A turn-on fluorescent probe with a dansyl fluorophore for hydrogen sulfide sensing

Hydrogen sulfide (H2S) is a biologically relevant molecule that has been newly identified as a gasotransmitter and is also a toxic gaseous pollutant. In this study, we report on a metal complex fluorescent probe to achieve the sensitive detection of H2S in a fluorescent “turn-on” mode. The probe bears a dansyl fluorophore with multidentate ligands for coordination with copper ions. The fluorescent “turn-on” mode is facilitated by the strong bonding between H2S and the Cu(ii) ions to form insoluble copper sulfide, which leads to the release of a strongly fluorescent product. The H2S limit of detection (LOD) for the proposed probe is estimated to be 11 nM in the aqueous solution, and the utilization of the probe is demonstrated for detecting H2S in actual lake and mineral water samples with good reproducibility. Furthermore, we designed detector vials and presented their successful application for the visual detection of gaseous H2S.


Introduction
Hydrogen sulde (H 2 S) is a highly toxic air pollutant that is widely generated from industrial production processes, involving sulfuric acid, sulfur, dyes, and cosmetics, as well as the microbial degradation processes by which anaerobic bacteria reduce inorganic sulfates and organic suldes. 1 Prolonged exposure to H 2 S may lead to respiratory paralysis, olfactory fatigue, vagueness of consciousness, and even permanent cerebral injury. 2,3 In addition, H 2 S is a newly iden-tied endogenous gaseous transmitter molecule like NO and CO, which has been associated with the regulation of cardiovascular, vasodilation, central nervous, respiratory, and immune systems. [4][5][6][7][8][9][10] The level of H 2 S in humans has been documented to be between 10 and 100 mM, and an abnormal concentration of H 2 S has been related with Down syndrome, diabetes, Alzheimer's disease, and arterial and pulmonary hypertension. [11][12][13] Therefore, analytical approaches capable of conducting selective and sensitive H 2 S detection in complicated environmental and biological systems are essential. Current techniques employed for H 2 S analysis include chemical titration, 14,15 colorimetry, 16 electrochemical assay, 17 gas chromatography, 18 and inductively coupled plasma atomic emission spectroscopy (ICP-AES). 19 Compared to these methods, uorescence analysis has proven to be a promising option because of its high sensitivity, high temporal and spatial detection resolutions, and the ability for conducting in situ monitoring of reactive and transient target analytes. [20][21][22][23][24][25][26][27] Generally, the uorescence probes for H 2 S mainly based on reduction reactions, 28-32 nucleophilic addition reactions, [33][34][35] and thiolysis reactions. [36][37][38] In addition, metal displacement approach utilizes the strong affinity of metal ions with suldes ions to rapidly attain reaction equilibrium, and achieve the realtime detection of H 2 S. 39,40 Based on metal displacement approach of copper metal complexes, some colorimetric or uorescent sensors for H 2 S have been emerging in recent years, owning to the low solubility product of CuS (K sp ¼ 6.3 Â 10 À36 ). For example, Li et al. synthesized a Cu-complex probe (BODIPY-DPA-Cu) by attaching di-(2-picolyl)amine (DPA) and BODIPY dye, in the presence of H 2 S, the uorescence at 546 nm enhanced 19-fold in the PBS buffer (10 mM, pH 7.4). It showed a high sensitivity and selectivity for sulde. 41 Kim J. Y. and co-workers developed a Cucyclen-dansyl (Cu-CD) probe for the quantication of H 2 S in PBS buffer containing 10% DMSO with good selectivity among competitive anions. 42 Kaushik and co-workers designed a probe for selective detection of H 2 S by copper complex embedded in vesicles, different from the MDA mechanism, in which both metal and indicator get displaced upon binding of H 2 S with metal center. 43 Other relevant probes for H 2 S detection based copper metal complexes are listed in Table S1. † [44][45][46][47][48][49][50][51] The present work capitalizes on this property by combining a dansyl uorophore with a Cu(II) metal complex to design a displacement-reaction-based probe for H 2 S detection in a uorescent "turn-on" mode. The uorescent metal complex probe includes a multidentate ligand that can coordinate with Cu 2+ ions to produce a stable dansyl uorophore Cu-complex (DNS-Cu). The DNS-Cu complex exhibits a very weak background uorescence because the empty d orbital of paramagnetic Cu(II) can accept the excited state electrons in the DNS uorophore, and hence, block the process of ligand uorescence generation. In addition, H 2 S strongly binds with the Cu 2+ ions to form CuS and releases a free uorescent DNS uorophore, resulting in a greatly enhanced probe uorescence. The H 2 S limit of detection (LOD) for the proposed probe (DNS-Cu) is determined to be 11 nM, and the probe demonstrates good H 2 S selectivity, reproducibility, and anti-interference performance. In addition, the probe was used to develop detector vials for visualizing gaseous H 2 S, which demonstrates the applicability of the proposed uorescent probe for sensing H 2 S gas in the environment.

