Analytica Chimica Acta 567 (2006) 189–195

A highly selective fluorescent sensor for Cu2+ based on

2-(2
-hydroxyphenyl)benzoxazole in a poly(vinyl chloride) matrix
Xiao-Bing Zhang a , Jing Peng a , Chun-Lian He b , Guo-Li Shen a , Ru-Qin Yu a,∗
a State Key Laboratory of Chemo/Biosensing and Chemometrics, College of Chemistry and Chemical Engineering,
Hunan University, Changsha, 410082, PR China
b Medical college, Hunan Normal University, Changsha 410006, PR China

Received 27 December 2005; received in revised form 1 March 2006; accepted 8 March 2006
Available online 16 March 2006

Abstract
This paper describes a copper selective optical chemical sensor based on static quenching of the fluorescence of 2-(2
-hydroxyphenyl)benzoxazole
entrapped in a poly(vinyl chloride) (PVC) membrane. The effect of the composition of the sensing membrane was studied, and experimental
conditions were optimized. The sensors exhibit stable response over the concentration range from 4.0 × 10−8 M to 5.0 × 10−5 M Cu2+ at pH
4.0–6.5, and a high selectivity. The response time for Cu2+ with concentration ≤5 × 10−6 M is less than 7 min. The optode can be regenerated using
0.1 M HCl and acetate buffer solution. The sensor has been used for direct measurement of copper content in river water samples with a relative
error less than 4% with reference to that obtained by atomic absorption spectrometry.
© 2006 Published by Elsevier B.V.

Keywords: Intramolecular proton transfer; Fluorescent chemical sensor; Cu(II) ion; Benzoxazole derivative

1. Introduction orders. Thus, the determination of trace amount of copper ions

in various samples is of great importance for environment and
Copper is widely used in electrical and electronic industry. human health.
This metal is used in production of coins, brazing alloys, as well Several methods for the determination of copper ions at
as in making water-supply pipes. It is also an essential trace ele- trace quantity level have been proposed. They include atomic
ment for both plants and animals, including humans. Following absorption spectrometry [5], inductively coupled plasma-mass
zinc and iron, copper ranks the third in abundance in human spectroscopy (ICP-MS) [6], inductively coupled plasma-atomic
bodies among the essential heavy metals [1]. It is required for emission spectrometry (ICP-AES) [7], spectrophotometry [8]
haemoglobin synthesis and acts as an indispensable cofactor and voltammetry [9]. These methods offer good limits of detec-
for several metalloenzymes with oxidase activity. Nevertheless, tion (LODs) and wide working concentration ranges. However,
copper is highly toxic to some organisms such as many bacteria except the last two kinds of methods, the aforementioned tech-
and viruses [2]. Owing to its toxicity to bacteria, elevated con- niques necessitate the use of sophisticated and relatively costly
centrations of copper hamper the self-purification capability of apparatus and require complicated pretreatment procedures, not
the sea or rivers and destroy the biological reprocessing systems suitable for on-line or in-field monitoring. Recent years have
in water. It is also found to be harmful to human at high concen- seen increasing interest in the development of optical chemical
trations. Copper has been suspected to cause infant liver damage sensors for Cu(II) with different chemical transducers [10–16],
in recent years. Both Indian childhood cirrhosis (ICC) [3] and as such “optodes” can offer advantages in terms of size, elec-
non-Indian childhood cirrhosis (NICC) [4] were reported to be trical safety, costs, and the fact that it is easy to perform long
connected with an excessive intake of copper and genetic dis- distance transmission with the application of fiber optics. Most
of these optodes are based on absorbance or reflectance measure-
ments of immobilized colorimetric reagents in various matrices.
∗ Corresponding author. Tel.: +86 731 882 2782. Although fluorescence signaling offers the advantage of high
E-mail address: [email protected] (R.-Q. Yu). sensitivity over absorbance or reflectance signaling, only few

0003-2670/$ – see front matter © 2006 Published by Elsevier B.V.

