The quantitative detection of HO˙ generated in a high temperature H₂O₂ bleaching system with a novel fluorescent probe benzenepentacarboxylic acid

Abstract

A novel fluorescent probe, benzenepentacarboxylic acid (BA), was developed for the quantitative detection of hydroxyl radicals (HO˙) in a simulated hydrogen peroxide (H₂O₂) bleaching system. Compared to terephthalic acid (TA), a commonly-used fluorescent probe of HO˙ with four possible reaction sites, the present probe has only one reaction site for HO˙ addition that results in the formation of a single fluorescent product, hydroxybenzenepentacarboxylic acid (HBA). The generation of the single pure fluorescent product by BA, rather than mixtures by TA, made the present probe more sensitive, accurate and reproducible in the quantitative detection of HO˙. Based on the working curve obtained from the fluorescence intensity and HBA concentration, a fluorimetric method for the quantitative detection of HO˙ was developed. The present probe BA was successfully used for the quantitative detection of HO˙ under alkaline and high temperature (80 °C) conditions, in a simulated H₂O₂ bleaching system. The method was highly sensitive, reproducible, and applicable in a wide temperature range of 20–98 °C. The present novel fluorescent probe will be potential tool for evaluating the generation and function kinetics of HO˙ generated in biological and environmental fields.

---

1. Introduction

Hydroxyl radicals (HO˙) play a crucial role in the physiological and pathological processes of biological molecules such as DNA, protein and fat.^1,2 Efforts on the detection of intracellular HO˙ has been made recently.³ In vitro, HO˙ has also gained great attention due to its strong oxidizing property and non-selectivity in rapid reactions with various organic and inorganic substances.^4,5 However, the high reactivity and short life of HO˙⁶ make it difficult to realize quantitative determination directly. Most of the approaches reported for HO˙ detection in vitro are focused on the qualitative description of standard Fenton (Fe²⁺/H₂O₂) and Fenton-like systems (pH ≤ 7).^7–9 Currently the developed methods for HO˙ detection mainly include electron paramagnetic resonance spectroscopy (ESR),¹⁰ chemiluminescence (CL),¹¹ high-performance liquid chromatography (HPLC)¹² and electrochemistry¹³ etc. However, these methods have non-ignorable defects, such as, complex operation, poor sensitivity and selectivity, and poor anti-interference ability.

Recently, the fluorimetric method has been considered to be promising for HO˙ determination,^14,15 because it is rapid, sensitive, selective, and easy to operate. Fluorescent dyes such as salicylfluorone¹⁶ and rhodamine 6G (ref. 17) have been used for HO˙ determination through the reduction of the fluorescence intensity, based on the strong oxidizing ability of HO˙. But these fluorescent dyes are prone to be degraded by any substance with a certain degree of oxidizing ability, causing poor selectivity and reproducibility in HO˙ detection, especially under complex conditions. Nitroxide-contained anthracene has also been utilized as a probe to detect HO˙ in the presence of dimethylsulfoxide (DMSO), where the reaction between DMSO and HO˙ can produce methyl radicals (˙CH₃) that could be further captured by the fluorescent probe to enhance the fluorescence intensity.¹⁸ This indirect method needs a long duration for the multistep reaction processes, resulting in poor sensitivity and reproducibility.

