Collaborative Construction of a Silver Nanocluster Fluorescent Probe Using the Pyridinium-Based Ionic Liquid [C4py][DCA].

A silver nanocluster fluorescent probe was synthesized by using the pyridinium-based ionic liquid [C4py][DCA] as the protective agent, AgNO3 as the precursor, and NaBH4 as the reducing agent. The presence of pyridine group enhanced the fluorescence intensity of Ag nanoclusters and facilitated the coordination interaction between Ag nanoclusters and AsO3 3-. Therefore, the collaborative construction of a silver nanocluster probe using the pyridinium-based ionic liquid [C4py][DCA] offered outstanding selectivity and sensitivity to detect AsO3 3- in water. More interestingly, the fluorescent probe quenched by AsO3 3- could be recovered with the addition of H2O2. This fluorescent probe provided a rapid and superior method for the detection of As(III) in the linear concentration range of 0-60 ppb with the lowest detection limit of 0.60 ppb. The mechanism of fluorescence quenching was a static quenching, considered to be due to electron migration between functional groups on the surface of Ag nanoclusters constructed with [C4py][DCA] and AsO3 3-.

Introduction

According to the Survey of the World Health Organization, 80% of human diseases are related to water pollution.[1,2] Trace elements in natural water are not only related to the geographical environment[3,4] but also related to human activities such as the direct discharge of domestic wastewater, industrial wastes, and medical wastewater.[5] Natural rock weathering and human exploitation and utilization of underground resources can bring arsenic into water sources, and the use of arsenic in chemical fertilizers and pesticides can also lead to trace arsenic in grains, vegetables, and water sources. Arsenic poisoning[6] can occur if people drink water with high levels of arsenic for a long time. Therefore, the arsenic content is one of the important analysis indexes of water quality. Accurate determination of the arsenic content in water can not only guarantee the water safety of humans and livestock but also help to formulate effective treatment and protection measures according to water quality status and changing trends.[7,8]

The toxicity of arsenic compounds is highly dependent on their form, and the toxicity of trivalent arsenic is much higher than that of pentavalent arsenic.[9] Therefore, by finding only the total arsenic content in water quality monitoring, one cannot effectively estimate the harm of arsenic and correctly evaluate the environmental quality. It is more important for the selective determination of trivalent arsenic.[10−12] At present, the main methods for the determination of arsenic contents are as follows: the colorimetric method with silver diethyl dithiocarbamate,[13] methylene blue,[14] and other traditional colorimetric reagents, atomic absorption spectroscopy,[15] inductively coupled plasma–mass spectrometry,[16] high-performance liquid chromatography,[17] and hydride generation atomic fluorescence spectrometry.[18,19] The first analytical method is cumbersome, takes a long time, has low sensitivity and a narrow detection limit, and the organic solvent used in the analysis is harmful to the human body. Other methods overcome the above shortcomings and can be used to measure the total amount of arsenic. At the same time, the use of large instruments and equipment in the abovementioned traditional detection methods restricts the real-time, in situ detection.

Nowadays, various synthesis methods of metal nanoclusters have been proposed, and based on surface-enhanced Raman scattering (SERS), spectral absorption, and fluorescence optical properties of these metal nanoclusters, the sensors used for the detection of toxic and harmful ions in water have been widely used because of their advantages of being rapid, simple, and with high sensitivity. Boruah[20] et al. and Banerjee[21] et al. detected As(III) in water using a colorimetric nanosensor with the LOD valued of 1 ppb and 0.86 ppb, respectively. Li[22] et al. constructed a portable colorimetric and a fluorescence nanosensor with the LOD values of 0.87 and 0.66 ppb, respectively, to detect As(III) in water. Pathan[23] et al. and Sun[24] et al. found a fluorescence nanosensor with the LOD values of 5.1 and 1 ppb, respectively. Although the LODs of colorimetric and fluorescence nanosensors of As(III) in water have reached the standard LOD of 10 ppb recommended by WHO, nanosensors with a lower LOD are still required for its detection in actual water. Song[25] et al. developed a SERS nanosensor to detect As(III) in the linear concentration range of 0.5–10 ppb with a LOD of 0.1 ppb. Although the LOD had obviously improved, the narrow linear range of concentrations still restricted its application in actual water. Therefore, it is imperative to develop an optical sensor for As(III) detection in water with a low LOD and a wide linear concentration range.

