Development of a new fluorescent probe: 1,3,5,7-tetramethyl-8-(4′-aminophenyl)-4,4-difluoro-4-bora-3a, 4a-diaza-s-indacence for the determination of trace nitrite

Article abstract of DOI:10.1016/S1386-1425(03)00329-9

A new fluorescent probe, 1,3,5,7-tetramethyl-8-(4′-aminophenyl)-4,4-difluoro-4-bora-3a, 4a-diaza-s-indacence (TMABODIPY) has been developed for the determination of trace nitrite in terms of the reaction of nitrite with TMABODIPY first in acidic solution and then in alkaline solution to form diazotate, a stable and highly fluorescent reagent. The method offered the advantage of specificity, sensitivity and simplicity. The linear calibration range for nitrite was 8-300nmoll^-1s with a 3σ detection limit of 0.65nmoll ^-1. The proposed method has been applied to monitor the trace nitrite in drinking water and vegetable without extraction.

Full text of DOI:10.1016/S1386-1425(03)00329-9

Spectrochimica Acta Part A 60 (2004) 987–993
Development of a new fluorescent probe:
1,3,5,7-tetramethyl-8-(4ꢀ-aminophenyl)-4,4-difluoro-4-bora-
3a,4a-diaza-s-indacence for the determination of trace nitrite
Mengling Li, Hong Wang, Xian Zhang, Hua-shan Zhang∗
Department of Chemistry, Wuhan University, Wuhan 430072, PR China
Received 15 May 2003; accepted 1 July 2003
Abstract
A new fluorescent probe, 1,3,5,7-tetramethyl-8-(4ꢀ-aminophenyl)-4,4-difluoro-4-bora-3a,4a-diaza-s-indacence (TMABODIPY) has been
developed for the determination of trace nitrite in terms of the reaction of nitrite with TMABODIPY first in acidic solution and then in alkaline
solution to form diazotate, a stable and highly fluorescent reagent. The method offered the advantage of specificity, sensitivity and simplicity.
The linear calibration range for nitrite was 8–300 nmol l−1 s with a 3σ detection limit of 0.65 nmol l−1. The proposed method has been applied
to monitor the trace nitrite in drinking water and vegetable without extraction.
© 2003 Elsevier B.V. All rights reserved.
Keywords: 1,3,5,7-Tetramethyl-8-(4ꢀ-aminophenyl)-4,4-difluoro-4-bora-3a,4a-diaza-s-indacence (TMABODIPY); Fluorescent probe; Nitrite;
Spectrofluorimetry
1. Introduction
Due to the importance of nitrite, great interests in
trace nitrite determination have increased in recent years.
The techniques mainly include spectrophotometry [5,6],
chemiluminescence [7], eletrometry [8], chromatogra-
phy [9,10], kinetic methods [11] and spectrofluorimetry
[12–19]. Among all of the techniques, spectrofluorimet-
ric protocol is widely used because of its simplicity,
sensitivity and low cost. Although a number of differ-
ent fluorescent reagents have been developed, such as
5-aminofluorescein [12], 2-amino-4-chloro-1-hydroxy-
benzene-6-sulphonicbenzidine [13], resorcinol [14],
4-hydroxycoumarin [15], 5,6-diamino-1,3-naphthalene
disulfonicacid [16], 2,3-diaminonaphthalene (DAN) [17],
tryptophan [18] and 2,6-diaminopyridine [19], difluoro-
boradiaza-s-indacences (BODIPY) have not been found to
be used in nitrite determination. BODIPY are an important
class of highly rigidized and polymer-like fluorescent dyes
that have found widespread applications in biochemistry and
molecular biology [20]. They present many advantageous
photonic properties, such as high extinction coefficients,
good photostability, and can be exited at wavelengths around
or over 500 nm. Due to the small Stokes shift, the fluores-
cence quantum yield of BODIPY is high. What is more,
they can be used as the mediators to form many kinds of ex-
Nitrite plays an important role in environmental, food,
industrial and physiological systems. On one hand, as it is
one of the stable metabolites of nitric oxide involved in the
regulation of numerous physiological and biological pro-
cesses as a messenger molecule [1] and nitric oxide can be
transformed into nitrite, a greater stable ion, through the re-
action with oxygen, the measurement of nitrite can there-
fore provide a reliable measurement of nitric oxide action
within the body and can be used as a biomarker that en-
ables physicians to gauge the health of an individual [2]. On
the other hand, nitrite is a widespread contaminant in envi-
ronmental, food, industry and human body. Passage of ni-
trite into the blood stream results in the irreversible conver-
sion of hemogloblin to metahemogloblin with oxygen up-
take and transportation compromised [3]. Furthermore, ni-
trite can produce nitroamine, a carcinogenic material within
the acidic conditions of the stomach and then subsequent
implication in the pathology of gastric cancer [4].
