Proton “off-on” behaviour of methylpiperazinyl derivative of naphthalimide: a pH sensor based on fluorescence enhancement

Abstract

In a search for new type pH sensing fluorophores, the possibility of using the proton “off–on” switch behaviour of naphthalimide derivatives for optical pH sensor preparation has been explored. A new compound, N-allyl-4-(4′-methyl-piperazinyl)-1,8-naphthalimide (AMPN), was synthesized. The enhancement of fluorescence of AMPN with the increase of hydrogen ion concentration is based on arresting photo-induced electron transfer to the naphthalimide fluorophore from aliphatic amine group after its protonation. The Stokes Shift of the proposed type of pH sensing fluorophore is significantly larger than that of the fluorescein counterparts. To avoid the leakage of the fluorophore, AMPN was photo-copolymerized with 2-hydroxyethyl methacrylate and acrylamide on the glass surface. The fluorescence intensity of membrane contacted with a pH 3.50 buffer is 4.7 times of that for pH 12.00 buffer solutions. The proposed pH sensor is not susceptible to ionic strength and shows good selectivity, repeatability and short response time. The membrane shows a good stability with a lifetime over two months. The sensor can be used for the determination of pH in the range of pH 4.5–9.0, without interference of most commonly co-existing inorganic ions and some organic species. The sensor has been applied to the analysis of urine samples.

---

1 Introduction

Over the last decade, optical pH sensors based on fluorescence phenomena have been the subject of a number of investigations. These sensors are based on the measurement of fluorescence intensity,^1–4 fluorescence intensity ratios at two emission wavelengths,^5–10 and fluorescence lifetime.^11,12 The most frequently used fluorophores are derivatives of fluorescein, one of the classical fluorophores used in analytical chemistry. McNamara et al.,¹³ for instance, used micrometric phospholipid labeled with fluorescein-coated polystyrene particles for the intracellular pH measurement. Other fluorescein derivatives used include fluorescein,¹ aminofluorescein,³ fluorescein isothiocyanate,^4,8 SNARF-1,⁶ acryloylfluorescein,⁹ Oregon green carboxylic acid 514, DM-NERF, and Cl-NERF,¹¹ fluorescein isothiocyanate-dextran,¹⁴N-(fluorescein-5-thiocarbamoyl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine and 5-octadecanoylaminofluorescein.² The photostability of fluorescein derivatives might cause problems especially when excited by laser sources, and photo-bleaching might occur which shortens the lifetime of sensors. The Stokes Shift of fluorescein derivatives is rather small causing a relatively high background. The rhodamine derivative which is very similar to fluorescein derivatives such as seminaphthorhodamine-1 carboxylate was used by Grant and Glass.¹⁰ Metal complexes such as ruthenium complexes^12,15 and europium complex¹⁶ were also used as fluorophores for pH sensing, though the variation of pH might cause complex disassociation, and sensors based on metal complexes might be susceptible to interference from other metal ions and complex agents. New reports on fluorescence pH sensors are continually being published. It seems, however, the scope of useful fluorophores for pH sensing remains rather limited. Searching new type fluorophores to cover a wider spectrum of organic compounds and tailored to specific analytical needs seems to be a research task which is equally important as the modification of sensor design using old fluorophores. In the search and design of new pH sensing fluorophores one should take into consideration covalent immobilization of the fluorescing compounds onto the sensor membrane phase to avoid the leaching problem and guarantee sufficient life-time of the sensors. Searching for new fluorophores with good property including excellent photostability, large Stokes Shifts and high quantum yield for optical pH sensing is a challenge for the analytical chemistry research efforts.