Ionic liquids have been widely used in the field of physical chemistry because of their unique properties,[26,27] such as non-combustion, very low vapor pressure, high thermal stability, a wide liquid range, a wide electrochemical window, and reusability. The excellent optical properties of ionic liquids containing metal elements have been recently reported, which indicated that it is feasible to construct metal nano-optical sensors by using ionic liquids.[28]

In this paper, a kind of silver fluorescent probe coconstructed with an ionic liquid was prepared: [C4py][DCA], a pyridine-based ionic liquid, and Ag nanoclusters were constructed in a cooperative construction mode. The positively charged silver nanoclusters were closely associated with AsO33– by a coordination bond between the lone electron pair of AsO33– and π molecular orbitals of the positively charged silver nanoclusters, and their original fluorescence can be quenched by the bond. More interestingly, AsO33– was oxidized to AsO43– with the addition of H2O2, and the original fluorescence gradually recovered with the disappearance of the lone electron pair of the anion.

Experimental Section

Reagents and Instruments

Table S1 in the Supporting Information shows the manufacturers and purity of all chemicals used. The synthesis of [C4py][DCA] was based on the previous work of our research group, and the specific synthesis method was reported in the literature.[28−31]

An ESCALAB 250 electron spectrometer with a monochromatic Al Kα excitation (hν = 1486.6 ev) source was used for XPS measurements of materials. A JEOL-2010 high-resolution transmission electron microscope was used for transmission electron microscopy (TEM) experiments. The fluorescence performance was measured using a F-4500 fluorescence spectrometer produced by Hitachi of Japan. A Shimadzu UV2000 UV–vis spectrometer was used to detect UV–vis spectra. A Nicolet380 Fourier transform infrared spectrometer was used to detect infrared spectra. A Malvern Zetasizer nano ZS90 nanoparticle size and potential detector was used to perform the dynamic light scattering (DLS) and apparent zeta potential experiments.

Preparation of Fluorescent Silver Nanoclusters

In a typical synthesis of silver nanoclusters, 5 mL of triple distilled water and methanol with a volume ratio of 1:1 was added to the reaction bottle and mixed evenly; 5 mL of 2 × 10–2 mol·L–1 AgNO3 solution and 2 mL 5 × 10–2 mol·L–1 [C4py][DCA] were added to the above mixture system in turns dropwise; after 30 min, 5 mL of 1 mol·L–1 NaOH solution was added dropwise into the above solution and was left to stand for 10 min; then, 0.5 mL of 0.1 mol·L–1 NaBH4 solution was added to the above mixture solution, which yielded silver nanoclusters after after 3 h and was named AgNC-[C4py][DCA] . The whole synthesis process was carried out in a thermostatic water bath, and the temperature was controlled at 40 °C.

Fluorescence Detection of AsO33–

The sensitive fluorescence detection of AgNC-[C4py][DCA] to AsO33– was as follows: 2 mL AgNC-[C4py][DCA] solution was mixed with 2 mL different concentrations of AsO33– from 10 to 600 μg/L, and 16 mL triple distilled water was added to incubate for 5 min. The fluorescence of the solution was measured in 2 mL AgNC-[C4py][DCA] solution mixed with 18 mL triple-distilled water as blank control.

