
ax3 þ bx2 þ cx þ d ¼ 0
ð3aÞ
With the following parameters:
a ¼ ꢀ4K2
b ¼ 4K2ðcL þ cM Þ
ð3bÞ
2
c ¼ ꢀðK2cL þ 4K2cLcM þ 1Þ
2
d ¼ K2cL cM
20000
25000
30000
35000
40000
45000
ν/cm-1
The values of the binding constants K1 and K2 were
obtained from fluorescence intensities using nonlinear least-
square fitting coded in FORTRAN.
Fig. 1 Absorption spectra of K1 in acetonitrile without (solid) and
with 0.01 M LiClO4 (dotted)

J Fluoresc (2010) 20:525–532
529
5
4
3
2
1
0
The fluorescence decay functions are monoexponential in
solvents of low polarities and clearly biexponential in solvents
of dielectric constants greater that 7 (i.e. more polar than THF)
i.e. IðtÞ ¼ A1 expðꢀt=t1Þ þ A2 expðꢀt=t2Þ. The analysis of
the decay functions in solvents of larger polarities gives the
long- and short-time characteristics, i.e. fluorescence life-
times τ1 and τ2 and corresponding amplitudes A1 and A2. It
should be noted that the fluorescence decay time τ1 is
weakly solvent polarity dependent (ca. 17 ns) (Fig. 2).
Fluorescence quantum yields and the fluorescence life-
times (τ2) of the dyes as solvent polarity function are
presented in Fig. 3.
1,0
0,8
0,6
0,4
0,2
0,0
τ
0
10 20 30
50 60 70
εs40
Φ
Upon addition of perchlorates to the acetonitrile solutions
of dyes L1 and K1 the fluorescence significantly increases. At
the same time, the amplitude A2 of the fluorescence decay
functions decreases. The fluorescence decay function
becomes single-exponential in the presence of perchlorates
exceeding 10−5 M (except of the compound L1 with sodium
perchlorate). The dependence of the fluorescence quantum
yield on salt concentration is presented in Fig. 4.
For the compound L1 the intensity of the dye changes
due to added salts is presented in Fig. 5.
Figure 6 presents the dependence of the relative
fluorescence of L1 in the presence of Mg(ClO4)2 in
acetonitrile. It has been found that the complex of this
dye with Mg2+ is 2/1 type. For other used salts the 1/1
complexation was observed.
0
10
20
30
40
50
60
70
εs
Fig. 3 Solvent polarity dependence of the fluorescence quantum yields
of the dyes L1(black squares) and K1 (open circles). The data in
alcohols are marked in the dashed box. Inset—the dependence of the
fluorescence lifetimes (τ2) for K1 on solvent polarity. Open circles—the
data in alcohols
The binding constants are presented in Table 2. For L1
with monovalent cations no complexation has been ob-
served up to 0.5 M concentration of the salts. In all cases
the 1/1 complexes have been detected.
Discussion
The equilibrium constants calculated using Eqs. 1 and 2
are collected in Table 1.
We performed the measurements also in tetrahydrofuran
(THF). In this case only 1/1 complexation in all investigat-
ed systems was observed.
The studied molecules do not show significant solvent
polarity influence on the ground state absorption and the
The relative fluorescence intensities with the best fits as
functions of the analytical salts concentration to the Eq. 1
are presented in Fig. 7.