doi:10.1016/j.aca.2006.03.025
190 X.-B. Zhang et al. / Analytica Chimica Acta 567 (2006) 189–195

optical sensors based on fluorescence are reported for Cu(II) 2. Experimental

determination. Fluorescent sensors for Cu(II) based on cal-
cein [17] or rhodamine derivatives [18,19] as fluorophores have 2.1. Reagents
been published. Unfortunately, these sensors suffer from being
highly unspecific. Werner and co-workers reported fluorescent Twice-distilled water was used throughout all experiments.
sensors based on Lucifer Yellow immobilized on cellulose par- High molecular weight poly(vinyl chloride) (PVC), dibutyl
ticles exhibiting improved selectivity towards Cu(II) [20,21]. phthalate (DBP), diethyl phthalate (DEP), dioctyl sebacate
For these sensors, Hg(II) was found to be an interferent for (DOS), dinonyl phthalate (DNP) and sodium tetraphenyl-
Cu(II) assay. Searching for new fluorophores which would react boron (NaTPB) were purchased from Shanghai Chemical
with Cu2+ with sufficient high selectivity is still an active field Reagents (Shanghai) and used as received. HPBO and 2-
as well as a challenge for the analytical chemistry research (2
-hydroxyphenyl)benzothiazole (HPBT) were obtained from
effort. Examples of recent advances along this line have been Sigma–Aldrich. 2-(2
-Hydroxy-phenyl)-4(3H)-quinazolinone
the design of some new fluorescent probes which show selec- (HPQ) was synthesized according to a literature procedure [39],
tively binding with Cu(II) in aqueous or polar organic solvents and its spectroscopic and physical properties concurred with
[22–26]. published data. Tetrahydrofuran (THF) was freshly distilled
Fluorescent dyes that exhibit excited state intramolecular pro- from sodium chips under argon with benzophenone/ketyl as
ton transfer (ESIPT) reactions have attracted great interest for indicator. Except when specified, other chemicals were of ana-
several decades because such compounds show good photo- lytical reagent grade and used without further purification.
physical properties such as intense fluorescence, large Stokes
shifts and significant photostability [27–29]. Several fluores-
2.2. Apparatus
cent chemical sensors for various analytes based on compounds
undergoing ESIPT reactions have been reported in the past
All fluorescence measurements were carried out on a
decade [30–32]. 2-(2
-Hydroxyphenyl)benzoxazole (HPBO),
Perkin–Elmer LS55 luminescence spectrometer with excita-
a typical compound exhibiting ESIPT reactions (Scheme 1)
tion slit set at 15.0 nm and emission at 3.0 nm. A home-made
with good photophysical properties, was reported to show a
poly(tetrafluoroethylene) flow-cell described elsewhere [37] and
strongly increasing fluorescence intensity after binding with
a bifurcated optical fiber (30 + 30 quartz fibers, diameter 6 mm
Li+ in acetonitrile [33,34] or Zn2+ in aqueous solution [35].
and length 1 m) were used for the Cu2+ measurements. The
In the context of our long-term interest in searching for novel
excitation light was carried to the cell through one arm of the
carriers for sensors [36–38], we try to immobilize HPBO
bifurcated optical fiber and the emission light collected through
derivatives in a PVC matrix and study its fluorescent response
the other. A glass plate (diameter 10 mm) covered with sens-
towards different metal ions in aqueous solution. Different
ing membrane was fixed on the top of the flow chamber by
from those of aforementioned literature reports, our experi-
the mounting screw nut with the membrane contacted with the
mental results show that the optode based on HPBO as a
sample solution. The sample solution was driven through the
fluoroionophore shows a selectively fluorescence quenching
flow-cell by a peristaltic pump (Guokang Instruments, Zhejiang)
towards Cu2+ with a wide working concentration range and high
at a flow rate of 1.4 ml min−1 . Experimental results show that
selectivity, which might be attributed to the different binding
with the flow rate larger than 1.0 ml min−1 , the optode membrane
capability to metal ions of HPBO in PVC matrix comparing
could obtain stable fluorescence signal. The standard solution
to that in solution. The sensor is applied to direct measure-
of Cu2+ was obtained by serial dilution of 1.0 × 10−1 M CuCl2
ment of copper content in river water samples with satisfied
solution and buffered with NaOAc–HOAc (pH 6.0). The pH
results.
measurements were carried out on a Mettler-Toledo Delta 320
pH meter.