The selective fluorescence detection of AgNC-[C4py][DCA] to AsO33– was as follows: 2 mL AgNC-[C4py][DCA] solution was added to 18 mL 60 μg/L AsO33– and interfering ion solution, respectively, the interfering ions include cations (Na+, Mg2+, Al3+, Fe3+, Ag+, and Cu2+) and anions (CO32–, SO42–, NO3–, PO43–, AsO43–, and PO33–), and the fluorescence detection conditions were the same as above.

The anti-interference ability detection of AgNC-[C4py][DCA] to AsO33– was as follows: 2 mL AgNC-[C4py][DCA] solution was added to 18 mL mixed solution of 10 μg/L AsO33– and 60 μg/L interfering ions, the interfering ions include cations (Na+, Mg2+, Al3+, Fe3+, Ag+, and Cu2+) and anions (CO32–, SO42–, NO3–, PO43–, AsO43–, PO33–), and the fluorescence detection conditions were the same as above.

The recovery experiment of the fluorescence probe AgNC-[C4py][DCA] was investigated by adding a certain concentration of hydrogen peroxide solution into the solution fluorescence quenched by AsO33–.

Results and Discussion

Characterization of AgNC-[C4py][DCA]

The size and morphology information of silver nanoclusters (AgNC-[C4py][DCA]) constructed in collaboration was obtained by TEM and DLS experiments. The data in Figures A and 3A showed that the size of the silver nanoclusters was small and uniform. The ordered lattice fringes with a spacing of 0.236 nm in Figure B were observed for the (111) crystal plane of silver. The TEM particle size distribution statistical results (Figure C) and DLS average results (Figure A) showed that the size distribution of silver nanoclusters was uniform, and the average size was 4 nm and 7 nm, respectively.

Figure 1

(A) TEM, (B) HRTEM, and (C) histogram of particle size distribution of AgNC-[C4py][DCA].

Figure 3

(A) DLS results of AgNC-[C4py][DCA], (B) FTIR spectra of (a) [C4py][DCA], (b) initial product of AgNC-[C4py][DCA] after dialysis purification, (c) final product of AgNC-[C4py][DCA] after repeated dialysis purification, (d) semipermeable extramembrane equilibrium solution of the initial product, and (e) semipermeable extramembrane equilibrium solution of the final product, and (C) apparent zeta potential of AgNC-[C4py][DCA].

The surface element composition and binding information of silver nanoclusters were detected by X-ray photoelectron spectroscopy and infrared spectroscopy. According to the full scan X-ray photoelectron spectral pattern of silver nanoclusters (AgNC-[C4py][DCA]) and one of the raw materials for synthesis ([C4py][DCA]) (Figure A), it can be seen that the N element in [C4py][DCA] was successfully linked to the surface of silver nanoclusters. The Ag 3d5/2 XPS peak in Figure B could be effectively divided into two peaks corresponding to the binding energies[32] at 368.8 and 369.5 eV, respectively. As can be seen from the binding energy data of the N 1s XPS peak of AgNC-[C4py][DCA] and [C4py][DCA] in Figure C, as [C4py][DCA] formed AgNC-[C4py][DCA], the binding energy of N 1s decreases from 399.1 to 398.6 eV, which means that the density of electron cloud around N atoms increased during this process.[33] Combined with the two splitting peaks of Ag 3d3/2 in Figure B, the two species of Ag in the synthesized Ag nanoclusters can be classified as two silver species, one was the free Ag nanoparticles[34] with a binding energy 368.8 eV and the other was Ag nanoparticles with a binding energy of 369.5 eV bound to N atoms in [C4py][DCA].

Figure 2

(A) Full scan X-ray photoelectron spectrum pattern and (C) N 1s high-resolution XPS patterns of AgNC-[C4py][DCA] and [C4py][DCA], (B) Ag 3d high-resolution XPS patterns of AgNC-[C4py][DCA].