Instrumentation
The uorescence spectra were obtained on a PerkinElmer LS-55 luminescence spectrometer. Mass spectra were recorded on a Thermo Proteome X-LTQ MS. 1 H-NMR spectrum was acquired on a Brucker Avance 400 MHz, using CDCl 3 as solvent and tetramethylsilane (TMS) as internal standard. Fluorescence photos were taken under a UV lamp, with a Canon 350D digital camera.

Synthesis of the dansyl uorophore and its Cu(II) metal complex (DNS-Cu)
The synthesis processes of DNS and the DNS-Cu complex were illustrated in Scheme 1. Firstly, we added 2-chloroethylamine hydrochloride (65 mg) to a dichloromethane (10 mL) solution of dansyl chloride (150 mg) and triethylamine (Et 3 N; 117 mL). Aer stirring for 1 h, a pale-yellow oil was obtained by removing the solvent, and then puried via column chromatography with a PE : EA ratio of 4 : 1. The resulting product is herein denoted as compound 1 (107 mg, 61%). For the synthesis of DNS, we added 62.5 mg of compound 1, 37.6 mg of compound 2, and 27.6 mg of anhydrous K 2 CO 3 into 3 mL of a dimethylformamide (DMF) solution. Aer stirring the mixture for 24 h, 15 mL of ice water was poured into the mixture to generate a precipitate. The

Determination of the binding constant of the DNS-Cu complex
The stability constant K of the DNS-Cu complex was calculated from a Benesi-Hildebrand plot according to the following Benesi-Hildebrand equation. 52,53 Here, F max is the uorescence intensity of the free DNS, F is the uorescence intensity of DNS with Cu 2+ , F min stands for the uorescence intensity of DNS in the presence of excessive Cu 2+

Measurement of quantum yields (QYs) of the DNS-Cu complex
The QY was measured by using uorescein (F s ¼ 0.95 in 0.1 M NaOH) as reference and calculated using the following equation.  The uorescence spectra were obtained between the range of 400 nm and 700 nm under an excitation wavelength of 338 nm. The primary uorescence peak of the DNS-Cu probe obtained at 534 nm was employed for S 2À sensitivity testing. The uorescence ratio F/F 0 of the DNS-Cu probe, where F 0 is the value of F obtained with no added S 2À ions, was plotted versus the S 2À ions concentration for quantitative analysis. All data were performed three times under equivalent conditions, and the average values are calculated.

Detection selectivity of the DNS-Cu complex probe for H 2 S
The selective responses of the DNS-Cu probe for other related anions and small molecules containing thiol groups were carefully examined using the same testing procedure as was employed for S 2À . The stock solutions of these species ( , F À , Cl À , Br À , I À , PO 4 3À , or ClO À (10 equiv.) into the DNS-Cu probe (1.0 mM) solution. Then, 1 equiv. of S 2À ions (1.0 mM) was added into the mixture solution, followed by recording the uorescence intensity at 534 nm. The selectivity of the DNS-Cu probe was also examined for H 2 S and other different common gases. Gaseous H 2 S was obtained by slowly dropping phosphoric acid on sodium sulde. Carbon monoxide (CO) gas was achieved from the reaction of formic acid with concentrated sulfuric acid. Carbon dioxide (CO 2 ) gas was prepared from the titration reaction between dilute sulfuric acid and sodium bicarbonate. Sulfur dioxide (SO 2 ) was got through a quantitative reaction of sodium hydrosulde and concentrated sulfuric acid. Ammonia (NH 3 ) gas was generated through the chemical reaction between NH 4 Cl and Ca(OH) 2 . Nitric oxide (NO) and nitrogen dioxide (NO 2 ) gas were obtained from pure gas. Then, different of these gas samples were injected into the DNS-Cu probe solution using a syringe, recording the uorescence spectra subsequently and taking the uorescence photos by a digital camera.