2.3. Preparation of optode membrane

The optode membrane solution was prepared by dissolving a

mixture of 3.2 mg of HPBO, 8 mg of NaTPB, 50 mg of PVC and
100 mg of DOS in 1.5 ml of freshly distilled THF. Before use,
the glass plates (diameter 10 mm) were activated by treatment
with concentrated HNO3 for 12 h, followed by 3% HF, and 10%
H2 O2 each for 30 min, then washed with distilled water and
ethanol. A glass plate was then mounted on a spin on device and
then rotated at a frequency of 800 rpm. Using asyringe, 0.1 ml
of the membrane solution was sprayed to the center of the plate.
A membrane of about 4 ␮m thickness was then coated on the
glass slides and dried in ambient air at room temperature for 24 h
Scheme 1. Basic mechanism for ESIPT reaction of HPBO. before use.
 X.-B. Zhang et al. / Analytica Chimica Acta 567 (2006) 189–195 191

2.4. Measurement procedure

Two arms of the bifurcated optical fiber were fixed in the

detecting chamber of the spectrofluorometer to carry the excita-
tion and emission light. The fluorescence intensity was measured
with the maximal excitation wavelength of 333 nm and at the
maximal emission wavelength of 484 nm. The sample solutions
of Cu2+ were fed through the detecting chamber of the flow-
cell by the peristaltic pump at a rate of 1.4 ml min−1 . After each
Fig. 2. Structure of HPQ, HPBO and HPBT.
measurement, the flow-cell was washed first with 0.1 M HCl
solution and then with blank buffer solution (NaOAc–HOAc,
pH 6.0) until the fluorescence intensity of the optode reached Cu2+ , ESIPT reactions of HPBO was blocked and its fluores-
the original blank value. cence quenched. From Fig. 1, one can see that the fluorescence
intensities of the optode membrane decrease with increasing
2.5. UV–vis spectra concentration of Cu2+ , which constitutes the basis for the deter-
mination of Cu2+ with the fluorescent sensor proposed in this
For studying the response mechanism, a chloroform solution paper.
of HPBO (1.0 × 10−5 M) and 0.01 M aqueous CuCl2 solution
were mixed and stirred at room temperature. After 3 h, the 3.2. Optimization of membrane compositions
organic layer was separated. Then its UV–vis spectrum was
recorded with a MultiSpec-1501 spectrophotometer and com- Several optode membranes were prepared using different sen-
pared with that of a chloroform solution of HPBO without CuCl2 sitive materials including HPBO, HPBT and HPQ (Fig. 2) with
solution. DOS as plasticizer to study the effect of sensing materials. The
experimental results are shown in Table 1.
3. Results and discussion From Table 1 we can see that the optode using HPBO shows
obviously better response characteristics than that using HPBT
3.1. Fluorescence quenching of optode by Cu2+ in terms of the working concentration range, while the optode
based on HPQ shows no response toward Cu2+ . This phe-
Fig. 1 shows the fluorescence spectra of the HPBO optode nomenon might be arising from the difference in their structure.
membrane exposed to a solution containing different con- It is surprising that the oxygen atom on the cycle linked with
centrations of Cu2+ , which are recorded at λex = 333 nm, benzo group on HPBO is more favorable for Cu2+ binding com-
λem = 400–600 nm. HPBO is a typical compound exhibiting paring to that of sulfur atom on HPBT, though the carbonyl group
ESIPT reactions with good photophysical properties. Owing on the same cycle of HPQ was expected to show a negative effect
to its ESIPT reactions, HPBO exhibits an intensive fluores- for Cu2+ binding.
cence emission at 484 nm with a large Stokes shift (151 nm) To study the influence of membrane compositions, several
when excited by 333 nm radiation. After selectively binding with optodes with different compositions were prepared using HPBO
as the sensitive material with DOS as plasticizer. Their response
characteristics are listed in Table 2.
As it can be seen, the intermediate content of HPBO (2%)
affords best response for Cu2+ . With HPBO content less than
1%, the response becomes worse. The response is also getting
worse when the content is larger than 3% which results in the
presence of some unsolved HPBO.
To investigation the effect of plasticizers, optodes with differ-
ent plasticizers, DOS, DBP, DNP and DEP were also prepared
using HPBO as the sensing membrane component. The results

Table 1
The effect of sensing materials on the response behaviors of optodes
Optode Sensing Working Fluorescence
material concentration intensity
range (M) change
Fig. 1. Fluorescence emission spectra of the HPBO sensing membrane equilib- 1 HPBO 4.0 × 10−8 to 5.0 × 10−5 520.21–195.27
rium with different concentration of Cu2+ : (1) blank solution; (2) 4.0 × 10−8 M; 2 HPBT 1.0 × 10−6 to 1.0 × 10−5 396.87–371.58
(3) 5.0 × 10−7 M; (4) 1.0 × 10−6 M; (5) 5.0 × 10−6 M; (6) 8.0 × 10−6 M; (7) 3 HPQ No response 591.37–592.63
2.5 × 10−5 M; (8) 2.8 × 10−5 M; (9) 5.0 × 10−5 M; (10) 7.5 × 10−5 M; (11)
1.0 × 10−4 M (Membrane composition: 2% HPBO, 5% NaTPB, 62% DOS and Membrane composition for each optode was: 2% sensing material, 5% NaTPB,
31% PVC). 62% DOS and 31% PVC.
192 X.-B. Zhang et al. / Analytica Chimica Acta 567 (2006) 189–195