Figure B showed the infrared spectral data of various materials of the silver nanoclusters in the synthesis and purification process. As shown in the infrared spectrum of Figure B, the absorption peak at 3443 cm–1 was attributed to the stretching vibration of the hydroxyl group (O–H). Because this peak was hardly visible in the infrared absorption spectrum of raw material [C4py][DCA] and O element was not seen in the full XPS spectrum of AgNC-[C4py][DCA], the infrared absorption peak at 3443 cm–1 of each substance in Figure B was attributed to the stretching vibration peak of the hydroxyl group (O–H) in water and the surface hydroxyl group (O–H) of silver nanoclusters. The absorption peak at 1635 cm–1 was attributed to C–N stretching vibrations, which was similar to the pyridine structure. The peak has been discovered in the infrared absorption spectra of the raw material ([C4py][DCA]) and different purities of silver nanoclusters AgNC-[C4py][DCA], and it has not been seen in the infrared absorption spectra of equilibrium solution outside the semipermeable membrane, so the phenomenon was explained as that 1-butylpyridine cations ([C4py]+) remained in the synthesized silver nanocluster AgNC-[C4py] [DCA] system by binding to silver atoms on the Ag nanocluster surface. Instead, the series of stretching vibration absorption peaks between 2124 and 2222 cm–1 were attributed to the cyano group (C≡N), which disappeared in the infrared absorption spectrum of the final product of AgNC-[C4py][DCA] after repeated dialysis purification. These results indicated that the anions of dicyandiamide ([DCA]−) were not bonded with the silver nanoclusters, so they left the silver nanocluster AgNC-[C4py][DCA] system through the semipermeable membrane during the purification process. The absorption peaks of the solution outside the semipermeable membrane after the initial purification were further evidence of the above process. The apparent zeta potential of AgNC-[C4py][DCA] in Figure C was +48.6 mV, indicating that the silver nanoclusters were stable. In combination with the infrared spectrogram information, the positive electrical property of the silver nanocluster AgNC-[C4py][DCA] was due to the bonding of 1-butylpyridine cations ([C4py]+) to silver atoms on the surface.

The optical properties of AgNC-[C4py][DCA] were detected by UV–vis and fluorescence spectrophotometry. As shown in Figure A, the absorption peak near 400 nm was classified as the surface plasmon resonance peak (SPR) of Ag nanoclusters with a small diameter,[34−36] which was consistent with the results of DLS and TEM. In addition, the fluorescence excitation and emission spectrum of AgNC-[C4py][DCA] (Figure B) showed that the maximum excitation and emission wavelengths were 347 and 448 nm, respectively, and the fluorescence emission spectrum with the best intensity of AgNC-[C4py][DCA] can be obtained at an excitation wavelength of 347 nm (Figure C). At the same time, it can be seen from Figure S1 that the fluorescence performance of the silver nanocluster was very stable under different concentrations of NaCl solution and UV lamp irradiation at different times.

Figure 4

(A) UV–vis, (B) fluorescence excitation and emission spectra, and (C) excitation-dependent photoluminescence emission spectra of AgNC-[C4py][DCA].

Detection of AsO33–

As a kind of fluorescence sensor of AsO33– in water, AgNC-[C4py][DCA] had excellent sensitivity and anti-interference ability. Figure displays the sensitivity of AgNC-[C4py][DCA] to AsO33– investigated by adding different concentrations of AsO33– to the AgNC-[C4py][DCA] solution. When the concentration of AsO33– increased from 1 to 60 ppb, the fluorescence intensity of AgNC-[C4py][DCA] decreased gradually (Figure A), indicating that AsO33– gradually quenched the fluorescence of AgNC-[C4py][DCA].

Figure 5

(A) Fluorescence emission spectra of AgNC-[C4py][DCA] at 347 nm with different concentrations of As(III), (B) linear relationship between the fluorescence change ratio and the AsO33– concentration, and (C) fluorescence decay curves of AgNC-[C4py][DCA] without and with AsO33– (the two samples were labeled AgNC-[C4py][DCA] and AgNC-[C4py][DCA]-As in the figure).