∗
Corresponding author. Tel.: +86-27-87685561;
fax: +86-27-87884130.
E-mail address: hshzhang@whu.edu.cn (H.-s. Zhang).
1386-1425/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S1386-1425(03)00329-9

988
M. Li et al. / Spectrochimica Acta Part A 60 (2004) 987–993
tremely fluorescent materials [21]. Since much of the litera-
ture on them is patent material designed to restrict unlicensed
commercial uses. BODIPY dyes tend to be sold prohibitively
expensive even in small quantities. Therefore, it is highly de-
sirable to synthesize and exploit this kind of reagents suitable
for various analytical application. For this purpose, a new
BODIPY reagent, 1,3,5,7-tetramethyl-8-(4ꢀ-aminophenyl)-
4,4-difluoro-4-bora-3a,4a-diaza-s-indacence (TMABOD-
IPY) has been developed to determine trace nitrite in terms of
the reaction of nitrite with TMABODIPY first in acidic solu-
tion and then in alkaline solution to form diazonate, a stable
and extremely fluorescent product, at the light of the work
of Axelrod and Enge [12]. The method presented the advan-
tage of specificity, sensitivity and simplicity. The proposed
method has been used to monitor the trace nitrite in real sam-
ples such as drinking water and vegetable without extraction.
TMABODIPY was synthesized in our laboratory and its
1.0 × 10−4 mol l−1 solution was prepared with ethanol and
stable at least for a month kept under 4 ◦C.
KH2PO4–Na2HPO4 buffer was prepared by mix-
ing 0.78 mmol l−1 KH2PO4 solution and 0.78 mmol l−1
Na2HPO4 solution to appropriate pH value. A standard ni-
trite solution (2.5 × 10−4 mol l−1) was prepared by drying
sodium nitrite at 110 ◦C for 4 h and dissolving it in water.
Twenty drops of chloroform and a pellet of sodium hydrox-
ide were added to prevent liberation of nitrous acid and to
inhibit bacterial growth and thus make the nitrite solution
stable. The standard solution was prepared weekly and kept
in refrigerator. The working solution was prepared daily by
appropriate dilution.
2.3. Synthesis of 1,3,5,7-tetramethyl-8-(4ꢀ-aminophenyl)-
4,4-difluoro-4-bora-3a, 4a-diaza-s-indacence
(TMABODIPY)
2. Experimental
The synthetic route employed to obtain the new probe is
outlined in Scheme 1. 2,4-Dimethylpyrrole was synthesized
according to the literature [22]. Ninety-five grams of 2,4-
dimethyl-3,5-dicarbethoxypyrrole was mixed with 145 g of
85% potassium hydroxide and heated in a steel bomb under
160 ◦C for 4–5 h. The reaction mixture was steam distilled
and the distillate was extracted several times with ether. The
extract was dried over potassium carbonate. The ether was
removed and the residue was distilled under vacuum, bp
72 ◦C (3.333 × 103 Pa).
2.1. Apparatus
Mass spectra were obtained by means of a VG ZAB-
3F GC–MS instrument (Manchester, UK). FT-IR spectra
were obtained for the products in KBr disks by means of
a Bruker IFS48 instrument (Karlsruhe, Germany). Fluores-
cence spectra were recorded with a Shimadzu RF-5000 spec-
trofluorimeter (Kyoto, Japan). pH was determined by means
of a DF-801 accurate acidimeter (Zhongshan University,
Guangzhou, China).
4-Nitrobenzalhade 0.7556 g (5 mmol) and 2,4-dimethyl-
pyrrole were dissolved in dry CH2Cl2 (50 ml) at room
temperature. The solution was purged with N2 for 30 min.
Several drops of CF3COOH were added to initiate the con-
densation. After 1.5 h, TLC (silica, CH2Cl2) showed that
all of the aldehyde had been consumed. The reaction mix-
2.2. Reagents
Unless otherwise specified, all reagents were of analytical
reagent grade without further purification. All solutions were
prepared in double-distilled water.