In the search of new fluorophores for pH sensing with improved analytical characteristics, our attention has been focused on the naphthalimide derivatives. Naphthalimides are excellent fluorophores with high stability and quantum yield which have been used as fluorescent brightening agents,^17–19 intracellular biomarkers,²⁰ DNA photocleavers,²¹ and electro-luminescent copolymers.²² The photo-induced electron transfer (PET) process of naphthalimide fluorophores were studied in detail.^23–25 A systematic study of naphthalimides used as fluorescence sensor carriers has been undertaken in authors' laboratory.^26,27 Gan et al.²⁸ synthesized piperazinyl derivatives of naphthalimide and studied the PET process of the compound. In the piperazinyl group, the nitrogen atom not directly connected with the naphthalimide fluorophore forms an aliphatic amino group. The 1,8-naphthalimide structure is a fluorescing one when an aliphatic amino group is connected with carbon chains. The fluorescence of the molecule is quenched by the PET process. Protonation of the amino group arrests the PET process and causes restoration of the fluorescence originated from 1,8-naphthalimide. Substituting the hydrogen atom of the amino group with a short chain alkyl group would turn the secondary amine into a tertiary one and increase the basicity of the fluorophore on the membrane phase of pH sensor. This might substantially alter the “off–on” performance characteristics of the pH sensing. In order to synthesize the desired piperazinyl derivative of naphthalimide, a bromide derivative of naphthalimide containing an allyl group is synthesized. The bromide atom facilitates the introduction of piperazinyl or alkylpiperazinyl group necessary for realizing the proton “off–on” effect in pH sensing, while the allyl group provides a carbon chain with a terminal double bond capable to copolymerize with monomers on the modified glass surface. N-allyl-4-(4′-methyl-piperazinyl)-1,8-napthalimide (AMPN) has been synthesized by the present authors from 4-bromo-1,8-naphthalic anhydride via an imination reaction with allylamine, nuclearphilic substitution reaction with piperazine, and alkylation of amine. The AMPN shows “off–on” fluorescence behaviour with hydrogen ions and can be utilized for preparing fluorescence pH sensor. Under UV radiation AMPN was photo-copolymerized with 2-hydroxyethyl methacrylate, acrylamide and covalently immobilized on the glass surface treated with a silanizing agent. This paper reports the experimental results of investigation of analytical performance of the prepared sensor which utilized the fluorescence enhancement by protonation as the basis of fluorometric monitoring of solution pH.

2 Experimental

2.1 Materials and apparatus

4-Bromo-1,8-naphthalic anhydride, purchased from Taizhou Chemicals (Zhejiang, China), was recrystallized twice from chlorobenzene (mp 218–220 °C). 3-(Trimethoxysilyl)propyl methacrylate (TSPM) was a product of ACROS Organics. Piperazine was from Jujing Chemicals (Xinxiang, China) and used as received (purity 99.0%). Allylamine was synthesized from allyl alcohol, hydrobromic acid and sodium thiocyanate by a three step reaction²⁹ (bp 57–59 °C, MS: base peak 56, molecular ion peak 57). 2-Hydroxyethyl methacrylate (HEMA), acrylamide and 1, 2-cyclohexanediol diacrylate of reagent grade were used without further purification for the membrane matrix preparation. Britton–Robinson (B-R) buffer solutions of different pH were prepared by mixing appropriate amounts phosphoric, acetic and boric acids of the same concentration (0.04 mol l⁻¹) and adjusting to desired pH with 0.20 mol l⁻¹ sodium hydroxide.³⁰ Other chemicals were of analytical-reagent grade. Doubly distilled water was used throughout.

All fluorescence measurements were conducted on a Hitachi F4500 fluorescence spectrometer (Japan). The pH values of the solutions were measured with a PHS-3C pH meter and E-201-C pH glass electrode (Rex Instruments, Shanghai, China).

2.2 The synthesis of N-allyl-4-(4′-methyl-piperazinyl)-1,8-naphthalimide (AMPN)

The synthesis of N-allyl-4-bromo-1,8-naphthalimide (ABN). ABN was synthesized by a modified method similar to that reported by Hu et al.²² for the synthesis of N-allyl-4-nitro-1, 8-naphthalimide through the reaction of imination. 4-Bromo-1, 8-naphthalic anhydride (2.20 g) was refluxed with allylamine (0.65 ml) for 2 h in 50 ml of ethanol. After cooling and filtering, the solid was washed with ethanol and dried at 105 °C. A pale yellow product (2.20 g) of ABN was obtained with a nominal yield of 87.6%; mp 142–144 °C; MS: base peak 300, molecular ion peak 317 and 315.