The plots with different concentrations of AsO33– as the abscissa corresponding to the fluorescence change ratio (F/F0) as the ordinate showed a very good linear relationship after three parallel experiments when the concentration of AsO33– was between 1 and 60 ppb (the linear equation and linear correlation coefficient were F/F0 = 0.90097–0.01101C and R2 = 0.9973, respectively), where the fluorescence change ratio (F/F0) is the ratio of the fluorescence intensity of AgNC-[C4py][DCA]-As (F) to that of original AgNC-[C4py][DCA] (F0, blank). The quenching curve was well fitted with a quenching constant of 4.12 × 10–2 ppb–1 using the Stern–Volmer equation, quinine sulfate (Φ = 54% in 0.1 mol L–1 H2SO4) was chosen as the reference substance, and the fluorescence yields of AgNC-[C4py][DCA] and AgNC-[C4py][DCA]-As were calculated as 41 and 7%, respectively, by using the following formula.

To explore the fluorescence-quenching mechanism of AgNC-[C4py][DCA] by AsO33–, time-correlated single-photon-counting experiments were performed to determine the fluorescence decay behavior of AgNC-[C4py][DCA] in the absence and presence of As(III). As shown in Figure C, the fluorescence decay of AgNC-[C4py][DCA] without and with AsO33– could be well fitted with the following exponential decay function to yield a life time of 7.31 and 7.03 ns, respectively. The measured fluorescence lifetimes (7.31 and 7.03 ns) did not noticeably change, suggesting a possible static quenching mechanism of AgNC-[C4py][DCA].[37]

Figures S4 and 6A display the fluorescence emission spectra and the fluorescence change ratio F/F0 of AgNC-[C4py][DCA] at 347 nm on addition of 60 ppb of different cations and anions, respectively, where the fluorescence change ratio is the ratio of the fluorescence intensity of the fluorescence probe after adding 60 ppb ionic solution to AgNC-[C4py][DCA] (F) to that of original AgNC-[C4py][DCA] (F0, blank). It could be clearly found that except for AsO33–, other cations (Na+, Mg2+, Al3+, Fe3+, Ag+, and Cu2+) and anions (CO32–, SO42–, NO3–, PO43–, AsO43–, and PO33–) exerted a weak effect on the fluorescence performance of AgNC-[C4py][DCA]. The above results showed that AgNC-[C4py][DCA] is sensitive to AsO33–, and it could be used as a fluorescent sensor to detect AsO33– efficiently.

Figure 6

(A) Fluorescence change ratio F/F0 of AgNC-[C4py][DCA] at 347 nm on addition of different cations and anions at a concentration of 60 ppb, (B) detection result of anti-interference ability of AgNC-[C4py][DCA] for detecting As(III). Fluorescence change ratio F/F0 of AgNC-[C4py][DCA] with the addition of 10 ppb AsO33– (light blue square column), 60 ppb of different interference cations or anions (purple square column), and mixed solution of 10 ppb of AsO33– and 60 ppb of interfering ions (dark blue square column). Note: all data in Figure are statistical results of three parallel experiments.

In order to study the practicability of AgNC-[C4py][DCA] as a fluorescence sensor for detecting AsO33–, the anti-interference ability in a complex environment was studied, and the results are shown in Figures S5 and 6B. The interferers were 60 ppb various cations (Na+, Mg2+, Al3+, Fe3+, Ag+, and Cu2+) and anions (CO32–, SO42–, NO3–, PO43–, AsO43–, and PO33–). Figures S5 and 6B display the fluorescence emission spectra and the fluorescence change ratio F/F0 of AgNC-[C4py][DCA] with that addition of 10 ppb AsO33– and upon the subsequent addition of 60 ppb of different interference ions, respectively, where the fluorescence change ratio is the ratio of the fluorescence intensity of F (after adding 10 ppb AsO33– and upon the subsequent addition of 60 ppb of interference ions) to that of F0 (original AgNC-[C4py][DCA], blank). As shown in Figures S5 and 6B, not only the interference ions had no obvious effect on the fluorescence performance of AgNC-[C4py][DCA] but also the fluorescence-quenching intensity of AgNC-[C4py][DCA] hardly changed after adding six times the concentration of interference ions to 10 ppb AsO33–. The above results indicated that the coexistence of AsO33– and most interference ions did not affect the quantitative detection of AsO33– by AgNC-[C4py][DCA]. Therefore, as a kind of fluorescence sensor for quantitative detection of the AsO33– content in water, AgNC-[C4py][DCA] had not only high sensitivity and selectivity but also strong anti-interference ability, which is feasible in reality.