NO
2
NO
2
C H OOC
2
5
KOH
F CCOOH
3
+
N , CH Cl
2
2
2
160ºC
COOC H
2
N
H
5
N
H
N
HN
CHO
NH
NH
2
2
1.DDQ
2.BF OEt
10%Pd/C,hydrazine hydrate
N , THF/methanol
NEt
2 ,
3
3
toluene
2
N
N
N
N
B
B
F
F
F
F
Scheme 1. Synthesis of TMABODIPY.

M. Li et al. / Spectrochimica Acta Part A 60 (2004) 987–993
989
ture was washed with saturated aqueous NaHCO3 solution
and water, dried with Na2SO4, filtered and rotary evapo-
rated. The product was dissolved in dry toluene (50 ml) and
2,3-dichloro-5,6-dicyanobenzoquinone (DDQ, 1.0 g) was
added in the solution. After 15 min, triethylamine (6.0 ml)
was added, followed immediately by BF3-etherate (6 ml).
After stirred for 1.5 h at room temperature, the mixture was
washed with water, dried with Na2SO4, filtered and rotary
evaporated.
flask with water and diluted to the mark. Two milliliters of
the solution was transferred to a 10 ml test-tube, and was
tested as described in Section 2.
3. Results and discussion
3.1. Spectrofluorimetric and spectrophotometric
properties
The compound was dissolved in warm THF (30 ml).
Ethanol (30 ml) was added as a co-solvent. After purged
with N2, 10% Pd/C (1.0 g) and 1.0 ml hydrazine were added.
The solution was stirred at reflux under N2 for 30 min,
cooled to 20 ◦C and poured into water. The aqueous mixture
was extracted with CH2Cl2. The extract was washed with
water, and the solvent was removed on a rotary evaporator.
The residue was taken up in CH2Cl2 and applied to a silica
gel column chromatography using hexane–CH2Cl2 (1:2) to
afford the desired compound.
The fluorescence and absorption are shown in Figs. 1
and 2, respectively. The excitation and emission maximum
of TMABODIPY was at 497/510 nm, respectively. The flu-
orescence intensity increased when nitrite was added. Under
the same condition, the maximum absorbance of TMABOD-
IPY was at 497 nm, while that of the product was at 507 nm,
Mass spectra: m/z: 339 (M), 301 (M–2F). IR: (KBr,
800.0
600.0
400.0
200.0
ν cm−1): 3441.3, 3201 (–NH2), 1091.9 (–B–F), 1713.3
=
(–C N), 2924.8 (–CH3).
2.4. Fluorometric analysis
Relative fluorescence (RF) quantum efficiencies were ob-
tained according to the literature [23] by comparing the area
under the corrected emission spectrum of the test sample at
492 nm with that of the solution of fluorescein in 0.1 mol l−1
NaOH whose quantum efficiency is 0.85. The slit width was
both 1.5 nm for excitation and emission, respectively.
0.0
-200.0
400
450
500
550
Wavelength/nm
Fig. 1. Fluorescence spectra of TMABODIPY and the product,
2.5. Determination of nitrite
CTMABODIPY
=
2.0 × 10−6 mol l−1
,
CNO
−
=
2.0 × 10−6 mol l−1
,
2
CHCl = 0.36 mol l−1, CNaOH = 0.52 mol l−1; (—) excitation spectrum of
the product, (· · ·) emission spectrum of the product, (- - -) excitation spec-
trum of the TMABODIPY, (- · -) emission spectrum of the TMABODIPY;
slit width of exitation and emission: 3 and 5 nm, respectively.
Two milliliters of TMABODIPY solution (2.5 ×
10−6 mol l−1) was transferred to a 10 ml test-tube. And a
working solution of 1 ml of nitrite (1.0 × 10−6 mol l−1) was
added followed by 1.0 ml of HCl solution (3.6 mol l−1).
The mixture was diluted to 5.0 ml with water and stood
at 40 ◦C for 30 min. After that, 2.0 ml of NaOH solution
(2.6 mol l−1) was added. Then, the solution was diluted to
the mask with water. Five minutes later, the relative fluo-
rescence intensity was measured at 510 nm with excitation
at 497 nm. The slit width was 3 and 5 nm for excitation and
emission, respectively.
0.30
0.25
0.20
0.15
0.10
2.6. Preparation of samples
0.05
0.00
Drinking water was determined according to the proce-
dure described above directly without further treatment. One
milliliter of the water was transferred to a 10.0 ml test-tube,
and the following operations were as in the procedure in
Section 2.