The synthesis of N-allyl-4-piperazinyl-1,8-naphthalimide (APN). APN and AMPN were synthesized by a modified method similar to that reported by Gan et al.²⁸ The compound ABN obtained was used as the starting compound for the synthesis of APN. Compound ABN (0.0063 mol) was mixed with an excess of anhydrous piperazine (0.019 mol) in 30 ml of ethylene glycol monomethyl ether and refluxed for 3 h. The cooled reaction mixture was poured into 200 ml of water and filtrated. The solid was washed with distilled water, dried at 120 °C. A yellow solid of APN (1.93 g) with a nominal yield of 95.0% was obtained; mp 168–170 °C; MS: [M + 1]⁺ 322.

The synthesis of AMPN. Compound APN (1.50 g) was mixed in 15 ml of formic acid (ω = 0.88) with 0.28 g of paraformaldehyde and stirred for 20 h at 80 °C. After removing the solvent under reduced pressure, 20 ml of 3 mol l⁻¹ hydrochloric acid was added and refluxed for 1 h. To the cooled mixture, 10 g of sodium carbonate was added. The yellow solid was washed with distilled water, dried at 120 °C and extracted with ethanol to give 1.41 g of yellow solid of AMPN with a nominal yield of 90.1%; mp 150–152 °C; MS: [M + 1]⁺ 336.

2.3 Preparation of the glass surface

For covalently immobilizing AMPN on the sensor surface, a terminal double bond was introduced onto the glass surface modified by silanization as described in the literature^31,32 with some modifications. Glass plates (diameter of 13 mm) were immersed successively in 3% HF and 10% H₂O₂ for 10 min each and then washed with double distilled water. A solution of TSPM was prepared by mixing 0.2 ml of TSPM, 2 ml of 0.2 mol l⁻¹ HOAc–NaOAc buffer solution of pH 3.6 and 8 ml of water and then stirring for 5 min. The glass plates were soaked in this solution for 2 h, then rinsed with distilled water and dried at room temperature.

2.4 Preparation of the pH sensing membrane

The AMPN polymer membrane was prepared according to the following procedure. Acrylamide (400 mg) was dissolved in 0.50 ml of N,N-dimethylformamide, with subsequent addition of HEMA (1.0 ml), 1,2-cyclohexanediol diacrylate (0.10 ml), AMPN (10 mg), benzoin ethyl ether (45 mg), and benzophenone (60 mg). The mixture was cast onto a cleaned poly(tetrafluoroethylene) plate. Silanized glass plates were placed over the droplets, and UV radiation (254 nm, 30 W, 10 cm over the glass plates) was directed onto the glass plates for ∼2 h. After the UV irradiation period, the glass plates were washed with water and methanol to remove any unreacted species until no leaching of the AMPN was detected. The thickness of films is about 40 µm.

2.5 Fluorescence measurements

The fluorescence was measured with a Hitachi F-4500 spectrofluorimeter controlled by a personal computer data processing unit. The light source was a 150 W Xe lamp and the detector was a R928F red-sensitive photomultiplier tube. The excitation and emission slits were both set at 5 nm. A home-made poly(tetrafluoroethylene) flow-cell described elsewhere²⁶ and a bifurcated optical fiber (30 + 30 quartz fibers, diameter of 8 mm and length of 1.2 m) were used for the pH sensing measurements. The excitation light was guided to the cell through one arm of the bifurcated optical fiber and the emission light collected through the other. A glass plate with covalently immobilized sensing membrane was fixed on the top of the flow chamber by the mounting screw nut with the membrane facing and contacting with the sample solution. The sample solution was driven through the flow cell by a peristaltic pump (Guokang Instruments, Zhejiang, China) at a flow rate of 3.0 ml min⁻¹. The sensing membrane was allowed to equilibrate with the sample solution until the fluorescence signal was stabilized. After each measurement, the fluorescence intensity of the membrane was recovered by pumping the pH 12.0 buffer solution through the cell prior to the next measurement.

2.6 Analysis of urine samples

The volunteers' urine samples were collected at different time periods. The samples were directly pumped through the flow cell and the pH was calculated by using calibration curve method.