At present, many methods for the detection of As have been proposed. However, traditional arsenic detection methods have great defects because they can not effectively distinguish the highly toxic arsenic trivalent. In recent years, the development of various optical sensors has realized the selective detection of As(III) to a certain extent, as shown in Table . AgNC-[C4py][DCA] is one of the best metallic fluorescent materials in linear range for detecting As(III) in water. At the same time, as shown in Figure S6, AgNC-[C4py][DCA] also is recyclable in a way other metallic nanofluorescent materials and carbon dots were not.

Table 1

Performance Comparison of Different As(III) Optical Sensors

sensor	mode	linear range (ppb)	LOD (ppb)	recyclable or not	reference
Fe-GODs	fluorescence	5–100	5.10	not	(23)
Au@Ag NPs	SERS	0.5–10	0.1	not	(25)
DTT-Au NRs	colorimetry	9.7–749.9	2.80	not	(38)
DTT-Fe3O4@Au	colorimetry	0–20	0.86	not	(21)
CDs/TTCA-QDs	fluorescence	0–100	1	not	(24)
CDs/TMT-Au NPs	dual mode (fluorescence and colorimetry)	0–100/50–100	0.66/0.87	not	(22)
AgNC-[C4py][DCA]	fluorescence	1–60	0.60	recyclable	this work

Possible Mechanism for the Fluorescence Response of AgNC-[C4py][DCA] to As(III)

By comparing the full scan X-ray photoelectron spectrum of solution AgNC-[C4py][DCA], AgNC-[C4py][DCA]-As, and AgNC-[C4py][DCA]-As-hp (Figure A), we found that with the addition of As(III), two new peaks appeared near 44 and 144 eV, while these two new peaks disappeared as H2O2 continued to be added to the above solution. According to the handbook of X-ray photoelectron spectroscopy, these two peaks were, respectively, attributed to As 3d and 3p. Besides, as was shown in the As 3d high-resolution XPS patterns of AgNC-[C4py][DCA]-As (Figure B), the binding energy of As was near 44.1 eV, which can be attributed to trivalent arsenic.[39] Before the full scan X-ray photoelectron spectra were detected, the three kinds of samples of AgNC-[C4py][DCA], AgNC-[C4py][DCA]-As and AgNC-[C4py][DCA]-As-hp were purified by dialysis, so As(III) was found in the AgNC-[C4py][DCA]-As, which illustrated that the As(III) was connected to the large size of the Ag nanoclusters, instead of entering the extramembrane equilibrium solution through the semipermeable membrane, and the As element was not checked out in the sample with re-entering H2O2, which illustrated that As(V), made from As(III) oxidized by H2O2, was separated from large Ag nanoclusters and purified by a semipermeable membrane into the extramembrane equilibrium solution. The Ag 3d3/2 high-resolution XPS peaks of the three samples in Figure D can be effectively divided into two peaks with binding energies of 368.8 and 369.5 eV corresponding to free and N-bonded Ag atoms, respectively. The area ratios of the two peaks of the three samples were all near 2, which indicated that the binding mode of Ag atoms did not change significantly in the process of adding As(III) and H2O2 to AgNC-[C4py][DCA] successively. Besides, as can be seen from Figure C, the binding energy of the N 1s high-resolution XPS peak of AgNC-[C4py][DCA]-As (trivalent arsenic was present in the solution) was significantly smaller than that of samples AgNC-[C4py][DCA] and AgNC-[C4py][DCA]-As-hp (trivalent arsenic was absent in either solution). Therefore, it can be concluded that As(III) was bound to large-size Ag nanoclusters by bonding with N atoms, which was the reason that it can not enter the extramembrane equilibrium solution through the semipermeable membrane during purification. As shown in Figure S2A, the apparent zeta potential of AgNC-[C4py][DCA]-As was 0 mV, which indicated that the positive charge of the Ag nanoclusters was neutralized by the negative charge of arsenite ions.