55
50
45
40
35
8
CHX
30
25
20
15
10
5
6
4
2
0
Mg++
6
Na+
THF
4
0
2
4
6
8
10
cM/cL
2
0
0,000
0,002
0,004
0,006
0,008
0,010
c/M
Fig. 4 Salt concentration dependence of the relative fluorescence
quantum yield of K1 upon addition of lithium (open circles), sodium
(black diamonds), calcium (open triangles), barium (black triangles)
and magnesium perchlorates (black squares) in acetonitrile. Dependence
of the relative fluorescence of the complex K1-Na+ and K1-Mg2+ on the
ratio of salt to dye concentration in ACN (inset)
20
30
40
50
τ/ns
60
70
80
Fig. 2 Fluorescence decay functions of K1 in CHX and THF measured
at 450 nm. The pulse profile (λmax = 400 nm) is also depicted

530
J Fluoresc (2010) 20:525–532
60
50
40
30
20
10
0
10
Mg++
Ca++
Ba++
0
0,00
0,01
0,02
0,03
0,04
0,05
1E-6
1E-5
1E-4
log(cs)
1E-3
0,01
salt concentration/M
Fig. 6 Dependence of the relative fluorescence quantum yield of L1
on logarithm of the salt concentration of magnesium (open circles),
calcium (black squares) and barium (open triangles) perchlorates in
ACN. Broken line represents the best fit to the Eq. 1, and solid line
represents the fit according to the Scheme 2 (2/1 complexation—
Eqs. 2–3) for magnesium salt. For other systems the correlation (solid)
lines are calculated by Eq. 1
12
10
8
d
c
6
with the previous observation, namely that the solvent
polarity influences fluorescence quenching. Increasing
solvent polarity decreases the energy of the charge transfer
pair, which is formed in monomolecular charge transfer
process, resulting in more efficient fluorescence quenching
in polar solvents. Some increment in the fluorescence
quantum yield is observed in alcohols, the effect is more
significant for compound L1 than K1. The effect is
certainly caused by the ground state interaction between
the lone pair at the nitrogen atom and hydrogen of the
hydroxyl group of the alcohols. Thus, the electron, which
may be transferred in the excited singlet state to the
pyrazoloquinoline skeleton (PQ), is now engaged in the
hydrogen bonding and the electron donative property of
the recognition moiety is significantly lowered. This is also
reflected in much longer fluorescence decay in higher
alcohols (cf. Fig. 3—inset).
b
4
a
2
20
30
40
50
60
t/ns
Fig. 5 Dependence of the relative fluorescence intensity of L1 on the
molar concentration of magnesium (open triangles), calcium (black
triangles) barium (black squares), lithium (black circles) and sodium
(open circles) perchlorates (top). The correlation lines represent the
predictions given by Eq. 1. Fluorescence decay functions for
compound L1 in the absence (a) and in the presence of 1.6×10−5
M
(b), 3.2×10−5 M (c) and 0.17 M Ca(ClO4)2. The excitation profile is
also presented (bottom)
position of the maximum of fluorescence. However, the
recognition moiety (amino alcohol and crown ether)
influences significantly the fate of the fluorescing state of
the fluorophore. In the excited singlet state the fluorescence
quenching occurs, as can be judged from the solvent
polarity dependence of the fluorescence decay functions
and quantum yields. In solvents of lowest polarity the
quenching is hardly observed but in THF the long-lived
fluorescence component occurs only in trace (cf. Fig. 2).
This behaviour is also observed for aprotic solvents of
larger polarity. The long fluorescence lifetime (τ1) is not
dependent on solvent polarity (it scatters around the value
of 17 ns). Similar data were observed for the parent
molecule i.e. 1,3,4-triphenyl-1H-pyrazolo[3,4-b]quinoline
(PQ) [18]. Fluorescence quantum yields clearly depend on
solvent polarity as can be judged from Fig. 3. They
decrease with increasing solvent polarity. This is consistent
These observations may be useful in designing these
molecules as fluorescence indicators of small inorganic
cations.
The change of the first absorption band upon addition of
the salts is not significant. Only small influence in the deep
UV-region is observed (cf. Fig. 1). Addition of small
amounts (up to 4×10−5 M) of perchlorates (lithium,
sodium, calcium, barium and magnesium) to the solution
of K1 in acetonitrile causes a ca. 8-fold increase of the
fluorescence intensity. It means that even a small concen-
tration of the cations is efficient to complexate the dye and
this complex is stable in the excited singlet state. It means a
large sensitivity of this dye on the presence of any
investigated cation. Thus, this dye cannot be used as a

J Fluoresc (2010) 20:525–532
531
Table 2 Binding constants i.e. equilibrium constants between
different inorganic cations and the compounds K1 and L1 in THF
determined from fluorimetric titration
60
Ca++
Ba++
Na+
50
Cation
(X)
Radius
(Å)
Charge
density
of cation
K1(L1X)/M−1
K1(K1X)/M−1
40
30
20
10
0
Mg++
Li+
Na+
Ba++
Ca++
Mg++
0.59
1.02
1.34
1.0
1.47
1.03
1.49
2.02
3.03
No complexation
44.4
No complexation 955
0.5
1.93×103
1.92×103
Li+
644
546
0.66
383.2
0,000
0,002
0,004
0,006
0,008
0,010
salt concentration/M
Fig. 7 Dependence of the relative fluorescence quantum yield of K1
on the salt concentration of lithium (open triangles), sodium (open
squares), magnesium (open circles), calcium (black diamonds) and
barium (black squares) perchlorates in THF. Solid lines represent the
best fit to the Eq. 1
lifetime of the complex was smaller (19.5 ns) in compar-
ison to that with the addition of the micromolar salt
concentration (21.5 ns).
