Development of a new water-soluble fluorescence probe for hypochlorous acid detection in drinking water.

Responsive small-molecule fluorescence probe specific for target analyte detection is an emerging technology for food safety and quality analysis. In this work, we report a new water soluble small-molecule fluorescence probe (PG) for the detection of hypochlorous acid (HOCl) in drinking water samples. Probe PG was developed by coupling of a glucosamine into 10-methyl-10H-phenothiazine fluorophore with a HOCl-responsive C=N bond. The thioether is another recognition site that can be oxidized to be sulfoxide in water. Due to the specific reactions triggered by HOCl, probe PG's absorption band is blue shifted from 388 to 340 nm, and fluorescence at 488 nm is more than 55-fold enhanced. Probe PG features high fluorescence stability in PBS buffer with varied pH, fast response and high selectivity to HOCl. The application of the probe PG for HOCl detection in real-world samples is demonstrated by HOCl detection in drinking water, including tap water, purified water, and spring water samples. The recoveries of this method for HOCl detection in drinking water are in the range of 99.17-102.3%. This work thus provides a new method for HOCl detection in drinking water with high precision and accuracy.

Introduction

Water, accounting on average of 60% of the body weight, is an essential component of a healthy balanced diet because the body relies on it to function properly. In living organisms, water is involved in all body’s chemical processes, such as digestion, nutrients absorption, waste products excretion, and body temperature regulation. Drinking fluids, particularly the drinking water, are the main source (80%) of the body water. To ensure safe and high-quality drinking water to be supplied, disinfection of drinking water by chlorine has been widely adopted in water treatment for over several hundred years (Jia et al., 2019). In water, chlorine reacts to form hypochlorous acid (Cl2 + H2O ⇌ HOCl + HCl) and hypochlorite (HOCl ⇌ OCl− + H+). HOCl and OCl− are strong oxidants that have disinfecting ability to kill 99.9% waterborne pathogens, such as bacteria, viruses, protozoa, and helminths (Block & Rowan, 2020). Nevertheless, exposure to large amount of HOCl/OCl− could result in several disorders, such as irritation of the oesophagus, vomiting, asthma (Del Rosso & Bhatia, 2018). The major routes of HOCl/OCl− exposure are through drinking water (5.6–28.2 μM in most disinfected drinking-water), foods, and contact with items either bleached or disinfected with it. For rats were administered OCl− in drinking water, an oral LD50 was found to be 850 mg/kg of body weight. Therefore, monitoring the HOCl level in drinking water is essential because drinking water is one of the most important substances for all living organisms’ survival.

In the past few decades, a variety of conventional techniques (Wu et al., 2019, Wu et al., 2019, Wu et al., 2019, Zhang et al., 2018), such as chemiluminescence (Zhu et al., 2018), colorimetric (Ren et al., 2021, Shiraishi et al., 2021), electrochemical (Ordeig et al., 2005), and chromatographic (Watanabe et al., 1998) methods have been widely used for HOCl analysis in environmental and biological samples (Kwon et al., 2021). Recent research has also demonstrated that the responsive small-molecule fluorescence probes are useful tools for the detection of HOCl in water and biological samples (Bruemmer et al., 2020). In comparison with traditional methods, fluorescence analysis using responsive probes are featured with high sensitivity and selectivity, simplicity for implementation, and good biocompatibility (Kaushik et al., 2017, Wu et al., 2019, Wu et al., 2019, Wu et al., 2019, Zhang et al., 2020). The development of responsive small-molecule probes has also been an attractive research field in food quality and safety analyses (Yadav et al., 2020, Duan et al., 2021, Feng et al., 2021). As a result, several fluorescence probes have recently been engineered for HOCl detection through employing different sensing reaction mechanisms, including oxidations with oxime derivatives (Zhao et al., 2011), amines (including dibenzoylhydrazine) (Chen et al., 2010, Goswami et al., 2014, Liu et al., 2011), electron-deficit C=C bond (Chen et al., 2010, Park et al., 2013, Wu et al., 2016), p-methoxyphenol and palkoxyaniline (Setsukinai et al., 2003, Sun et al., 2008), and group 16 elements (such as S, Se, and Te) (Ge et al., 2020, Ikeno et al., 2019, Venkatesan and Wu, 2015). The probes were thus developed by incorporating one or two these HOCl-responsive moieties with well-established fluorophores (Gangopadhyay et al., 2018, Han et al., 2020, Li et al., 2020a, Li et al., 2020b), such as fluorescein, cyanine, coumarin, naphthalimide, and metal complexes (Wu et al., 2019, Wu et al., 2019, Wu et al., 2019, Zhang et al., 2018). HOCl detection was then achieved through recording the changes of fluorescence signals after the response reaction. Despite the rapid progress in the development of fluorescence probes for HOCl detection, drinking water analysis necessitates a reliable analytical probe for the detection of HOCl in pure water with rapid response, high sensitivity and selectivity.