Table 2 5%. Our experimental results also show that the intermediate
The effect of amount of HPBO on the response behavior of optodes content of DOS affords best response for Cu2+ . The low content
Optode Membrane Working Fluorescence of DOS causes long response time of the optode, while over-
composition concentration intensity loading of DOS (larger than 70%) might lead to worse stability
HPBO:NaTPB: range (M) change of the membrane. The experimental results show that the optode
DOS:PVC
membrane composition consisting of 2% HPBO, 5% NaTPB,
1 1:5:63:31 1.0 × 10−7 to 5.0 × 10−5 475.24–201.86 62% DOS, and 31% PVC (w/w) provides the best response for
2 2:5:62:31 4.0 × 10−8 to 5.0 × 10−5 520.21–195.27 Cu2+ , which is chosen for further experiments.
3 3:5:61:31 6.5 × 10−8 to 5.0 × 10−5 554.53–341.66
4 4:5:61:30 8.5 × 10−8 to 5.0 × 10−5 512.23–310.13
3.3. Effect of pH on the optode membrane response

Table 3 To study the effect of pH on the optode membrane response to

The effect of plasticizers on the response behavior of optodes copper, the fluorescence intensity versus pH plot for the HPBO
Optode Plasticizer Working concentration range (M) optode was obtained by changing the solution pH with differ-
ent acetate buffer solutions and fixing the Cu2+ concentration at
1 DOS 2.0 × 10−7 to 8.0 × 10−5
2 DBP 1.0 × 10−6 to 1.0 × 10−4 1.0 × 10−5 M (Fig. 3). It can be seen that, in the range of lower
3 DNP 5.0 × 10−7 to 1.0 × 10−4 pH value, the fluorescence intensity of the optode decreased
4 DEP 1.0 × 10−6 to 1.0 × 10−4 with decreasing pH value. This phenomenon might be caused
Membrane composition for each optode was: 2% sensing material, 65% DOS by extraction of H+ from aqueous solution into the optode mem-
and 33% PVC. brane at high acidity, which induces the protonation of the
nitrogen atom on HPBO and block its ESIPT reaction. In the
are shown in Table 3. The optode with DOS as the plasticizer region of higher pH value, the increase of fluorescence inten-
gave the best response for Cu2+ , probably due to its highest sity of the optode is mostly caused by the deprotonation of the
lipophilicity and suitable polarity, which seems to be favorable phenolic oxygen on HPBO under basic condition and result-
for the extraction–complexation of the Cu2+ into the optode ing in the increase of basic fluorescence intensity of the optode
membrane [16]. membrane [34]. Moreover, the partial precipitation of Cu(OH)2
In the experiment, we also find that, with the addition of under strong basic condition might decrease the actual concen-
moderate amount of lipophilic ion, sodium tetraphenylborate tration of Cu2+ in the sample solution, which might also cause
(NaTPB), the response characteristics of the optode membrane the increase of fluorescence intensity of the optode under basic
are improved. The results are shown in Table 4. condition. From Fig. 3 we can see that, in a range of pH from
From Table 4 one can see that the response concentration 4.0 to 6.5, acidity does not affect the determination of Cu2+
range of the optode membrane becomes wider as the amount with the proposed optode. In subsequent experiments, a pH 6.0
of NaTPB in the optode membrane increases, which might NaOAc–HOAc buffer solution was selected for further studies.
be caused by the increasing hydrophilicity owing to the addi-
tion of NaTPB. This seems favorable for Cu2+ to approach the 3.4. Response characteristics of the optode
solution/membrane interface and undergo reversible coordina-
tion/dissociation processes. However, from Table 4 one can also The response curve of HPBO based optode is shown in Fig. 4.
see that the intensity of the fluorescence signal decreased with The working concentration range covers from 4.0 × 10−8 M to
increasing amount of NaTPB. The increase of the response con- 5.0 × 10−5 M of Cu(II) in a pH 6.0 NaOAc–HOAc buffered
centration range seems to reach a balance with the decrease of
the fluorescence intensity signal when 5% NaTPB was added,
and the response concentration range of the optode membrane
becomes narrower when the content of NaTPB is larger than