As a new class of trapping ligand, benzoic acid and terephthalic acid (TA) have been extensively employed for HO˙ detection via the formation of fluorescent hydroxylated products.^19,20 There exist 5 vacant sites available for HO˙ additions on the asymmetric benzoic acid molecule, but only ortho-hydroxylated derivatives emit fluorescence. Thus, the non-selectivity in the HO˙ addition leads to the generation of non-fluorescent meta- and para-hydroxylated derivatives as well as multiple-hydroxylated derivatives. This means that invalid HO˙ addition reactions are inevitable, leading to poor sensitivity and inaccuracy in HO˙ determination. Thus symmetrical TA has been developed to replace benzoic acid as a fluorescent probe for HO˙.²⁰ Regardless of any vacant site combined with HO˙, fluorescent ortho-hydroxylated products could be derived, making TA widely applicable in HO˙ determination in biological and environmental systems.^7,20–26 However, it is obvious that multiple-hydroxylated derivatives, including double, three, and four hydroxylated derivatives should be produced simultaneously in TA systems because of the non-selectivity in the HO˙ addition. As a result, the TA-based fluorescent probe for HO˙ detection will show a poor reproducibility that limits its practical application in quantitative detection of HO˙. Hence, it is urgent to develop a novel fluorescent probe for HO˙ with excellent specificity, reproducibility and stability to realize the quantitative determination of HO˙, particularly under practical complex conditions, such as a high temperature H₂O₂ bleaching system which is an important pretreatment process in the textile industry to remove impurities to facilitate the subsequent dyeing and finishing processes. Unlike standard Fenton (Fe²⁺/H₂O₂) and Fenton-like reaction systems (pH ≤ 7, room temperature), the high temperature H₂O₂ bleaching system (20–98 °C, pH ≥ 7) is representative of practical complex conditions which have rarely been used to generate HO˙. Therefore, the development of a fluorescent probe that is capable of detecting HO˙ at a high temperature, will be helpful for monitoring the HO˙ yield under practical complex conditions, but also for understanding the H₂O₂ bleaching mechanism.

To get a good reproducibility from the TA system, one concept is considered that only a single pure fluorescent product is generated after addition reaction with HO˙. Bearing this in mind, we envisaged the introduction of substitution on the benzene ring of TA, leaving one site feasible for HO˙ addition, which means that a single pure fluorescent product should be obtained. Herein benzenepentacarboxylic acid (BA), which has five carboxylic acid groups on the benzene ring and only one vacant site available for HO˙ addition, was firstly used as a novel fluorescent probe for HO˙ detection. Compared with the commonly-used TA system, a superior performance such as a better water-solubility, high sensitivity, and good reproducibility, was found in the BA system, especially under high temperature conditions. The present novel probe was practically applied to the quantitative evaluation of HO˙ generation in high temperature alkaline H₂O₂ bleaching systems.

2. Experimental

2.1. Apparatus

The fluorescence spectra and relative fluorescence intensity was measured on an F-7000 spectrophotofluorometer (Hitachi, Japan) with a xenon lamp excitation and a quartz cuvette (1.0 cm optical path) as the container. The excitation and emission band-pass of spectrometer slits were set at 5.0 nm, with the PMT voltage of 400 V.

The ESI-MS spectra were obtained on a Varian 310 mass spectrometer. The ¹H NMR and ¹³C NMR spectra were recorded in D₂O on a Bruker AVANCE III 400 MHz spectrometer with TMS as the standard. All of the H₂O₂ systems were heated by an RY-25012 room temperature oscillation dyeing machine (Long Ling electronic technology co., LTD, Shanghai, China).

2.2. Reagents

Benzenepentacarboxylic acid (BA, purity > 98%, TCI), terephthalic acid (TA, purity 99%), H₂O₂ (30 wt%), tetra-acetylethylenediamine (TAED) and dimethylsulfoxide (DMSO) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Sodium benzenepentacarboxylate and terephthalate stock solution (0.01 mol l⁻¹) were prepared from benzenepentacarboxylic acid and terephthalic acid through neutralization with NaOH, respectively. The working solution was prepared from a dilution of the corresponding stock solution. All the chemicals used were of analytical reagent grade. Double distilled water was used throughout.

2.3. Procedures for HO˙ determination

H₂O₂ (2.2 × 10⁻⁴ mol l⁻¹) was added into a reaction solution containing BA/TA (4 × 10⁻⁴ mol l⁻¹) and NaOH (4 × 10⁻³ mol l⁻¹). The reaction mixture was kept at a certain temperature and then promptly quenched by flowing cold water. A certain amount of reaction solution was placed in the four-pass optical quartz cuvette, then the fluorescence spectra and relative fluorescence intensity was measured.