In this work, we report a new water-soluble fluorescence probe (PG) for the detection of HOCl in drinking water samples. PG was developed by incorporating a glucosamine into 10-methyl-10H-phenothiazine fluorophore by a HOCl-responsive C=N bond (Scheme 1). The thioether in phenothiazine fluorophore is another responsive moiety for HOCl detection. The glucose-conjugated probe (PG) has very good water solubility, allowing for HOCl detection in aqueous solution (20 mM PBS buffer, pH 7.4). PG is almost non-fluorescent in PBS buffer, while the fluorescence intensity is significantly increased upon HOCl-triggered cleavage of C=N bond and oxidation of thioether within 5 s. Overall, PG has large Stokes shift (148 nm), excellent solubility and stability in aqueous solution, high sensitivity and selectivity, rapid fluorescence response to HOCl, allowing it to be used as a tool for the detection of HOCl in drinking water samples.

Scheme 1

Schematic illustration of the response mechanism of probe (PG) for HOCl detection.

Experimental section

Materials

10-Methyl-10H-phenothiazine-3-carbaldehyde was synthesized by following the published method (Feng et al., 2018). 10-Methyl-10H-phenothiazine and D(+)-glucosamine hydrochloride were purchased from Aladdin reagent Co. (Shanghai, China). Triethylamine, phosphorus oxychloride, metal ions (nitrate salts), anions (sodium salts) and sodium hypochlorite (NaOCl) were obtained from Sinopharm Chemical Reagent Co., Ltd. (China). Unless otherwise stated, solvents and reagents were of analytical grade from commercial suppliers and were used without further purification.

Synthesis of probe PG

Scheme S1 illustrates the stepwise synthesis procedure of PG. Briefly, triethylamine (0.15 mL) was added into D(+)-Glucosamine hydrochloride (0.216 g, 1.1 mmol) in 8 mL MeOH. The mixture was refluxed for 30 min, and then 10-methyl-10H-phenothiazine-3-carbaldehyde (0.241 g, 1 mmol) in 10 mL MeOH was added into the solution. The reaction mixture was refluxed for another 3 h to form a yellow precipitate. PG was then obtained in 67% yield after filtration and washing thrice with cooled MeOH. 1H NMR (DMSO‑d6, 400 MHz) δ (ppm) 8.08 (s, 1H), 7.52 (d, J = 1.7 Hz, 1H), 7.24–7.15 (m, 2H), 6.99 (d, J = 7.9 Hz, 3H), 6.51 (s, 1H), 4.88 (d, J = 5.3 Hz, 1H), 4.78 (d, J = 5.6 Hz, 1H), 4.69 (t, J = 7.2 Hz, 1H), 4.51 (t, J = 5.8 Hz, 1H), 3.73 (dd, J = 9.7, 5.6 Hz, 1H), 3.53–3.46 (m, 1H), 3.42 (d, J = 5.6 Hz, 1H), 3.35 (s, 3H), 3.24 (s, J = 9.7, 5.8, 2.1 Hz, 1H), 3.16 (dd, J = 11.3, 5.3 Hz, 2H), 2.82–2.75 (m, 1H). 13C NMR (DMSO‑d6, 100 MHz) δ (ppm) 160.98, 147.46, 145.03, 131.30, 128.79, 128.36, 127.29, 126.02, 123.35, 122.68, 122.04, 115.39, 114.78, 96.06, 78.58, 77.34, 75.03, 70.81, 61.75, 35.86. ESI-MS (positive mode, m/z) calcd for C20H23N2O5S+: 403.1328, [PG + H]+: found: 403.1313. M.p.: 188.1–189.1 °C.