Characterization of DNS and DNS-Cu
The structure of the DNS conrmed by the analysis of MS and 1 H-NMR ( Fig. S1 and S2 †). As illustrated in Scheme 1, the DNS uorophore exhibits a bright uorescence at 534 nm, aer coordinating with Cu 2+ , DNS-Cu displays a very weak uorescence (Fig. S3 †). The MS analysis veried the formation of the DNS-Cu complex. ESI-MS of DNS-Cu (m/z): calculated 526.1100 found 526.0400 (Fig. S4 †). The value of stability constant (K) for the DNS-Cu complex was determined to be 2.7 Â 10 4 based on Benesi-Hildebrand method, as seen in Fig. S5. †

Stability and sensitivity of the DNS-Cu probe for H 2 S
We rstly investigated the stability of the DNS-Cu complex probe before the sensitivity experiment. As noted from Fig. S6, † the uorescence intensity of the probe at 534 nm exhibits no distinct changes aer six consecutive irradiations for 60 min, indicating that the DNS-Cu complex probe is stable against photobleaching. When the probe solution (1.0 mM) is exposed to S 2À ions, as seen in Fig. 1A, the uorescence intensity of the probe greatly increased as the S 2À ions concentration increased from 0 to 1.375 mM. This can be attributed to the strong binding of S 2À with the Cu(II) metal center of the DNS-Cu complex to form stable CuS precipitation, subsequently releases the DNS uorophore. The MS and 1 H-NMR analysis proved the release of the DNS in Fig. S7, † and the quantum yields were increased from 1.8% to 25.5% aer the DNS-Cu probe reaction with S 2À (Fig. S8 †). In addition, the uorescence intensity exhibits a dose-response with increasing S 2À concentration up to an S 2À dose of 1.0 mM, aer which further increases in the S 2À concentration produce no further increase in the uorescence intensity. This indicates that the stoichiometric reaction between the DNS-Cu probe and S 2À was 1 : 1. The plot of F/F 0 versus the S 2À concentration in Fig. 1B exhibits a good linear relationship with a coefficient of determination R 2 ¼ 0.9988 in the S 2À concentration range of 0 to 1.0 mM. The LOD was estimated to be 11 nM based on the denition of LOD ¼ 3 Â S.D./k, where k is the slope of the curve in Fig. 1B, and 3 Â S.D. stands for 3 times standard deviation of the blank signal.

Selectivity of the DNS-Cu probe for H 2 S
As shown in Fig. 2A, the uorescence intensity of the DNS-Cu probe solution increased sharply aer the addition of S 2À (1.0 mM), whereas no apparent changes in the uorescence intensity were observed aer adding the NO 3 À , NO 2 À , SO 4 2À , SO 3 2À , SCN À , S 2 O 3 2À , F À , Cl À , Br À , I À , PO 4 3À , and ClO À anionic species (1.0 mM), indicating a high selectivity for S 2À ions. For further practical application, the anti-interference experiments of the probe were conducted by adding NO 3 À , NO 2 À , SO 4 2À , SO 3 2À (50 equiv.), and SCN À , S 2 O 3 2À , F À , Cl À , Br À , I À , PO 4 3À ClO À (10 equiv.) into the probe solution, respectively. Then 1 equiv. of S 2À ions (1.0 mM) were added subsequently into the mixture solution, followed by recording the uorescence intensity at 534 nm.
The results of the anti-interference experiments shown in Fig. 2B, indicating that the probe exhibits a good antiinterference against other anionic species.
In order to verify the practicability of probe in complex environments, some biological phosphates (PPi, ATP, ADP, AMP) and endogenous biomolecule (GSH, L-Cys) were investigated with DNS-Cu probe. The results show that the PPi, ATP, ADP and AMP could not turn on the uorescence of the probe system (Fig. S9 †). Even in the presence of 10 equiv. of PPi, ATP, ADP and AMP (10 mM), H 2 S (1.0 mM) also can enhance the uorescence of the DNS-Cu probe in the same manner as in the absence of these analytes. These results suggest that biological phosphates dose not interference the detection system. However, it is noted that GSH and cysteine could enhance the uorescence of the DNS-Cu probe, which may be attributed to the affinity of Cu 2+ with -SH groups. Fortunately, their interference with these thiol-containing molecules can be readily eliminated through treatment with dimethyl sulfoxide (DMSO), which oxidizes the thiol groups to form disuldes that have less affinity with Cu(II), which minimizes their interference effect (Fig. S10 †).