3. Results and discussion

3.1. The excitation and emission spectra

The excitation and emission spectra of BA were recorded in an alkaline solution with and without H₂O₂ addition. As shown in Fig. 1, BA only was nonfluorescent, whereas a strong fluorescence emission was observed in the presence of H₂O₂ (λ_ex = 311 nm, λ_em = 435 nm), revealing that the HO˙ generated from the reaction system promoted the formation of a relative fluorescent product.

Fig. 1 The excitation and emission spectra of BA. (1) BA (4 × 10⁻⁴ mol l⁻¹); (2) BA (4 × 10⁻⁴ mol l⁻¹), H₂O₂ (2.2 × 10⁻⁴ mol l⁻¹). T = 80 °C, pH = 10.0; t = 60 min.

3.2. A comparison of the performances between the BA and TA systems

To exemplify the superior performance of BA, the results of 9 parallel determinations both in the Fenton (Fe²⁺/H₂O₂) reaction system at an ambient temperature and the alkaline H₂O₂ reaction system at a high temperature (80 °C), were compared by using BA and TA as the fluorescent probe to capture HO˙ respectively (Fig. 2). It can be seen that BA showed a comparable reproducibility with that of TA in the Fenton system (Fig. 2a), with a respective relative standard deviation (RSD) of 0.8% and 1.03%. However, in the high temperature alkaline H₂O₂ reaction system, while the reproducibility of TA became much worse (RSD, 7.59%), BA still exhibited an excellent reproducibility for HO˙ detection (RSD, 2.39%). The superiority of BA in the high temperature H₂O₂ reaction system may result from its one vacant site on the benzene ring allowing for HO˙ addition to generate a single fluorescent hydroxylated product HBA (Scheme 1). In contrast, various multiple-hydroxylated derivatives could be produced in the case of TA which cause a poor reproducibility, which is due to the fact that there is more than one site available for HO˙ combination (Scheme 1).

Fig. 2 A reproducibility comparison between BA and TA. (a) The Fe²⁺/H₂O₂ Fenton reaction system: Fe(II) (80 μmol l⁻¹), KH₂PO₄–K₂HPO₄ (0.02 mol l⁻¹, pH 7.4), ambient temperature (25 °C), t = 30 min; (b) the alkaline H₂O₂ reaction system: NaOH (4 × 10⁻³ mol l⁻¹), T = 80 °C, t = 60 min. H₂O₂ (2.2 × 10⁻⁴ mol l⁻¹), BA (4 × 10⁻⁴ mol l⁻¹), TA (4 × 10⁻⁴ mol l⁻¹).

Scheme 1 The processes of HO˙ addition to BA and TA.

To prove such an assumption, ESI-MS was used to analyse the reaction mixtures of BA and TA obtained in the high temperature system respectively (Fig. S1, ESI†). One main peak was found at m/z 470.7 for BA, which is ascribed to a single hydroxylated BA product (M + OH + 7Na)⁺, whereas various peaks were observed at m/z = 227.8, 242.0, 340.7 and 362.5 for TA, which are ascribed to (M + OH + 2Na)⁺, (M + 2OH + 2Na)⁺, (M + 4OH + 5Na)⁺ and (M + 4OH + 6Na)⁺. Clearly, a single hydroxylated fluorescent product was produced by BA as described in Scheme 1, whereas single and multiple-hydroxylated products were formed by TA which is different from the previous assumption that only one single hydroxylated product is formed.^4,19–26 The better performance allows BA to be a novel fluorescent probe for HO˙ determination, even in severe alkaline and high temperature conditions, in a simulated H₂O₂ bleaching system.

DMSO has been known to be an efficient scavenger for HO˙.²⁷ It was found that the addition of DMSO into the alkaline H₂O₂ reaction system efficiently quenched the fluorescence originated from hydroxylated BA (Fig. S2, ESI†), indicating the suppression effects of DMSO on the formation of fluorescent HBA. This further confirms that the fluorescence enhancement indeed results from the formation of HBA, a fluorescent product formed through the addition of HO˙ to BA molecules (Scheme 1).