General procedures for spectrometric analysis

Anions and biomolecules in deionized water (20 mM) were freshly prepared. HOCl solution was prepared by diluting the commercial NaOCl solution into deionized water and diluted solution was used as the stock solution stored at 4 °C (Chen et al., 2010). The HOCl concentration was determined by UV–vis analysis using its molar extinction coefficient (ε292 nm = 391 M−1 cm−1) (Chen et al., 2010). Singlet oxygen (1O2) was produced in a Na2MoO4-H2O2 system in 0.05 m carbonate buffer of pH 10.5 (Song et al., 2005). SIN-1 was used as the donor of peroxynitrite (ONOO−) (Wu et al., 2018, Yang et al., 2006). Commercial H2O2 (30%) solution was diluted with deionized water to prepare H2O2 stock solution. The concentration was measured by UV–vis using the molar absorption coefficient of H2O2 in water (ε240 nm = 43.6 M−1 cm−1) (Chen et al., 2017). Hydroxyl radical (•OH) was obtained through a Fenton reaction between ammonium sulfate and hydrogen peroxide (Wu et al., 2019, Wu et al., 2019, Wu et al., 2019, Zhuang et al., 2014). For spectrometric analysis, PG (10 mM) in 20 mM PBS buffer (pH 7.4) was added with different concentrations of analytes (total volume 3 mL). The solution was stabilized for 5 min before fluorescence and UV–vis analyses.

HOCL detection in drinking water samples

Bottled spring water and purified water samples were obtained from local supermarket. Tap water was obtained from the laboratory without any further treatment. To demonstrate the feasibility of probe PG for HOCl detection in these drinking water samples, PG (10 μM) in above drinking water was added with different concentration of HOCl (0, 0.01, 0.05, 0.1, and 0.5 mM). The mixture was stabilized for 5 min and then the fluorescence color changes of the solution was recorded under 365 nm UV light. For quantitative detection of HOCl in drinking water samples, tap water and bottled spring water were added with HOCl, and then the HOCl-spiked water samples were analysed by using PG (10 μM) as a probe.

Results and discussion

Synthesis and characterization of probe PG for HOCl detection

The development of responsive fluorescence probes for selective and sensitive detection of various analytes in drinking water samples is contributing significantly to the food analysis (Yadav et al., 2020, Duan et al., 2021, Feng et al., 2021). In this work, we report a water soluble small-molecule probe (PG) for HOCl detection in drinking water samples. The probe was developed through coupling of glucosamine into 10-methyl-10H-phenothiazine fluorophore with a HOCl-responsive C=N bond. The conjugation of glucose offered PG good water solubility. Together with the thioether recognition moiety in phenothiazine fluorophore, probe PG is capable of reacting with HOCl to cleave the C=N bond and oxidize the thioether to sulfoxide. It has been reported that the unbridged C=N bonds containing fluorescent dyes are usually non-emissive because the C=N isomerization is the predominant decay process of excited states (Zhang et al., 2018). As a result of the HOCl-mediated C=N cleavage and formation of aldehyde and sulfoxide functionalized phenothiazine fluorophore, fluorescence of PG is switched on and the changes of fluorescence intensity can be recorded for HOCl detection.

As shown in Scheme S1, probe PG was synthesized through a two-step reaction procedure. Chemical structure of PG was characterized by 1H NMR and 13C NMR, and HRMS characterizations (Figs. S1-S3). The reaction between probe PG and HOCl was then examined by HRMS analyses. As shown in Fig. S4A, MS analysis (positive model) of probe PG showed a molecular ionic peak at m/z = 403.1313 ([PG + H]+). Upon addition of HOCl, cleavage of C=N bond was observed immediately as the emergence of new peak at m/z = 242.0636 ([C14H12NOS]+) (Fig. S4B). After 5 min reaction, oxidation reaction was observed as the emergence of another peak at m/z = 258.0588 (C14H12NO2S)+) (Fig. S4C). The data of HRMS analysis suggest the cleavage of C=N bond and oxidation of thioether to sulfoxide after reaction of PG with HOCl (Scheme 1).

UV–vis absorption and fluorescence responses of probe PG to HOCl

To investigate the UV–vis absorption and fluorescence responses of PG to HOCl, PG (10 μM) in 20 mM PBS buffer was treated with HOCl at different concentrations, and then the UV–vis absorption and fluorescence spectra were obtained after 5 min reactions. As shown in Fig. 1A, in the absence of HOCl, PG exhibited two intense absorption bands centred at 286 and 388 nm. Upon addition of HOCl, both absorption bands were decreased. A new absorption band (λabs = 340 nm) was emerged, and the absorbance was gradually increased with the increasing concentration of HOCl (Fig. 1B). Maximum change of the absorption spectra at 340 nm was obtained when 0.7 mM HOCl was added into the PBS solution containing PG.