Table 4
The effect of NaTPB on the response behavior of optodes
Optode Membrane Working Fluorescence
composition concentration intensitya
HPBO:NaTPB: range (M)
DOS:PVC

1 2:0:65:33 2.0 × 10−7 to 8.0 × 10−5 564.90

2 2:1:65:32 1.0 × 10−7 to 8.0 × 10−5 556.07
3 2:3:63:32 8.0 × 10−8 to 7.5 × 10−5 539.33
4 2:5:62:31 4.0 × 10−8 to 5.0 × 10−5 520.21
5 2:6:61:31 7.5 × 10−8 to 5.0 × 10−5 430.19
a Fluorescence intensity of optode membrane contacting with blank buffer
Fig. 3. Effect of pH on the optode membrane response (the concentration of
solution. Cu2+ was fixed at 1.0 × 10−5 M).
 X.-B. Zhang et al. / Analytica Chimica Acta 567 (2006) 189–195 193

Table 5
Interference of different species to the fluorescence determination of Cu2+ with
the proposed optode
Interferent Concentration Fluorescence Relative error % Binding
(M)a (F = F0 − F) b (F/F0 × 100) constant
(M−1 )

Na+ 1.0 × 10−2 8.59 2.29 –c

K+ 1.0 × 10−2 −0.14 −0.04 –c
Li+ 1.0 × 10−2 −11.25 −3.00 –c
Mn2+ 1.0 × 10−2 −5.55 −1.48 –c
Hg2+ 1.0 × 10−2 6.56 1.75 –c
Mg2+ 1.0 × 10−2 3.26 0.82 –c
Ca2+ 1.0 × 10−2 2.81 0.75 –c
Al3+ 1.0 × 10−2 16.16 4.31 48
Ag+ 1.0 × 10−2 13.28 3.54 41
Cd2+ 1.0 × 10−2 −2.66 −0.71 –c
Zn2+ 1.0 × 10−2 −15.90 −4.24 45
Pb2+ 1.0 × 10−2 0.34 0.09 –c
Fig. 4. The plot of log (F/F) as a function of the log[Cu2+ ] (membrane com-
Ni2+ 1.0 × 10−3 1.58 0.42 2.04 × 102
position: 2% HPBO, 5% NaTPB, 62% DOS and 31% PVC; F0 and F denote
Co2+ 1.0 × 10−3 2.78 0.74 2.19 × 102
the fluorescence intensities of the optode membrane in absence and presence of
Fe3+ 1.0 × 10−5 5.36 1.43 2.41 × 103
Cu2+ , respectively).
a The concentration of Cu2+ is fixed at 1.0 × 10−5 M (pH 6.0).
b
solution. The response time for Cu2+
with concentration F and F0 are the fluorescence intensities of the optode contacting
with 1.0 × 10−5 M Cu2+ solution with and without adding the interferant
≤5 × 10−6 M is less than 7 min. The relatively long response
(F0 = 375.85).
time of the optode membrane might be arising from the slow c Not available because of the minor spectral change.
complex reaction between HPBO and Cu2+ in PVC membrane.
The short-time stability of the optode was tested by recording
the fluorescence intensity of 5.0 × 10−6 M Cu(II) over a period by pretreatment of the sample solution with appropriate mask-
of 1 h for eight measurements. A relative standard deviation of ing agents. The experimental results for common metal cations
fluorescence reading of 1.2% was achieved. The response of presented in Table 5 reveal that most alkali, and alkaline earth
membrane to Cu(II) is not fully reversible, but can be regener- metal cations, and many transition metal cations such as Mn2+ ,
ated with HCl solution. The membrane should be washed first Cd2+ , Al3+ and Hg2+ exiting in a concentration of 1.0 × 10−2 M
with 0.1 M HCl solution and then with blank buffer solution do not show significant interfering effect on the Cu2+ assay with
(NaOAc–HOAc, pH 6.0) after each measurement. The optode relative error less than ±5%. Ni2+ and Co2+ presenting in a
membrane can be easily regenerated by twice washing with concentration less than 1.0 × 10−3 M do not show appreciable
0.1 M HCl followed by blank buffer solution at the concentration interfering effect on the Cu2+ assay. However, Fe3+ exiting in
level 4.0 × 10−8 M–1.0 × 10−5 M Cu2+ . When the Cu2+ con- the same concentration show slight interfering effect on the Cu2+
centration level is larger than 5.0 × 10−5 M, three times repeated assay. Further research shows that Fe3+ present in a concentra-
washing is necessary for regeneration. The optode membranes tion less than 1.0 × 10−5 M can be tolerated with relative error
are stored in buffer solution in the dark when not in use. No less than 5%.
detectable change of the intensity of fluorescence signal is found In order to further investigate the selectivity of HPBO based
after one month, implying that the HPBO used is stable in a sensor towards Cu2+ , the binding constants (Ka ) between the
membrane contacting with water. sensor and metal ions were calculated by Stern–Volmer equa-
tion from the fluorescence titration experimental data [40]. The
3.5. Selectivity results are also shown in Table 5. All metal ions were supposed
to form 1:1 complexes with HPBO. The binding constant for
A number of common species was investigated concerning Cu2+ was calculated to be 2.79 × 104 M−1 . As shown in Table 5,
the interference with the fluorescence determination of Cu2+ . except for Fe3+ , the binding constants for most metal ions were
The experiments were carried out by fixing the concentration much smaller than that of Cu2+ . These results implied that HPBO
of Cu2+ at 1.0 × 10−5 M and then recording the change of the based sensor showed high selectivity towards Cu2+ over other
fluorescence intensity before and after adding the interferants metal ions, which makes it feasible for practical applications.
into the Cu2+ solution buffered with a NaOAc–HOAc solu-
tion (pH 6.0). Many common anions, such as Cl− , SO4 2− , and 3.6. Response mechanism of the optode
NO3 2− , which might co-exist with Cu2+ , were investigated for
their effect on the determination of Cu2+ . No observable effects The change of UV–vis spectra of HPBO when contacting
were found in the interference studied. Some other anions, such with Cu2+ solution might shed some light on the response mech-
as S2− , CO3 2− , and OH− , which might react with Cu2+ and pro- anism of the optode. Fig. 5 shows that the absorbance peaks of
duce precipitate, showed expectable obvious interfering effects HPBO occurred at 331, 318, 291, 284, 279, and 271 nm, respec-
on the Cu2+ assay. However, these effects could be eliminated tively, with chloroform as the solvent. While chloroform solution
194 X.-B. Zhang et al. / Analytica Chimica Acta 567 (2006) 189–195