3.3. Optimization of analytical parameters for HO˙ detection

To get optimal conditions for HO˙ detection in a high temperature H₂O₂ bleaching system, the effect of the analytical parameters including the temperature, pH, BA concentration and reaction duration on the HO˙ yield characterized by fluorescence intensity was discussed respectively in BA systems.

3.3.1. Effect of temperature. For the alkaline H₂O₂ bleaching system, the temperature is a significant factor that efficiently controls the decomposition rate of H₂O₂ and therefore the bleaching performance. Herein, HO˙ generation in an alkaline H₂O₂ reaction system at different temperatures (20 °C, 40 °C, 60 °C, 80 °C, 98 °C) was detected with BA. As displayed in Fig. 3, the HO˙ yield was enhanced with the increasing temperature. At lower temperatures, from 20 °C to 40 °C, only a very weak fluorescence was detected, indicating the slow decomposition of H₂O₂ and the generation of HO˙. In contrast, at higher temperatures (>60 °C), the fluorescence intensity at 435 nm increased significantly, indicating that a large amount of HO˙ was generated due to the fact that the activation energy of the decomposition reaction of H₂O₂ was efficiently overcome. In order to avoid the quenching of HO˙ that is probably caused by the rapid and excessive generation at too a high temperature (98 °C), 80 °C was selected for HO˙ detection in this work.

Fig. 3 The effect of temperature on HO˙ generation. T (20–98 °C); H₂O₂ (2.2 × 10⁻⁴ mol l⁻¹); BA (4 × 10⁻⁴ mol l⁻¹); pH = 10.0; t = 60 min.

3.3.2. The effect of pH. It is well known that H₂O₂ bleaching system is generally conducted under alkaline conditions due to the fact that the high pH condition facilitates the decomposition of H₂O₂. As shown in Fig. 4, the fluorescence intensity reached a maximum around pH 10–11, even if the probe BA can work for HO˙ detection at pH 7.0–13.0. Therefore, pH 10–11 was selected as the optimal condition for HO˙ generation in the H₂O₂ bleaching system.

Fig. 4 The effect of pH on HO˙ generation. pH (7.0–13.0); H₂O₂ (2.2 × 10⁻⁴ mol l⁻¹); BA (4 × 10⁻⁴ mol l⁻¹); T = 80 °C; t = 60 min.

3.3.3. The effect of BA concentration. The momentary presentation of HO˙ in an aqueous solution make it necessary to be captured immediately by the probe. The effect of the BA concentration on the HO˙ yield was tested in the range 1 × 10⁻⁵ to 8 × 10⁻⁴ mol l⁻¹ (Fig. 5). It can be seen that fluorescence intensity became enhanced evidently with the increasing BA concentration from 1 × 10⁻⁵ mol l⁻¹ to 4 × 10⁻⁴ mol l⁻¹ and then reached a steady state. An insufficient BA concentration led to the spontaneous quenching of partial HO˙ which has not been promptly captured. Hence, the BA concentration for HO˙ detection was chosen to be 4 × 10⁻⁴ mol l⁻¹.

Fig. 5 The effect of the BA concentration on HO˙ generation. BA (1 × 10⁻⁵ to 8 × 10⁻⁴ mol l⁻¹); H₂O₂ (2.2 × 10⁻⁴ mol l⁻¹); T = 80 °C; pH = 10.0; t = 60 min.

3.3.4. The effect of the reaction duration. The generation and capture of HO˙ in the H₂O₂ reaction system is actually a dynamic balance process. HO˙ releases from H₂O₂ continuously and adds to the single vacant site of BA, eventually forming the fluorescent product HBA. As displayed in Fig. 6, the fluorescence intensity became enhanced rapidly up to 60 min and remained steady after that, indicating that the HO˙ generated has been almost completely captured within 60 min. Therefore, 60 min was chosen as the reaction duration.

Fig. 6 The effect of the reaction time on HO˙ generation. Time (15–180 min); H₂O₂ (2.2 × 10⁻⁴ mol l⁻¹); BA (4 × 10⁻⁴ mol l⁻¹); T = 80 °C; pH = 10.0.