The compound L1 shows larger sensitivity with respect to
the cation, meaning that the ratio of the binding constant
between Ca2+ and Na+ containing complexes exceeds 1,300.
Also the difference between the binding constants with
barium and calcium exceeds factor 14 (see Table 1). A small
amount of calcium perchlorate (ca. 0.001 M) forms a stable
complex with the dye causing an increase of the fluorescence
intensity of the factor ca. 6. The effect is a little bit smaller
upon addition of barium perchlorate, much smaller in the
presence of lithium salt and almost not observed upon
addition of sodium cation. Magnesium cation, which has the
fluorescing indicator for the cations in acetonitrile. An
interesting behaviour has been observed upon addition of
the salts concentrations which many times exceed the
concentration at which the maximal fluorescence quantum
yield has been reached. Upon successive addition of the salt
a decrease of the fluorescence quantum yield has been
observed. This effect is most pronounced with LiClO4, as
presented in Fig. 8.
Most probably the initially formed 2/1 complex under-
goes consecutive complexation with the excess of the
lithium cations to form 1/1 complex as it has been
mentioned in Introduction. We measured also the fluores-
cence decay function of K1 with the addition of Mg(ClO4)2
and LiClO4 at higher salt concentration. We found that at
higher concentration the fluorescence lifetime is signifi-
cantly decreased (12.1 ns) compared to this at micromolar
salt concentration (15.5 ns). Similar (although not so
pronounced) effect was observed with the addition of
magnesium perchlorate. At higher salt concentration the
40
35
30
3
cLi+=3x10-5
cLi+=0.2M
cLi+=0.0
M
25
20
15
10
5
B
C
A
2
1
0
Table 1 Binding constants i.e. equilibrium constants between
different inorganic cations and the compounds K1 and L1 in
acetonitrile determined from fluorimetric titration
-5
0
5
10 15
t/ns
20 25
30 35
40
K1(L1M)/M−1 or K2(K12M)/M−2
K2(L12M)/M−2
0,00
0,05
0,10
0,15
0,20
Cation Radius
Charge
cLiClO4/M
(X)
(Å) [16] density of
cation [17]
Fig. 8 Dependence of relative fluorescence of K1 on the concentra-
tion of lithium perchlorate in ACN. Inset—Fluorescence decay
functions of K1 without LiClO4 (A) and with addition of the salt (B)
and (C) and their fits to the biexponential (A) and monoexponential
decay function (B and C). The residuals for decay function A are also
plotted. The decay functions are vertically shifted for clarity of the
picture. The decay parameters are as follows: (A) τ1=17.0 ns
(A1=0.07), τ2=0.17 ns (A2=0.37), (B) τ1=14.5 ns, (C) τ1=12.1 ns
Li+
Na+
Ba++ 1.34
Ca++ 1.0
Mg++ 0.66
0.59
1.02
1.47
1.03
1.49
2.02
3.03
589
1.4×108
1.1×109
2.9×109
1.5×109
2.6×109
16.8
1,632
2.32×104
K2=1.5×109

532
J Fluoresc (2010) 20:525–532
largest charge density from the cations, investigated here
forms a 2/1 complex in acetonitrile.
Conclusions
The binding constants (L1M in acetonitrile) correlate
well with the charge density of the cations, which are
engaged in the complex. The larger value of the charge
density implies the larger value of the binding constant of
the complex. Similar trends of the binding constants with
ion size (or ion charge density) were reported in several
papers [6–8, 19]. However, our compounds show a larger
increase of the fluorescence intensity upon addition of
the salts than those proposed by other authors [6–12]. The
reason may be a relatively large fluorescence lifetime of the
parent compound 1,3,4-triphenyl-1H-pyrazolo[3,4-b]quino-
line (PQ) (up to 20 ns, hardly depending on solvent
polarity), thus the efficient fluorescence quenching by the
recognition moiety occurs, making the fluorescence of
the uncomplexed dye low efficient. The complexation
changes the fluorescence properties a lot, and in the
presence of small amounts of the salt an intensive green-
colored fluorescence appears. The electron from the lone
pair of the nitrogen atom of the recognition unit is more or
less engaged in the relatively strong binding with the
cation, which is preserved in the excited singlet state.
From Tables 1 and 2 we can recognize that complexation
in less polar tetrahydrofurane (THF) is less efficient
compared to that in strongly polar acetonitrile. Moreover,
only 1/1 complexes have been detected in all investigated
systems. The reason as such behavior may lie in less
dissociation of the perchlorates in less polar solvents.