Potential reusability of the DNS-Cu probe
The uorescence of the DNS-Cu probe can be activated by the addition of S 2À , quenched by the addition of 1 equiv. of Cu(II), and then recovered again by adding another 1 equiv. of S 2À . As shown in Fig. 3, such uorescence on-off cycles with S 2À and Cu(II) could be repeated for 4 times with little degradation in the uorescence intensity (Fig. 3B), and their corresponding uorescent images recorded in Fig. 3A. The good reusability of the uorescent probe for the alternate detection of S 2À and Cu 2+ can be developed as a logic gate by operating the two inputs In(S 2À ) and In(Cu 2+ ), as indicated in Fig. 3C. Optical logic gates, such as the YES, 54 NOT, 55 AND, 56 OR, 57 NOR, 58 XOR, 59 and INHIBIT gates 60 have been investigated widely in recent years. Here, the two inputs can be set as 0 and 1 to represent the absence and presence of uorescence, respectively, where exposure to S 2À results in a uorescence on state, and the output readout is 1, while the uorescence would be turned off by exposure to Cu 2+ ions, leading to an output of 0. When there are no inputs of H 2 S and Cu 2+ to the initial logic solution of the probe DNS-Cu complex, it shows no uorescence, then the output signal is 0. While both of H 2 S and Cu 2+ exist at the same time, it still no uorescence, the output signal is 0 as well. Briey, no uorescence results from both inputs set as 0 or 1 simultaneously, the output signals were both 0, which match with the INHIBIT logic function. , F À , Cl À , Br À , I À , PO 4 3À and ClO À ), the dense bars represent the subsequent addition of 1 equiv. of S 2À (1.0 mM) into the mixture solution.

Application of the DNS-Cu probe to actual water samples
Spike and recovery experiments were conducted using actual lake and mineral water samples. The lake water samples were collected from Shushan lake and ltered through a 0.45 mm microporous lter to remove insoluble particles. The mineral water samples were bought from a local supermarket and used directly without any pretreatment. The original concentrations of S 2À in lake water and mineral water samples were rstly conrmed to be 4.2 nM and 5.3 nM by ICPMS (iCAP RQ). These values were much lower than the maximum allowable level of S 2À (15 mM) set by World Health Organization (WHO) for drinking water. Thus, they do not pose any health concerns.
Then the spiked recovery tests were conducted with three different S 2À concentrations (0, 500, and 1000 nM) in a mixed solution of water/ethanol (v/v ¼ 1 : 1). The average concentration and standard deviations of S 2À in the spiked lake and mineral water samples are presented in Table 1. The concentrations of sulde estimated in the non-spiked lake and mineral water samples were less than the LOD of the method. The two values are lower than the detection limit, suggesting that the sensitivity of the method has its own limitation and cannot be directly used to real water samples with trace contents, but can be used in the further recovery calculation. The recovery results in the lake water samples are slightly greater than 100%, which may be attributed to the microbial degradation process. Generally, the recovery rates ranged from 98.2% to 102.3% for    the lake and mineral water samples, which are statistically near 100%, and therefore validates the ability of the proposed probe for S 2À ions sensing in complex samples.

Visualization of H 2 S gas using detector vials
Since S 2À and H 2 S can quickly reach equilibrium in aqueous solution, this method can be applied to detect gaseous H 2 S in aqueous solution. For this measurement, we fabricated detector vials to achieve on-site and rapid detection of gaseous H 2 S. The vials itself has no uorescence with a rubber stopper, different levels of H 2 S gas (0, 0.5, 1.0, 5.0, 10, and 20 ppm) were syringed into the vials. The images as presented in Fig. 4 indicate that the uorescence color intensity in the vials increased with increasing H 2 S concentration. The limit of detection of this method was determined as 0.5 ppm, based on the minimum amount of H 2 S to produce a slightly different uorescent color, which could be visible by ve persons. The uorescence intensity of the detector vial achieved a maximum level at an H 2 S concentration of 20 ppm. In addition, the selectivity of the proposed probe for H 2 S gas was investigated by evaluating the incidence of uorescence when lling the detector vials with other common gases, such as N 2 , NH 3 , NO 2 , SO 2 , CO 2 , CO, and NO (100 ppm). As presented in Fig. 5. The results indicated that the injection of these other gaseous compounds has no substantial effect on the uorescence intensity of the probe, and that only H 2 S gas activated the probe uorescence. These results demonstrated a sensitivity and selectivity of this method for H 2 S determination in the gas state.

Conclusions
We have fabricated a dansyl-based copper complex for use as a "turn-on" uorescence probe in H 2 S sensing. The subsequent addition of H 2 S effectively snatches Cu(II) from the DNS-Cu complex, and thus releases a free dansyl moiety, leading to the uorescence enhancement of the probe. The LOD for this method was determined to be 11 nM in aqueous solutions. In addition, the probe provided good H 2 S detection results with actual water samples. Moreover, the good reusability of the uorescent probe for cyclical detection of H 2 S and Cu(II), which can be used to develop "INHIBIT" logic circuit.

Conflicts of interest
There are no conicts to declare.