3 Results and discussion

3.1 Spectral characteristics

In a systematic search of new pH sensing fluorophores of proton “off–on” action type, a series of naphthalimide derivatives were synthesized in this laboratory. The structures of some representative ones are shown in Fig. 1. One notices that naphthalimides without substitution group at position 4 do not possess strong fluorescence. When at position 4 there exist a monosubstituted aminogroup leaving a hydroxyethyl or methylacryloxyethyl terminus, the naphthalimide obtained, (N-allyl-4-(N-2′-hydroxyethyl)amino-1,8-naphthalimide (AHEAN) or N-propyl-4-(N-methylacryloxyethyl)amino-1,8-naphthalimide (PMEAN) respectively (Fig. 1), shows strong fluorescence with Stokes Shift of about 80 nm which is larger than that of fluorescein (ca. 50 nm). The fluorescence of these compounds does not change, however, with pH and remains constant in a wide pH range (3–11). By introducing the piperizine structure at position 4, the proton “off–on” behaviour could be realized.

Fig. 1 Structure of naphthalimide derivatives.

de Silva and Rice²⁵ and Gan et al.²⁸ reported the electron transfer originated from dialkylamino unit attached to the 4-amino position of naphthalimide derivatives. In the case of AMPN, we can assume that PET occurs between the naphthalimide fluorophore and electron-donating NCH₃ in the piperizine structure. The fluorescence of the naphthalimide fluorophore is being quenched by the PET process. The PET process can be stopped by the protonation of the NCH₃ with the fluorescence of the fluorophore restored. An “off–on” switch behaviour is observed. The peaks of excitation and emission spectra are located at 400 nm and 517 nm, respectively (Fig. 2). A substantially increased Stokes Shift of about 117 nm was observed which is favourable for fluorescent pH sensing.

Fig. 2 Fluorescence excitation and emission spectra of three naphthalimide derivatives: (a) AMPN pH 3.50; (a′) AMPN pH 12.00; (b) PMEAN pH 3–11 (measured at pH 8.36); (c) AHEAN pH 3–11 (measured at pH 7.46).

The fluorescence spectra of the AMPN copolymer membrane in B-R buffer solutions of different pH are showed in Fig. 3. The excitation spectrum was obtained by fixing the emission wavelength at 517 nm, the peak of the emission spectrum. The emission spectrum was obtained by fixing the excitation wavelength at 400 nm. Hydrogen ion can strongly enhance the fluorescence intensity of AMPN copolymer and the fluorescence intensity is 4.7 times of the original values. B-R buffer solutions of various pH values were driven through the flow cell by a peristaltic pump at a flow rate of 3.0 ml min⁻¹. The excitation spectrum shows maximum intensity peak of 434 nm in pH 12.00 B-R buffer solutions and 400 nm in pH 3.50 buffer solution, while the peak of the emission spectrum is located at 517 nm. With the decreasing of pH from 12.00 to 3.50, the peak of 432 nm in the excitation spectrum shows a blue shift toward 400 nm. Upon the decreasing of pH from 12.00 to 3.50 there is no shift of the emission spectrum peak of 517 nm.

Fig. 3 Fluorescence excitation and emission spectra of the sensing membrane contacted with different pH B-R buffer solutions (1) 12.00; (2) 11.00; (3) 10.00; (4) 9.50; (5) 9.00; (6) 8.50; (7) 8.00; (8) 7.50, (9) 7.00, (10) 6.50; (11) 6.00; (12) 5.50; (13) 5.00; (14) 4.50; (15) 4.00; (16) 3.50.

3.2 Effect of ionic strength

In the optical pH sensing methods, ionic strength often affects the pH determinations. The effect of ionic strength on the fluorescence intensity of membrane was examined with pH 6.50 B-R buffer solutions. The buffer solutions were adjusted to ionic strength from 0.01 to 2 mol l⁻¹ with NaCl for sensor based on AMPN. It was found that with increasing ionic strength, no significant changes of the fluorescence intensity of membrane were observed for AMPN. This indicates that the proposed sensor based on AMPN is not susceptible to ionic strength variations. The AMPN is good in the fabrication of optical pH sensor for situations when maintaining a constant ionic strength is a problem.