Fresh vegetable was cleaned and dried, then 10 g of which
was mortared and immerged in 100 ml water. After being
filtered, the solution was transferred to 100 ml volumetric
400
450
500
550
600
Wavelength/nm
Fig. 2. Absorption spectra of TMABODIPY and the product.
CTMABODIPY
=
2.0 × 10−6 mol l−1
,
CNO
−
=
1.0 × 10−3 mol l−1
,
2
CHCl = 0.36 mol l−1, CNaOH = 0.52 mol l−1; (—) absorbance spectrum
of TMABODIPY, (· · ·) absorbance spectrum of the product.

990
M. Li et al. / Spectrochimica Acta Part A 60 (2004) 987–993
N=N+
NH2
-
NO2
HCl
N
N
N
N
B
B
F
F
F
F
N=NO-Na+
N=NOH
OH-
N
N
N
B
N
B
F
F
F
F
Scheme 2. TMABODIPY reacts with nitrite.
with a red shift of 10 nm, showing that there is a new reagent
formed. The proposed product should be a diazonate deriva-
tive according to the literature [12]. The reaction scheme of
TMABODIPY and nitrite is shown in Scheme 2. The dia-
zotate extends the conjugated system compared with TMA-
BODIPY and then the maximum absorbance of which has a
bathochromic shift. Diazotate in alkaline is electron rich as
it carries positive charge. The electron donation of diazotate
leads to the fluorescence intensity of the product increasing
greatly compared to TMABODIPY. What is more, the fluo-
rescence intensity of TMABODIPY keeps stable over a wide
pH range, as indicated in Fig. 3, from 6.0 to 12.0 (Table 1).
Table 1
Spectroscopic properties of TMABODIPY and the product
Solvent
εmax (abs)
λabs,max
(nm)
λem,max
(nm)
φf
× 104
M
−1 cm−1
Water
Ethanol
Acetonitrile
Dichloromethene 1.04
THF
1.11
1.23
1.41
497
473
501
505
503
505
507
510
507
505
509
508
510
521
0.32
0.31
0.29
0.33
0.32
0.37
0.84
1.10
0.94
0.64
Toluene
Watera
a
The spectroscopic properties of the product.
3.2. Effect of TMABODIPY concentration
Fig. 4 indicates the effect of TMABODIPY concentration
on the relative fluorescence intensity. TMABODIPY con-
centration varied from 2 × 10−7 to 8 × 10−7 mol l−1. In this
concentration range, there was a linear relationship between
250
200
150
100
50
300
200
100
0
0
0
2
4
6
8
10
CTMABODIPY/10-7 mol l-1
5
7
9
11
13
Fig. 4. Effect of TMABODIPY amount, CNO − = 1.0 × 10−7 mol l−1
,
pH
2
CHCl = 0.36 mol l−1, CNaOH = 0.52 mol l−1; T = 30 ◦C; reaction time:
Fig. 3. Effect of pH on TMABODIPY, CTMABODIPY = 5.0×10−7 mol l−1
.
60 min.

M. Li et al. / Spectrochimica Acta Part A 60 (2004) 987–993
991
300
250
200
150
100
50
250
200
150
100
50
0
0.1
0.2
0.3
0.4
0.5
0.6
CHCl/mol l-1
0
20
40
60
80
100
T/min
Fig. 5. Effect of HCl amount, CTMABODIPY = 5.0 × 10−7 mol l−1
,
CNO − = 1.0 × 10−7 mol l−1, CNaOH–CHCl = 0.08 mol l−1; T = 30 ◦C;
Fig. 7. Effect of time and temperature, CTMABODIPY = 5.0×10−7 mol l−1
,
;
2
CNO − = 1.0×10−7 mol l−1, CHCl = 0.36 mol l−1, CNaOH = 0.52 mol l−1
reaction time: 60 min.
2
(᭜) 20 ◦C, (᭿) 30 ◦C, (ꢀ) 40 ◦C, (×) 50 ◦C.
relative fluorescence intensity and the concentration of ni-
trite solution. A 5×10−7 mol l−1 of TMABODIPY solution
was chosen as optimal for at which the highest sensitivity
was obtained.
The highest determination sensitivity was obtained at NaOH
concentration of 0.52 mol l−1. Therefore, the final concen-
tration of NaOH concentration was chosen as 0.52 mol l−1
.