Figure 7

(A) Full scan X-ray photoelectron spectrum pattern, (C) N 1s and (D) Ag 3d high-resolution XPS patterns of the three samples with AgNC-[C4py][DCA], after adding As(III) (the sample is labeled as AgNC-[C4py][DCA]-As) and after re-entering H2O2 (the sample is labeled as AgNC-[C4py][DCA]-As-hp), and (B) high-resolution As 3d XPS patterns of AgNC-[C4py][DCA]-As.

With the addition of As(III) and H2O2 successively, the peak shape and half peak width of the absorption peak of pyridine structure were changed at 1635 cm–1, while the absorption peak at 1489 cm–1 disappeared first and then reappeared. Combined with XPS results, the change of the infrared absorption peak near 1500 cm–1 can be attributed to the interaction between As(III) and N in AgNC-[C4py][DCA] (Figure A). Based on the data in Figure S6, the loss rates of fluorescence intensity of AgNC-[C4py][DCA]-As-hp without and with semipermeable membrane purification were calculated to be 8.5% and 4.4%, respectively. Combined with the information in Figures B,C, S2A,B, and S3A,B, it can be seen that with the addition of As(III), the positivity of Ag nanoclusters decreased, which led to coalescence and fluorescence quenching. After adding H2O2 to oxidize As(III), the above process was partially reversible.

Figure 8

(A) FTIR spectra of AgNC-[C4py][DCA], AgNC-[C4py][DCA]-As, and AgNC-[C4py][DCA]-As-hp, (B) DLS results, and (C) apparent zeta potential of AgNC-[C4py][DCA]-As-hp.

As shown in Table S3, after three rounds of recovery and purification, the probe was used to detect 60 ppb of As(III) ions, and the spiked recoveries were calculated as 105, 111.67, and 126.67% using the equation F/F0 = 0.90097–0.01101C. Meanwhile, in tap and river water samples, the recoveries for As(III) detection ranged from 93 to 108%, as well as the relative standard deviation was less than 3.78%, illustrating that the fluorescent probe AgNC-[C4py][DCA] was relatively accurate for detecting As(III) in environmental water samples.

According to the fluorescence performance and structure analysis of the above materials, the working mechanism of AgNC-[C4py][DCA] is shown in Figure , including the selective response to As(III) (fluorescence quenching) and the fluorescence recurrence after adding H2O2.

Figure 9

Working mechanism of AgNC-[C4py][DCA] including the selective response to As(III) (fluorescence quenching) and the fluorescence recurrence after adding H2O2.

Conclusions

In summary, AgNC-[C4py][DCA] was synthesized as a fluorescent probe with high selectivity and sensitivity for quantitative detection of the As(III) content in water. It showed excellent performance in terms of detection limit and linear range. What is more, the probe can be recycled with hydrogen peroxide. At the same time, the working and recovery mechanisms of the AgNC-[C4py][DCA] probe were proposed based on the full combination of fluorescence performance, XPS spectrum, infrared spectrum, zeta potential, and other structural performance detection results, which provided a new method for the design, synthesis, and recovery of various optical probes in accordance with the idea of green chemistry.