3.3 Repeatability, reversibility and response time

The repeatability and reversibility of the pH induced fluorescence signal change were evaluated by recording the fluorescence readings during alternatively pumping buffer solutions of pH 8.50, 7.50, and 6.00 through the sensor system. The buffer solution of pH 12.00 was used as the reference blank. Four successive cycles were carried out. Fig. 4 shows the fluorescence intensity change upon switching from one solution to another. The confidence interval of mean fluorescence intensity values for four replicates were found to be 157.3 ± 0.8, 285.2 ± 1.0 and 341.1 ± 1.6 (P = 0.95) for buffer solutions of pH 8.50, 7.50 and 6.00, respectively. A confidence interval of mean fluorescence intensity value of 80.1 ± 0.2 (n = 13, P = 0.95) was obtained for buffer solutions of pH 12.00. The repeatability and reversibility of different films were very similar and satisfactory. These results were obtained, of course, by using a standard spectrofluorimeter with stabilized light source and detecting system. As the purpose of the present study is to investigate the possibility of constructing a robust convenient device for pH measurement, the use of fluorescence lifetime rather than intensity measurement might be a better choice. This requires some further experimental investigations. The recovering time is independent of the order of concentration change i.e. it actually remains the same no matter whether one switches from low to high concentration level or the reverse. The response time is of the order of 90 s. The protonation of the aliphatic amine group involved seems quite reversible and no noticeable hysteresis effect was observed.

Fig. 4 Fluorescence responses vs. time at 517 nm by alternatively pumping different pH B-R buffer solutions (a) 12.00; (b) 8.50; (c) 7.50; (d) 6.00. The fluorescence intensities are in arbitrary units given by the time scanning mode of an Hitachi F-4500.

3.4 Stability and lifetime

For investigating the stability of the proposed sensor, a buffer solution of pH 6.50 was continuously pumped through the flow cell in contact with the sensing membrane over a period of 6 h and the excitation light was illuminated on the membrane continuously. The membrane fluorescence intensity was recorded with an interval of 30 min. A relative standard deviation of 0.83% was obtained and no photo-decomposition of AMPN was observed. The naphthamilide derivatives show some advantages in this respect comparing to the fluorescein counterparts. The membrane shows good stability and has a lifetime of at least two months. The covalent immobilization of the active component in the copolymer can hinder the leaching effect and improve the lifetime of sensor remarkably.

3.5 Selectivity

The effect of interferents on the fluorescence determination of pH was investigated with the solution pH fixed at 6.48 by 0.2 mol l⁻¹ HOAc–NaOAc. For common inorganic ions and some possible co-existing organics less than 5% relative error was obtained for concentrations less than the 1 × 10⁻³ or 1 × 10⁻⁴ mol l⁻¹ level tested (Table 1). The organic cations, for example quaternary ammonium and berberine, do not interfere with the pH sensing. The sensor has a good selectivity for hydrogen ion, making it feasible for practical applications in the monitoring of pH.

Table 1 Effect of different interferents on the fluorescence intensity of the optical membrane (pH 6.50). ΔI = I − I₀, I₀ and I are the fluorescence intensity of the optical membrane contacted with the pH 6.50 HOAc–NaOAc buffer solutions without and with interferents, respectively

Interferent	Concentration (mol l^−1)	Relative fluorescence change value(%) ΔI/I0 × 100
CaCl2	1.0 × 10^−3	−2.0
Cd(NO3)2	1.0 × 10^−4	−1.3
CrCl3	1.0 × 10^−4	−0.8
CuSO4	1.0 × 10^−4	−1.1
Fe2(SO4)3	1.0 × 10^−4	−0.5
HgCl2	1.0 × 10^−4	3.2
KCl	1.0 × 10^−3	0.8
MgCl2	1.0 × 10^−3	−1.8
NaNO2	1.0 × 10^−3	−1.3
NaNO3	1.0 × 10^−3	−1.0
NH4Cl	1.0 × 10^−3	1.0
NiSO4	1.0 × 10^−3	−1.2
Pb(NO3)2	1.0 × 10^−4	1.3
ZnCl2	1.0 × 10^−4	−2.9
Berberine hydrochloride	1.0 × 10^−4	1.3
Phenol	1.0 × 10^−3	−0.4
Tetraethylammonium bromide	1.0 × 10^−3	0.5
Tetramethylammonium bromide	1.0 × 10^−3	0.4
Urea	1.0 × 10^−3	−1.1