3.4. Quantitative determination of HO˙

3.4.1. The linear relationship between the fluorescence intensity and the H₂O₂ concentration. As shown in Fig. 7, a linear relationship was found between the fluorescence intensity at 435 nm and the H₂O₂ concentration in the range of 1.1 × 10⁻⁵ to 1.1 × 10⁻³ mol l⁻¹, with a detection limit of 1 × 10⁻⁶ mol l⁻¹. It is clear that the HO˙ yield is proportional to the fluorescence intensity and therefore to the H₂O₂ concentration. This indicates that only a small amount of H₂O₂ decomposed into HO˙ in the alkaline H₂O₂ high temperature reaction system. This was further supported by the observation of more HO˙ generation in the room temperature Fenton reaction system than that in the high temperature alkaline H₂O₂ system (Fig. S3, ESI†).

Fig. 7 The relationship between the H₂O₂ and HO˙ concentration. H₂O₂ (1.1 × 10⁻⁵ to 1.1 × 10⁻³ mol l⁻¹); BA (4 × 10⁻⁴ mol l⁻¹); T = 80 °C; pH = 10.0; t = 60 min.

3.4.2. The working curve for the quantitative detection of HO˙. To achieve a quantitative detection of HO˙, a pure single hydroxylated fluorescent product HBA was separated to obtain the working curve. The product HBA was prepared from the reaction of BA with alkaline H₂O₂ at 98 °C, and purified by recrystallization from water and characterized by ¹H NMR, ¹³C NMR, and ESI-MS (Fig. S4–S6, ESI†). The disappearance of the aromatic proton signal of BA in the ¹H NMR spectrum clearly demonstrated the substitution by the hydroxyl group, agreeing well with the ESI-MS analysis.

Based on the one-to-one correlation between the HO˙ and HBA amounts, the working curve for the quantitative detection of HO˙ was obtained by measuring the fluorescence intensity of HBA in a series of concentrations (4.06 × 10⁻³ to 20 μmol l⁻¹) at pH 10. As displayed in Fig. 8, a good linear relationship between the HBA concentration and the fluorescence intensity was gained with R² = 0.99979. It can be described by the equation

Y = 144.66042X + 0.407, (1)

where X and Y refer to the concentration of HO˙ (μmol l⁻¹) and the corresponding fluorescence intensity, respectively. Thus, by measuring the fluorescence intensity, the HO˙ amount can be easily calculated, making it possible for the quantitative detection of HO˙.

Fig. 8 The standard curve between the fluorescence intensity and HBA concentration.

3.4.3. The application of BA for the evaluation of HO˙ generated in H₂O₂ bleaching systems. BA was then used as a novel fluorescent probe for the practical quantitative detection of HO˙ in alkaline H₂O₂ high temperature bleaching systems. At the same time, the HO˙ generation was also evaluated in an activated H₂O₂ bleaching system with a classic activator tetra-acetylethylenediamine (TAED).^28,29 This indicates that only small amount of HO˙ (1.0 × 10⁻⁶ mol l⁻¹) was produced from 2.2 × 10⁻⁴ mol l⁻¹ H₂O₂ (Table 1) in the absence of TAED. The HO˙ yield, however, became enhanced apparently with increasing TAED concentration up to 100 mg l⁻¹, and became constant with more TAED addition (Fig. 9). Compared to the regular H₂O₂ bleaching system, the HO˙ amount was increased 8 times with the addition of 100 mg l⁻¹ TAED. These findings revealed that TAED greatly promoted the HO˙ yield in the alkaline H₂O₂ high temperature reaction system. All measurements showed an excellent reproducibility (RSD 1.5–3.4%) for HO˙ detection in the high temperature H₂O₂/TAED reaction system. The developed quantitative detection method for HO˙ will be helpful for better understanding the bleaching mechanism of the high temperature alkaline H₂O₂ system.