Fig. 1

UV–vis absorption and fluorescence responses of probe PG to HOCl in 20 mM PBS buffer of pH 7.4. (A) UV–vis absorption spectra of probe PG (10 μM) after reacting with increasing concentrations of HOCl (0–1 mM). (B) Absorbances of probe PG at 340 nm against the concentrations of HOCl. (C) Fluorescence spectra of PG (10 μM) after reacting with increasing concentrations of HOCl (0–1 mM). Inset: fluorescence color change of probe PG with (a) and without (b) HOCl. (D) Fluorescence intensities of PG at 488 nm against the increasing concentration of HOCl. Excitation was performed at 340 nm.

Upon excitation at 340 nm, PG displayed weak fluorescence at 488 nm (Stokes shift 148 nm) in PBS buffer (20 mM, pH 7.4) with a quantum yield of 0.48%. As shown in Fig. 1C, fluorescence of PG at 488 nm was emerged and increased upon addition of HOCl. Green color of the PG solution containing HOCl was observed, indicating that the enhanced fluorescence of PG in the presence of HOCl can be clearly perceived even by the naked eyes. The emission intensity was over 55-fold enhanced and reached to a plateau after PG reacting with 0.7 mM HOCl (Fig. 1D), which is in consistent with the results of UV–vis analyses. The quantum yield was then determined to be 8.52% after PG reacting with HOCl.

In PBS buffer, the fluorescence stability of probe PG within 32 h incubation and its fluorescence response to HOCl at different time point were investigated. As shown in Fig. S5, fluorescence intensity of PG in PBS buffer was not changed over 32 h incubation, while remarkable enhancement of PG’s emission was obtained after reacting with HOCl. Moreover, increment of fluorescence intensity of PG at 488 nm displayed a good linearity with the concentrations of HOCl in the range from 5 to 280 μM (R2 = 0.995) (Fig. S6). Based on the method defined by IUPAC (3σ/k) (Mocak et al., 1997), the detection limit of PG for HOCl was then determined to be 1.97 μM.

Selective response of probe PG to HOCl was then investigated by UV–vis absorption and fluorescence spectrometric analyses. In PBS buffer of pH 7.4, PG (10 μM) in water was added with 0.8 mM Br−, AcO−, Cl−, F−, S2−, NO2−, NO3−, 1O2, OH−, ONOO−, P2O74−, PO43−, HSO3−, SO32−, SO42−, HCO3−, HSO4−, H2PO4−, H2O2, •OH, Cys, Hcy, and GSH, respectively. As shown in Fig. 2A, negligible changes of absorption spectra were obtained after mixing of interference species with probe PG. In contrast, the absorption band was clearly shifted from 388 to 340 nm upon reacting with HOCl. Selective response of PG to HOCl was further confirmed by fluorescence spectra analyses. As shown in Fig. 2B, fluorescence of PG was remarkably increased upon reacting with HOCl over other interference species. Under UV-light, green color of the PG solution was observed after reacting with HOCl, while no color changes were obtained upon the addition of other interference species (Fig. 2C). In addition, response of PG to HOCl was not affected in the presence of other interference species (Fig. S7), indicating that PG can specifically detect HOCl in complicated water samples.

Fig. 2

UV–vis absorption and fluorescence selectivity of probe PG (10 μM) for HOCl detection over other interference species (0.8 mM) in 20 mM PBS buffer of pH 7.4. (A) Changes of UV–vis spectra of probe PG with and without various interference species and HOCl. (B) Fluorescence spectra response of probe PG to various interreference species and HOCl. Excitation was performed at 340 nm. (C) Fluorescence color images of PG in the absence and presence of various interference species and HOCl. These species include: 1. Blank, 2. HOCl, 3. Br−, 4. AcO−, 5. Cl−, 6. F−, 7. S2−, 8. NO2−, 9. NO3−, 10. 1O2, 11. OH−, 12. ONOO−, 13. P2O74−, 14. PO43−, 15. HSO3−, 16. SO32−, 17. SO42−, 18. HCO3−, 19. HSO4−, 20. H2PO4−, 21·H2O2, 22. •OH, 23. Cys, 24. Hcy, 25. GSH.