---

3.6 Measuring range

To eliminate the influence of possibly varied experimental conditions and other factors, we can express the change of fluorescence intensity with α which is defined as follows:

Here F₀ is the fluorescence intensity when the membrane contacted with the pH 12.00 reference buffer solutions, F_∞ is the fluorescence intensity when AMPN in the membrane is completely protonated, which corresponds to the fluorescence intensity recorded at pH 3.5 or lower, F is the fluorescence intensity of the membrane exposed to the different pH B-R buffer solutions.

The calibration curve of AMPN sensor was constructed by recording the fluorescence intensity values of buffer solutions of different pH and plotting a versus pH. The α is related to pH as a linear function in the pH ranges of 4.5–7.0 and 7.0–9.0:

α = 1.3058 − 0.07275pH (R = 0.9986, pH 4.5–7.0) (2)

α = 3.1953 − 0.3400pH (R = 0.9974, pH 7.0–9.0) (3)

These linear equations can served as the quantitative basis for the determination of pH in a range of pH 4.5 to 9.0.

The pH range of the proposed sensor is limited by the pK_a of the amino group of the fluorescence carrier. Some preliminary experiments show that if the tertiary amino group is replaced by secondary amino group, the pH range could be slightly extended.

3.7 Application to real sample analysis

The practical application of the proposed sensor has been tested by pH determinations of urine samples. The urine samples provided by volunteers were directly pumped into the flow cell and the pH was determined using the proposed sensor by calibration curve method. The results were compared with those obtained using a conventional glass electrode. Table 2 collects the results obtained by the two methods. Considering the fact that the proposed pH sensor is of a radically different type compared to a glass pH electrode, the results seem to be reasonably acceptable for practical use.

Table 2 Determination of the pH of three urine samples using the proposed sensor based on AMPN

Samples	pH^a^c	pH^b^c	Error
a Measured by present sensor. b Measured by conventional glass electrode. c Mean ± S.D. of triplicate measurements.
1	6.76 ± 0.03	6.63 ± 0.03	0.13
2	6.70 ± 0.05	6.86 ± 0.02	−0.16
3	6.33 ± 0.04	6.25 ± 0.02	0.08

---

4 Conclusions

Naphthalimide derivatives have been shown to be promising fluorophores possessing some advantages over the so far reported ones such as fluorescein derivatives. The outstanding features of the methylpiperazinyl derivative of naphthalimide include, among others, significantly larger Stokes Shift and excellent photostability. The new compound N-allyl-4-piperazinyl-1,8-naphthalimide (AMPN) synthesized has been examined experimentally as a feasible pH sensing fluorophore for sensor preparation. The sensor with covalently immobilized AMPN shows fluorescence enhancement with the increase of hydrogen ion concentration, which is the consequence of the hindering of PET from aliphatic amine group in AMPN by protonation. The proposed pH sensor can be used in the pH range 4.5 to 9.0. The sensor is not susceptible to ionic strength variations and shows good selectivity, repeatability and satisfactory response. The lifetime of the sensing membrane is guaranteed by covalent immobilization. The feasibility of the practical use of the sensor has been demonstrated by analysis of urine samples.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grants 20075006, 29975006 and 29735150).