3.3. Effect of HCl concentration
3.5. Effect of time and temperature
As diazonium salt must be prepared in acidic condition,
proper HCl concentration should be selected to ensure TMA-
BODIPY react with nitrite as completely as possible. The
effect of HCl concentration was studied in the range of
0.25–0.54 mol l−1. It was found that the highest relative
fluorescence intensity of TMABODIPY was obtained at
HCl concentration of 0.36 mol l−1 with a tolerance range of
0.27–0.45 mol l−1, as can be seen in Fig. 5.
Time and temperature are both critical factors in reaction
of TMABODIPY and nitrite. According to the literature,
the TMABODIPY reacts with nitrite first in acidic solution
to form diazonium salt then in alkaline, the diazonium salt
transforms to diazotate soon. The effect of time and temper-
ature during the first reaction period was investigated. The
effect of time and temperature was indicated in Fig. 7. It was
found that it needed a longer time to reach the maximal rel-
ative fluorescence intensity at low temperature (20–30 ◦C)
than at high temperature (40 ◦C). Furthermore, high tem-
perature (40 ◦C) resulted into more relative fluorescence in-
tensity than low temperature. But as temperature climbed
up to 50 ◦C, the relative fluorescence intensity began to de-
cline and was hard to keep stable because diazonium salt is
apt to decomposed at too high temperature, while the reac-
tion temperature of 40 ◦C offered the RF stable at half an
hour. Therefore, the reaction was first performed at 40 ◦C
for 30 min in acidic solution, then at room temperature for
only 5 min in alkaline solution. After formation, the product
kept stable for at least 24 h.
3.4. Effect of NaOH concentration
Proper alkaline solution is needed to change diazonium
salt to diazotate, a stable and extremely fluorescence prod-
uct. The effect of NaOH concentration is shown in Fig. 6.
350
300
250
200
150
100
50
3.6. Linearity, sensitivity and precision
Under the optimal conditions described above, the lin-
ear calibration curve, detection limit and precision were ob-
tained. The linear calibration curve was as follows: Y =
1914.9X − 3.904 (γ = 0.9987, n = 10; X: nitrite concen-
tration (␮mol l−1); Y: RF between the blank and the stan-
dard sample). Linear range was from 0.08 × 10−7 to 3.00 ×
10−7 mol l−1 and the detect limit was 0.65 × 10−9 mol l−1
with a signal to noise ratio of 3. The repeatability of the
method was tested by the analysis of a standard nitrite solu-
0.2
0.4
0.6
0.8
1
CNaOH/mol l-1
Fig. 6. Effect of NaOH amount, CTMABODIPY = 5.0 × 10−7 mol l−1
,
CNO − = 1.0 × 10−7 mol l−1, CHCl = 0.36 mol l−1; T = 30 ◦C; reaction
2
time: 60 min.

992
M. Li et al. / Spectrochimica Acta Part A 60 (2004) 987–993
Table 2
Interference of foreign ions
Foreign ions
Tolerance (␮g l−1
)
Foreign ions
Tolerance (␮g l−1
)
Foreign ions
Tolerance (␮g l−1
)
2−
Ca2+
Mg2+
Al3+
Fe3+
Fe2+
Hg2+
Zn2+
100000
30000
20000
300
150
300
40000
Pb2+
Mn2+
Ni2+
Cu2+
1500
150
300
1000
4000
2000
10000
CO3 −
20000
50000
30000
20000
2000
0.2
NO3
EDTA
Citric
Br−
3−
PO4
2−
2−
C2O4
I−
I−a
SO4
2500
a
Masked by Fe3+
.
Table 3
Analytical result of nitrite in samples with TMABODIPY
Sample
Added
Found
R.S.D. (%) (n = 6)
Recovery (%)
Drinking water (␮mol l−1
)
0.0000
0.0500
0.1000
0.0251
0.0726
0.1239
1.65
1.68
2.34
95.0
98.8
Vegetable (cole)∗ (ng mg−1
)
0.0000
11.5000
23.0000
9.0390
20.4000
33.2810
1.15
3.93
2.34
98.8
105.0
tion over a period of 10 days (n = 6). The relative standard
Table 4
deviation was found to be 4.53% at 1.00 × 10−7 mol l−1
.