Fig. 5. UV–vis spectra (CHCl3 solution): (1) HPBO; (2) HPBO + Cu2+ .

of HPBO was treated with a 0.01 M aqueous CuCl2 solution for Fig. 6. . Relative fluorescence intensity (α) as a function of log[Cu2+ ]. The
3 h, the intensity of the absorbance peaks at 331 and 318 nm was curves fitting the experimental data were calculated from Eq. (3). (1) m:n = 1:2,
decreased obviously. Moreover, a new peak at 369 nm emerged K = 4.0 × 108 ; (2) m:n = 1:1, K = 4.0 × 104 ; (3) m:n = 2:1, K = 1.60 × 109 ; (䊉)
after treating with Cu2+ . Similar to the case of Li+ with HPBO Data points experimentally obtained.
in acetonitrile [33], we suggest that the new absorbance peak at
369 nm is a sign of the formation of a complex between deproto- [B](aq) can be represented as:
nated HPBO and Cu2+ . The isosbestic point at 339 nm supports
αAn 1
the presence of two species in solution (free HPBO and the = n−1
(3)
HPBO–Cu2+ complex). 1 − αÁ nK[A](org) [B]m
(aq)
The complex ratio of Cu2+ with HPBO can be calculated
from the fluorescence titration experimental data [37,38]. Sup- The response of the optode for different concentrations of Cu2+
pose a complex equilibrium between Cu2+ (B) in the aqueous is shown in Fig. 6. Three curves are calculated using Eq. (3)
sample solution and HPBO (A) in the organic membrane phase with different ratios of Cu2+ :HPBO and K. It can be seen that
is established with formation of a complex with a complex ratio the curve with 1:1 complex ratio and an appropriate K of 4 × 104
of m:n, one has: fits best to the experimental data. Therefore, we suggest that the
response mechanism of the proposed optode is based on the
Kd βA formation of a complex between deprotonated HPBO and Cu2+
mB(aq)
mB(org), mB(org) + nA(org)
An Bm (org)
with a complex ratio of 1:1 (Scheme 2).
i.e.
3.7. Determination of Cu2+ in river water samples
K
mB(aq) + nA(org)
An Bm (org) (1)
The sensing membrane was applied to testing river water
Here Kd , β and K are the distribution coefficient, the apparent samples (obtained from different locations of Xiangjiang river,
complex formation constant of An Bm and the over-all equilib- Changsha) and the results were compared with those given by
rium constant of the reaction, respectively. the atomic absorption spectrometry reference method. In order to
The relative fluorescence intensity α is defined as the ratio reduce the pH influence in the detection, 7 ml of buffer solution
of free A, [A]f , to the total amount of A, [A]t in the membrane was added to 1 ml of the water samples to keep pH value at
phase. It can be experimentally determined by measuring the 6.0. The samples were then analyzed using the proposed sensor.
fluorescence intensity of the optode: Results are shown in Table 6.
From Table 6 one can see that the content of copper ions in
[A]f F − F0 river water samples as determined by the HPBO optical fiber
αA = = (2)
[A]t Fb − F 0 sensor was in good agreement with that obtained by atomic