Perchlorates are easily soluble even in solvents of lower
polarities and they occur in the form of ionic pairs and
triple ions [20]. The dissociation constant of the ionic pair
(Kd) is strongly dependent on solvent polarity and may be
expressed by Fuoss equation:
Novel fluorescing cation indicators were synthesized and
investigated by means of fluorescence spectroscopy. The
stability of the fluorescing complexes (K1X and L1X) was
determined by fluorescence titration. The crown-containing
system (K1) shows the large sensitivity towards the
presence of small cation concentration, which means that
even micromolar ion concentration of the small inorganic
cations causes a 8-fold increase of the fluorescence
quantum yield. Better selectivity is exhibited by the
chelate-type system (L1), which is very sensitive to the
presence of two-valued ions (Mg2+,Ca2+ and Ba2+). These
systems may be applicative in practice. In less polar THF
better selectivity was observed in both compounds.
Acknowledgments The authors would like to thank Prof. Jan
Najbar for various assistance during the completion of this project
and to Mariusz Kosla for editorial comments.
References
1. Valeur B (2002) Molecular fluorescence. Principles and applica-
tions. Ch. 10. Wiley-VCH, Weinheim
2. Li YQ, Bricks JL, Resch-Genger U, Spieles M, Rettig W (2006) J
Phys Chem A 110:10972
3. Leflon P, Plaquet R, Rose F, Hennon G, Ledeme N (1996) Anal
Chim Acta 327:301
4. Jacobs IA, Taddeo J, Kelly K, Valenziano C (2002) Am J Ind Med
41:285
5. Hung Y-M, Chung H-M (2004) Nephrol Dial Transplant 19:1308
6. Bergamini G, Ceroni P, Balzani V, Cornelissen L, van Heyst J,
Lee S-K, Vögtle F (2005) J Mater Chem 15:2959
7. Licchelli M, Orbelli Biroli A, Poggi A, Sacchi D, Sangermani C,
Zema M (2003) Dalton Trans 4537
8. Bouas-Laurent H, Castellan A, Daney M, Desvergne JP, Guinand
G, Marsau P, Riffaud MH (1986) J Am Chem Soc 108:315
9. Yamauchi A, Hayashita T, Nishizawa S, Watanabe M, Teramae N
(1999) J Am Chem Soc 121:2319
10. Xia WS, Shmehl RH, Li CJ (1999) J Am Chem Soc 121:5599
11. Leray I, Habib-Jiwan J-L, Branger C, J-Ph Soumillion, Valeur B
(2000) J Photochem Photobiol A Chem 135:163
12. Balzani V, Credi A, Venturi M (2003) Molecular device and machines.
A journey into the nanoworld. Wiley-VCH Verlag, Weinheim
13. Chaczatrian K, Chaczatrian G, Danel A, Tomasik P (2003) Polish
J Chem 77:1141
14. Meech SR, Phillips D (1983) J Photochem 23:193
15. Bourson J, Pouget J, Valeur B (1993) J Phys Chem 97:4552
16. Barkici H, Koner AL, Nau WM (2005) Chem Commun 5411
17. Leray I, Lefevre J-P, Delouis J-F, Delaire J, Valuer B (2001) Chem
Eur 7:4590
18. Rechthaler K, Rotkiewicz K, Danel A, Tomasik P, Köhler G
(1997) J Fluoresc 7:301
19. Bourson J, Valeur B (1989) J Phys Chem 93:3871
20. Petrucci S, Eyring EM (1991) J Phys Chem 95:1731
21. Mac M (1997) J Photochem Photobiol A Chem 107:107
22. Parvatalu D, Srivastava AK (2008) J Chem Eng Data 53:933
À
Á
3
Kd ¼
exp zAzBe2="sakT
ð4Þ
4pNAa3
where parameter a is the distance of closest approach of
the ions, zA and zB are charges of the ions and εs is the
dielectric constant of the solvent. This linear dependence of
the dissociation constant of lithium perchlorate on inverse of
dielectric constant was presented in our previous paper [21].
Similar situation is thus expected for other perchlorates.
Lower concentration of the free cations in THF is thus
responsible for smaller equilibrium constant for complexa-
tion of investigated dyes with cations in THF. Moreover, the
perchlorates of divalence cations (Me2+ = Ba2+, Ca2+ and
Mg2+) exist in THF rather as MeClO4 cations [22],
+
therefore the 1/1 complexation seems to be rationalized in
less polar solvents, as it is observed in the investigated
systems.