Table 1 The quantitative detection of HO˙ generated in the H₂O₂ and H₂O₂/TAED systems

[TAED], mg l^−1	[HO˙], μmol l^−1	X̄, μmol l^−1	RSD, %
0	1.05, 1.03, 1.01, 1.07, 1.02	1.04	2.26
10	1.87, 1.76, 1.86, 1.83, 1.92	1.85	3.19
20	2.71, 2.62, 2.55, 2.67, 2.78	2.66	3.35
30	4.55, 4.41, 4.35, 4.59, 4.61	4.5	2.56
40	5.51, 5.36, 5.46, 5.61, 5.53	5.49	1.68
60	6.67, 6.71, 6.57, 6.78, 6.82	6.71	1.46
80	7.71, 7.32, 7.43, 7.52, 7.69	7.53	2.28
100	8.27, 8.13, 8.1, 8.46, 8.39	8.27	1.9
200	8.89, 8.62, 8.92, 9.16, 8.82	8.88	2.19
300	9.62, 9.5, 9.37, 9.73, 9.68	9.58	1.52
400	9.83, 9.6, 9.89, 9.62, 9.94	9.78	1.6

---

Fig. 9 The quantitative detection of HO˙ generated in the H₂O₂ systems at various TAED concentrations. TAED (0–400 mg l⁻¹); H₂O₂ (2.2 × 10⁻⁴ mol l⁻¹); BA (4 × 10⁻⁴ mol l⁻¹); T = 80 °C; pH = 10.0; t = 60 min.

3.4.4. The effect of interferences. Cotton fabrics are treated during practical high temperature hydrogen peroxide bleaching. Thus, the potential interferences of several organic components including starch, pectin and polyvinyl alcohol that exist in the cotton fabrics and the inorganic metal ions in the tap water used in the industrial bleaching process were studied. As displayed in Fig. 10, no obvious interference was encountered from 50 times of K⁺, Cl⁻, Na⁺, SO₄²⁻ (in molar ratio of H₂O₂), and a certain amount of starch, pectin, polyvinyl alcohol, Fe³⁺, Ca²⁺, Mg²⁺, Al³⁺, Sn²⁺ and Zn²⁺. However, some metal ions such as Fe²⁺, Co²⁺ and Cu²⁺ showed interference, which may result from complex formation between BA and the metal ions, leading to a decrease of the fluorescence intensity. Actually, most metal ions in the higher concentration form a precipitate with OH⁻ in the alkaline hydrogen peroxide bleaching system. The interference experiment proved that most components that exist in the industrial high temperature hydrogen peroxide bleaching have little effect on the detection of HO˙ with BA. The detection method has a satisfactory selectivity and stability in the determination of HO˙.

Fig. 10 The interferences of the organic components and inorganic ions in the detection of HO˙. (0) Blank. (1–4) K⁺, Cl⁻, Na⁺, SO₄²⁻ (1.1 × 10⁻² mol l⁻¹); (5–7) starch, pectin, polyvinyl alcohol (50 mg l⁻¹); (8–16) Fe³⁺, Ca²⁺, Mg²⁺, Al³⁺, Sn²⁺, Zn²⁺, Fe²⁺, Co²⁺, Cu²⁺ (2 × 10⁻⁵ mol l⁻¹). BA (4 × 10⁻⁴ mol l⁻¹); H₂O₂ (2.2 × 10⁻⁴ mol l⁻¹); T = 80 °C; pH = 10.0; t = 60 min.

4. Conclusions

A quantitative detection for HO˙ was proposed based on a novel fluorescent probe BA in this work. The present probe BA was for the first time used to quantitatively determine HO˙ in a high temperature alkaline H₂O₂ bleaching system. It proves that BA, with only one reaction site, can overcome the defects of a poor reproducibility and a lack of quantitative detection which are encountered by a previous fluorescent probe TA. The pure hydroxylated product HBA was obtained to establish the standard curve between the HO˙ concentration and the fluorescence intensity. A fluorimetric method for the quantitative detection of HO˙ was developed with an excellent reproducibility under severe conditions. It is simple, rapid, sensitive, and available to precisely detect HO˙ in a wide temperature range 20–98 °C. The present fluorescent probe will be helpful for monitoring the HO˙ yield under practical complex conditions, but also for understanding the H₂O₂ bleaching mechanism.