Here Fb is the fluorescence intensity of the optode in the blank

buffer solution and F0 represents the fluorescence intensity of
the same membrane when HPBO is completely complexed with
Cu2+ . F is the fluorescence intensity of the optode actually mea-
sured when contacting with a Cu2+ solutions of given concen-
tration. The relationship between the α and Cu2+ concentration Scheme 2. Response mechanism of HPBO based optode towards Cu2+ .
 X.-B. Zhang et al. / Analytica Chimica Acta 567 (2006) 189–195 195

Table 6 [8] J.J. Pinto, C. Moreno, M. Garcia-Vargas, Talanta 64 (2004) 562.

Results of Cu2+ determination in river water samples [9] V. Beni, V.I. Ogurtsov, N.V. Bakunin, D.W.M. Arrigan, M. Hill, Anal.
Chim. Acta 552 (2005) 190.
Samples Concentrationa (M) Concentrationb (M) Relative
[10] C. Sanchez-Pedeno, J.A. Ortuno, M.I. Albero, M.S. Garcia, Fresenius J.
error (%)
Anal. Chem. 366 (2000) 811.
River water 1 (9.60 ± 0.10) × 10−6 (9.44 ± 0.06) × 10−6 −1.67 [11] T.J. Sands, T.J. Cardwell, R.W. Cattrall, J.R. Farrell, P.J. Iles, S.D. Kolev,
River water 2 (8.00 ± 0.08) × 10−6 (8.29 ± 0.05) × 10−6 3.83 Sens. Actuators B 85 (2002) 33.
River water 3 (17.09 ± 0.10) × 10−6 (16.47 ± 0.15) × 10−6 −3.63 [12] N. Alizadeh, A. Moemeni, M. Shamsipur, Anal. Chim. Acta 462 (2002)
a
187.
Determined by atomic absorption spectrometry. [13] N. Mahendra, P. Gangaiya, S. Sotheeswaran, R. Narayanaswamy, Sens.
b Average of three determinations found by the proposed method. Actuators B 90 (2003) 118.
[14] P.C.A. Jeronimo, A.N. Araujo, M. Conceicao, B.S.M. Montenegro,
absorption spectrometry with a relative error less than 4%. The C. Pasquini, I.M. Raimundo Jr., Anal. Bioanal. Chem. 380 (2004)
108.
present sensor seems useful for the determination of Cu2+ in real [15] A. Safavi, M. Bagheri, Sens. Actuators B 107 (2005) 53.
samples. [16] M.B. Gholivand, P. Niroomandi, A. Yari, M. Joshagani, Anal. Chim.
Acta 538 (2005) 225.
4. Conclusion [17] L.A. Saari, W.R. Seitz, Anal. Chem. 56 (1984) 810.
[18] F.V. Bright, G.E. Poirer, G.M. Hieftje, Talanta (1988) 113.
[19] D.J.S. Birch, O.J. Rolinski, D. Hatrick, Rev. Sci. Instrum. 67 (1996)
The optode described in this work shows a high selectiv- 2732.
ity for Cu2+ over other common metal ions except for Fe3+ , [20] T. Mayr, T. Werner, Analyst 127 (2002) 248.
which shows slight interfering effect on the Cu2+ assay with [21] T. Mayer, I. Klimant, O.S. Wolfbeis, T. Werner, Anal. Chim. Acta 462
its concentration larger than 1.0 × 10−5 M. The sensors exhibit (2002) 1.
stable response over the concentration range from 4.0 × 10−8 M [22] V. Amendola, L. Fabbrizzi, C. Mangano, H. Miller, P. Pallavicini, A.