References

1. M. Cajlakovic, A. Lobnik and T. Werner, Anal. Chim. Acta, 2002, 455, 207–213.
2. D. R. Fry and D. R. Bobbitt, Microchem. J., 2001, 69, 123–131.
3. A. Lobnik, I. Oehme, I. Murkovic and O. S. Wolfbeis, Anal. Chim. Acta, 1998, 367, 159–165.
4. D. A. Nivens, Y. Zhang and S. M. Angel, Anal. Chim. Acta, 1998, 376, 235–245.
5. D. A. Nivens, M. V. Schiza and S. M. Angel, Talanta, 2002, 58, 543–550.
6. S. Brasselet and W. E. Moerner, Single Mol., 2000, 1, 17–23.
7. J. Ji and Z. Rosenzweig, Anal. Chim. Acta, 1999, 397, 93–102.
8. K. L. Michael, L. G. Taylor and D. R. Walt, Anal. Chem., 1999, 71, 2766–2773.
9. J. A. Ferguson, B. G. Healey, K. S. Bronk, S. M. Barnard and D. R. Walt, Anal. Chim. Acta, 1997, 340, 123–131.
10. S. A. Grant and R. S. Glass, Sens. Actuators, B, 1997, 45, 35–42.
11. H. J. Lin, H. Szmacinski and J. R. Lakowicz, Anal. Biochem., 1999, 269, 162–167.
12. U. Kosch, I. Klimant and O. S. Wolfbeis, Fresenius’ J. Anal. Chem., 1999, 364, 48–53.
13. K. P. McNamara, T. Nguyen, G. Dumitrascu, J. Ji, N. Rosenzweig and Z. Rosenzweig, Anal. Chem., 2001, 73, 3240–3246.
14. Y. H. Liu, T. H. Dam and P. Pantano, Anal. Chim. Acta, 2000, 419, 215–225.
15. C. Malins, H. G. Glever, T. E. Keyes, J. G. Vos, W. J. Dressick and B. D. MacCraith, Sens. Actuators, B, 2000, 67, 89–95.
16. A. Lobnik, N. Majcen, K. Niederreiter and G. Uray, Sens. Actuators, B, 2001, 74, 200–206.
17. T. Konstantinova, A. Spirieva and T. Petkova, Dyes Pigm., 2000, 45, 125–129.
18. L. G. F. Patrick and A. Whiting, Dyes Pigm., 2002, 52, 137–143.
19. M. Q. Chen, H. B. Yan, Y. Y. Hu and Y. J. Shen, J. East China Univ. Sci. Technol., 2001, 27, 195–197.
20. W. W. Stewart, J. Am. Chem. Soc., 1981, 103, 7615–7620.
21. W. Yao, X. H. Qian and Q. Y. Hu, Tetrahedron Lett., 2000, 41, 7711–7715.
22. C. Hu, W. H. Zhu and H. Tian, Acta Polym. Sin. (Chinese), 1999, 2, 232–235.
23. H. Tian, T. Xu, Y. B. Zhao and K. C. Chen, J. Chem. Soc., Perkin Trans. 2, 1999, 545–549.
24. T. P. Le, J. E. Rogers and L. A. Kelly, J. Phys. Chem., A, 2000, 104, 6778–6785.
25. A. P. de Silva and T. E. Rice, J. Chem. Soc., Chem. Commun., 1999, 163–164.
26. C. G. Niu, Z. Z. Li, X. B. Zhang, W. Q. Lin, G. L. Shen and R. Q. Yu, Anal. Bioanal. Chem., 2002, 372, 519–524.
27. C. G. Niu, X. Yang, W. Q. Lin, G. L. Shen and R. Q. Yu, Analyst, 2002, 127, 512–517.
28. J. A. Gan, K. C. Chen and H. Tian, J. East China Univ. Sci. Technol., 2001, 27, 217–220.
29. Handbook of Fine Chemicals (Organics Part), ed. S. G. Zhang, Chemical Industry Press, Beijing, 1996, pp. 101–105.
30. Lange's Handbook of Chemistry, ed. J. A. Dean, McGraw-Hill Book Company Press, New York, 13th edn., 1985, Sec. 5.102, Table 5-24.
31. C. Munkholm, D. R. Walt, F. P. Milanovich and S. M. Klainer, Anal. Chem., 1986, 58, 1427–1430.
32. M. P. Xavier, D. Garcia-Fresnadillo, M. C. Moreno-Bondi and G. Orellana, Anal. Chem., 1998, 70, 5184–5189.

---