Acknowledgements

The work was a part of the national science and technology support program (no. 2009BAE88B02) and is supported by the National Ministry of Science and Technology. This work is also supported by the Graduate Student Degree Thesis Innovation Foundation Projects of Donghua University (Grant no. CUSF-DH-D2013050).

References

1. M. Dizdaroglu and P. Jaruga, Free Radical Res., 2012, 46, 382.
2. J. Du and J. M. Gebicki, Int. J. Biochem. Cell Biol., 2004, 36, 2334.
3. L. Yuan, W. Y. Lin and J. Z. Song, Chem. Commun., 2010, 46, 7930.
4. F. Ay, E. C. Catalkaya and F. Kargi, J. Hazard. Mater., 2009, 162, 230.
5. N. H. Zhang, Z. T. Zhang, M. D. Bai, C. Chen, X. Y. Meng and Y. P. Tian, Mar. Pollut. Bull., 2012, 64, 2742.
6. F. C. Cheng, J. F. Jen and T. H. Tsai, J. Chromatogr., B: Anal. Technol. Biomed. Life Sci., 2002, 781, 481.
7. B. Tang, L. Zhang and Y. Gen, Talanta, 2005, 65, 769.
8. L. X. Li, Y. Abe, Y. Nagasawa, R. Kudo, N. Usui, K. Imai, T. Mashino, M. Mochizuki and N. Miyata, Biomed. Chromatogr., 2004, 17, 470.
9. L. X. Li, Y. Abe, K. Kanagawa, T. Shoji, T. Mashino, M. Mochizuki, M. Tanaka and N. Miyata, Anal. Chim. Acta, 2007, 599, 315.
10. K. Q. Zhou, J. J. Yin and L. L. Yu, Food Chem., 2006, 95, 446.
11. C. J. Miller, A. L. Rose and T. D. Waite, Anal. Chem., 2011, 83, 261.
12. C. Tai, J. F. Peng, J. F. Liu, G. B. Jiang and H. Zou, Anal. Chim. Acta, 2004, 527, 73.
13. E. Kilinc, Talanta, 2005, 65, 876.
14. N. Soh, Anal. Bioanal. Chem., 2006, 386, 532.
15. X. Q. Chen, X. Z. Tian, I. Shin and J. Yoon, Chem. Soc. Rev., 2011, 40, 4783.
16. F. L. Ren, N. Wu and S. H. Si, Chin. J. Anal. Chem., 2001, 29, 60.
17. J. Fan, H. Q. Guo and S. L. Feng, J. Fluoresc., 2007, 17, 257.
18. X. F. Yang and X. Q. Guo, Analyst, 2001, 126, 1800.
19. E. Vidrio, C. H. Phuah, A. M. Dillner and C. Anastasio, Environ. Sci. Technol., 2009, 43, 922.
20. M. Saran and K. H. Summer, Free Radical Res., 1999, 31, 429.
21. X. H. Qu, L. J. Kirschenbaum and E. T. Borish, Photochem. Photobiol., 2000, 71, 307.
22. K. R. Millington and L. J. Kirschenbaum, Rev. Prog. Color. Relat. Top., 2002, 118, 6.
23. S. E. Page, W. A. Arnold and K. McNeill, J. Environ. Monit., 2010, 12, 1658.
24. S. A. V. Eremia, D. C. Lucia, G. L. Radu and J. L. Marty, Talanta, 2008, 77, 858.
25. W. J. Jiang, J. A. Joens, D. D. Dionysiou and K. E. O'Shea, J. Photochem. Photobiol., A, 2013, 262, 7.
26. T. Hirakawa, K. Yawata and Y. Nosaka, Appl. Catal., A, 2007, 325, 105.
27. L. Ren, H. K. Kim and W. W. Zhong, Anal. Chem., 2009, 81, 5510.
28. J. Z. Shao, Y. Huang, Z. H. Wang and J. Q. Liu, Color. Technol., 2010, 126, 103.
29. J. Y. Cai, D. J. Evans and S. M. Smith, AATCC Rev., 2001, 1, 31.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra45975f

---