Perotti, A. Taglietti, Angew. Chem. Int. Ed. 41 (2002) 2253.
to 5.0 × 10−5 M Cu2+ at pH 4.0–6.5, and it can be regenerated [23] C. Wolf, X. Mei, J. Am. Chem. Soc. 125 (2003) 10651.
readily by washing first with 0.1 M HCl solution and then with [24] R. Meallet-Renault, R. Pansu, S. Amigoni-Gerbier, C. Larpent, Chem.
blank buffer solution. The main disadvantage of this optode is Commun. (2004) 2334.
the relatively long response time (7 min). [25] Q. Wu, E. Anslyn, J. Am. Chem. Soc. 126 (2004) 14682.
[26] M. Royzen, Z. Dai, J.W. Canary, J. Am. Chem. Soc. 127 (2005)
1612.
Acknowledgements [27] D.L. Williams, A. Heller, J. Phys. Chem. 74 (1970) 4473.
[28] J. Catalán, F. Fabero, M.S. Guijarro, R.M. Claramunt, M.D.S. Maria,
Financial support from the National Natural Science Foun- M.C. Foces-Foces, F.H. Cano, J. Elguero, R. Sastre, J. Am. Chem. Soc.
dation of China (Grants 20435003 and 20505008) is gratefully 112 (1990) 747.
acknowledged. [29] L.M. Tolbert, K.M. Solntsev, Acc. Chem. Res. 35 (2002) 19.
[30] K. Choi, A.D. Hamilton, Angew. Chem. Int. Ed. 40 (2001) 3912.
[31] A.S. Klymchenko, A.P. Demchenko, J. Am. Chem. Soc. 124 (2002)
References 12372.
[32] M.M. Henary, Y. Wu, C.J. Fahrni, Chem. Eur. J. 10 (2004) 3015.
[1] D.G. Barceloux, J. Toxicol-Clin. Toxicol. 37 (1999) 217. [33] S.O. Obare, C.J. Murphy, New J. Chem. 25 (2001) 1600.
[2] C. Barranguet, F.P. van den Ende, M. Rutgers, A.M. Breure, M. Grei- [34] W. Qin, S.O. Obare, C.J. Murphy, S.M. Angel, Anal. Chem. 74 (2002)
jdanus, J.J. Sinke, W. Admiraal, Environ. Toxicol. Chem. 22 (2003) 4757.
1340. [35] M.M. Henary, C.J. Fahrni, J. Phys. Chem. A 106 (2002) 5210.
[3] S.H. Hahn, M.S. Tanner, D.M. Danke, W.A. Gahl, Biochem. Mol. Med. [36] X.B. Zhang, C.C. Guo, J.B. Xu, G.L. Shen, R.Q. Yu, Analyst 125 (2000)
54 (1995) 142. 867.
[4] B.P. Zietz, J.D. de Vergara, H. Dunkelberg, Environ. Res. 92 (2003) [37] X.B. Zhang, Z.Z. Li, C.C. Guo, S.H. Chen, G.L. Shen, R.Q. Yu, Anal.
129. Chim. Acta 439 (2001) 65.
[5] N. Pourreza, R. Hoveizavi, Anal. Chim. Acta 549 (2005) 124. [38] X.B. Zhang, C.C. Guo, Z.Z. Li, G.L. Shen, R.Q. Yu, Anal. Chem. 74
[6] J.S. Becker, M.V. Zoriy, C. Pickhardt, N. Palomero-Gallagher, K. Zilles, (2002) 821.
Anal. Chem. 77 (2005) 3208. [39] J.J. Naleway, C.M.J. Fox, D. Robinhold, E. Terpetschnig, N.A. Olson,
[7] J. Otero-Romani, A. Moreda-Pineiro, A. Bermejo-Barrera, P. Bermejo- R.P. Haugland, Tetrahedron Lett. 35 (1994) 8569.
Barrera, Anal. Chim. Acta 536 (2005) 213. [40] Q.Y. Chen, C.F. Chen, Tetrahedron Lett. 46 (2005) 165.