Comparison of detection limits for spectrofluorimetric determinations for
nitrite with different reagents
3.7. Interference of foreign ions
Reagents
Detection limit
References
(ng ml−1
)
The interference of a number of foreign ions was investi-
gated according to the recommended procedure at a nitrite
concentration of 1.00×10−7 mol l−1. The tolerance limit of
an ion was taken as the maximum amount causing an error
of ≤5% in the fluorescence of sample. As Table 2 indicated,
most of the ions studied did not interfere with the deter-
mination, even when present in large excess. Only I− had
2,3-Diaminonaphthalene
2,6-Diaminopyridine
Resorcinol
4-Hydroxycoumarin
5-Aminofluorescein
Tryptophan
5,6-Diamiono-1,3-naphthalene
disulfonic acid
TMABODIPY
<0.3
2
33
3
0.5
1
0.09
[17]
[19]
[14]
[15]
[12]
[18]
[16]
obvious interference, but could be well masked by Fe3+
.
0.03
This paper
3.8. Sample analysis
Acknowledgements
The method has been applied to the determination of ni-
trite in real samples such as drinking water and fresh veg-
etable (cole). Drinking water was tested without further
treatment. Fresh vegetable (cole) was prepared according to
the procedure described above. Both samples were analysis
in the recommended procedure. The results were shown in
Table 3.
The research presented in this manuscript was supported
by the Research Foundation for the Doctoral Program of
Higher Education and the National Natural Science Foun-
dation of China (No. 39970206).
References
[1] S. Moncada, R.M.J. Palmer, E.A. Higgs, Pharmacol. Revision 43
(1991) 109.
4. Conclusion
[2] D. Tsikas, F.M. Gutzki, S. Rossa, H. Bauer, C. Neumann, K. Dock-
endorff, J. Sandmann, J.C. Frolich, Anal. Biochem. 244 (1997)
208.
[3] C.S. Bruning-Fann, J.B. Kaneene, Vet. Human Toxicol. 35 (1993)
521.
[4] M.A. Kuiper, J.J. Visser, P.I. Bergmans, P. Scheltens, E.C. Wolters,
J. Neurol. Sci. 121 (1994) 46.
[5] P. Griess, Chem. Ber. 12 (1879) 426.
A new fluorescence probe, TMABODIPY, which pro-
vides high quantum efficiency, modest extinction coeffi-
cients, rigid and tolerate wide pH range, was developed to
determine trace nitrite in real samples. The method offered
the advantages of specificity, sensitivity and simplicity. The
sensitivity of the method is more highly than other spec-
trofluoremetric methods (see Table 4).
[6] R.A. Nicholas, J.B. Fox, J. Assoc. Off. Anal. Chem. 56 (1973) 922.

M. Li et al. / Spectrochimica Acta Part A 60 (2004) 987–993
993
[7] K. Yoshizumi, K. Aoki, Anal. Chem. 57 (1985) 737.
[8] R. Perez-Olomos, I. Yoldi, M.P. Ruiz, J.M. Merino, Anal. Sci. 14
(1998) 1001.
[9] Z. Iskandarani, D.J. Pietrzyk, Anal. Chem. 54 (1982) 2601.
[10] H.J. Cortes, J. Chromatogr. 234 (1982) 517.
[11] I.A. Pettas, S.I. Lafis, M.I. Karayannis, Anal. Chim. Acta. 376 (1998)
331.
[12] H.D. Axelrod, N.A. Engel, Anal. Chem. 47 (1975) 922.
[13] A. Lapat, L. Szekelyhidi, I. Hornyak, Biomed. Chomatogr. B 7
(1997) 120.
[15] T. Ohta, Y. Arai, S. Takitani, Anal. Chem. 58 (1986) 3132.
[16] N. Jie, J. Yang, F. Meng, Talanta 40 (1993) 1009.
[17] H. Wang, W. Yang, S.C. Liang, H.S. Zhang, Anal. Chim. Acta. 419
(2000) 169.
[18] P. Damiani, G. Burini, Talanta 33 (1986) 649.
[19] N. Jie, J. Yang, F. Meng, Talanta 40 (1993) 1009.
[20] R.W. Wager, J.S. Lindsey, Pure Appl. Chem. 68 (1996) 1373.
[21] M.P. Debeczeny, W.A. Svec, M.R. Wasielewski, New J. Chem. 20
(1996) 815.
[22] H.M. Alsoph, H.K. Robert, J. Am. Chem. Soc. 42 (1941) 1831.
[23] A.-M. Zhang, Chin. J. Anal Chem. 29 (2001) 202.
[14] M. Nakamura, Anal. Lett. 13 (1980) 771.