Time and pH dependent fluorescence response of PG to HOCl

Through monitoring the emission intensity change at 488 nm with excitation at 340 nm, time dependent fluorescence response of probe PG to HOCl was investigated in 20 mM PBS buffer of pH 7.4. As shown in Fig. 3A, PG displayed weak and stable emission in buffer solution. Upon the addition of HOCl, PG solution showed fast response by increasing the emission intensity and the intensity reached the maximum value within 5 s (Fig. 3A inset). The emission intensity then remained at a steady level until further addition of HOCl. The results indicate that the sensing reaction of PG to HOCl is very fast and the produced product is fluorescence stable in PBS buffer.

Fig. 3

Time and pH dependent fluorescence response of probe PG (10 μM) to HOCl. (A) Time-dependent fluorescence changes of PG in 20 mM PBS buffer of pH 7.4 upon the sequential addition of 0.3 mM, 0.6 mM, 1.0 mM of HOCl. (B) pH effects on the emission intensities of PG before and after reacting with HOCl (0.8 mM) in aqueous solution of different pH. Excitation and emission were performed at 340 and 488 nm, respectively.

Then, the pH effect on PG’s fluorescence response to HOCl was investigated in PBS buffer at different pH. As shown in Fig. 3B, PG displayed weak and stable emission in PBS buffer with pH from 3 to 11.5. Upon reacting with HOCl, remarkable enhancement in emission intensity was observed in pH 3–9. In the pH 9–10.5, response of PG to HOCl was gradually decreased, and in pH above 11, limited fluorescence response of PG to HOCl was observed. The result suggests that probe PG is able to detect HOCl in acidic, neutral, and weakly basic buffer solutions.

Detection of HOCl in drinking water samples

To validate the capability of probe PG for HOCl detection in real-world samples, PG was added into three kinds of drinking water samples, including bottled purified water, spring water, and tap water. Then, HOCl at different concentrations (0.01, 0.05, 0.1, and 0.5 mM) were added into the water samples. As shown in Fig. 4, under UV light, fluorescence colors of three drinking water samples were gradually changed to green with the HOCl concentration increasing from 0.01 to 0.5 mM. The enhancement of green fluorescence is attributed to the response of PG to HOCl in these three drinking water samples. Therefore, probe PG is promising to be used as a tool for the detection of HOCl by “naked eye” in drinking water samples.

Fig. 4

Fluorescence color responses of PG (10 µM) in drinking water samples (purified water, tap water, and spring water) spiked with different concentrations of HOCl (0, 0.01, 0.05, 0.1, 0.5 mM) under 365 nm UV light.

The practical utility of probe PG for quantitative detection of HOCl in drinking water samples was further evaluated. As shown in Table 1, HOCl concentrations were found to be 5.72 and 0 μM in tap water and spring water sample, respectively. Through investigating the HOCl-spiked drinking water, the recoveries were in the range of 99.17–102.30% and the relative standard deviation (RSD) of all detections was less than 2.99%. These results clearly indicate that PG can be used as a tool for HOCl detection in drinking water with high precision and accuracy.

Table 1

Detection of HOCl in drinking water samples using PG (10 μM) as a probe.

Samples	HOCl level (μM)	HOCl added (μM)	HOCl found (μM)	RSD (%, n = 3)	Recovery (%)
Tap water	5.72	30	36.3	1.95	101.61
		50	56.59	2.33	101.56
		70	75.09	1.10	99.17
Spring Water	0	30	30.69	2.23	102.3
		50	49.59	2.19	99.18
		70	71.13	2.99	101.62

Conclusions

In summary, a new water-soluble responsive probe (PG) for HOCl detection in drinking water was developed in this work. The probe was designed by conjugating glucosamine into 10-methyl-10H-phenothiazine fluorophore with a HOCl-responsive C=N bond. The reaction of PG with HOCl cleaved C=N bond and oxidized the thioether in phenothiazine within 5 s. More than 55-fold enhancement in emission at 488 nm was observed, which allowed for the HOCl detection in PBS buffer. The application of this probe for the detection of HOCl in drinking water was also validated by quantification of HOCl in tap and spring water samples. The successful development of this probe thus offers a reliable tool for HOCl detection in real-world food samples.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.