Evolution of Silver-Mediated, Enhanced Fluorescent Au-Ag Nanoclusters under UV Activation: A Platform for Sensing.

Here, we report the synthesis of dopamine (DA)-mediated Au-Ag bimetallic nanoclusters in aqueous solution under UV activation. The success story emerges from monometallic fluorescent nanocluster evolution from photoactivation of gold as well as silver precursor compounds along with DA. The intriguing fluorescence property of the nanocluster relates to facile incorporation of Ag in Au, showing a 6-fold enhancement of the emission profile than simply DA-mediated Au nanoclusters. Silver effect, which is classified under the synergism, is the main reason behind such enhancement of fluorescence. The as-synthesized nanoclusters are robust and can be vacuum-dried and redispersed for repetitive application. The intriguing fluorescence of bimetallic nanoclusters is found to be quenched selectively in the presence of sulfide ion in an aqueous medium, paving the way for nanomolar detection of sulfide in water. The utility of the sensing platform has been verified employing different environmental water effluents.

Introduction

Metal nanoclusters are classified under metal nanoparticles of less than 2 nm diameter. Because of their tiny size, the energy levels split into a discrete state. This causes the extraordinary optical, catalytic, electronic, and magnetic properties of metal nanoclusters.[1] The metal nanoclusters are widely implemented in the fields of sensing, catalysis, bioimaging, nanodevice, bioconjugation, and so forth.[2−7] Coinage metals, specially gold and silver nanoclusters, have become an interesting fluorescent probe. They have sizes comparable to the Fermi wavelength of an electron and exhibit molecular-like behavior. They show size-dependent fluorescence. Jellium model is used to describe the size-dependent electronic structure and relative electronic transitions of nanoclusters.[8] However, silver nanoclusters show brighter luminescence over gold nanoclusters synthesized by the same motif.[9] Dickson et al.[10,11] have synthesized fluorescent Ag nanoclusters with two to eight atoms. Different templates such as peptides, gelatin, proteins, DNA, and so forth are widely used for the synthesis of fluorescent Au and Ag nanoclusters.[12] The presence of both gold and silver in the same platform results in synergistic interaction, causing much higher fluorescence than individual gold or silver fluorescing nanoclusters. Pradeep et al. synthesized the mercaptosuccinic acid-protected bimetallic Ag7Au6 nanocluster through a galvanic exchange reaction[13] Zhang et al. synthesized Au–Ag nanoclusters from the hydrothermal method and used the bimetallic cluster selective analysis of Hg2+ and Cu2+ ions.[12] Wang et al. prepared highly fluorescent Au–Ag nanoclusters using lipoic acid to sense Fe(III).[14] In some of these cases, the silver effect has been utilized to achieve intense fluorescence. Shi et al.[15] have demonstrated that in gold catalysis, the “silver effect” provides more catalytic activity than individual gold nanoparticles. However the main problem during the synthesis is that the reactions involve multiple steps, and quite a long time is needed for the completion of the reaction. Hence, we report a simple and facile synthetic process for the synthesis of fluorescent bimetallic nanoclusters.

Noble metal nanoclusters are widely used for catalysis,[16] bio as well as chemical sensing,[17,18] cell imaging,[19] and several other fields of applications. Sensing of hazardous metals or anions in the environment is much needed to prevent pollution. Sulfide ion is widely found in nature as industrial wastage and as a harmful water pollutant.[20] Excessive exposure to sulfide causes chronic disorders in the respiratory tract, blood, eyes, skin, and digestive system and also causes headaches, vertigo, impaired hearing, and autonomic dysfunction.[21] There are several techniques involving voltametric,[22] chromatographic,[23] potentiometric,[24] electrochemical,[25] fluorometric,[26] and colorimetric[27] methods to detect sulfide ions. Among these, fluorescence technique is considered as the most sensitive and selective technique.

The synthesis of fluorescent nanoclusters is quite a time-consuming process that involves multiple steps. However, in our work, we have synthesized a bimetallic nanocluster through ultraviolet (UV) treatment of an aqueous mixture of dopamine (DA), Au(III) salt, and Ag(I) salt through a simple technique. The synthetic procedure has been achieved within 1 h of time. DA has been used as a moderate reducing agent for the stepwise reduction. The concentration of silver ions in the reaction medium has been varied to achieve the most intense fluorescence. The silver effect has been seen to be effective in this case. Synthesis of gold nanoclusters has also been done using DA and Au(III) salt under the same experimental conditions. It is found that the presence of silver(I) in the reaction medium always produces better fluorescent particles than that without silver(I) salt. The as-obtained fluorescent clusters are then characterized properly. It is found that the fluorescence of the as-synthesized gold–silver nanoclusters gets quenched in the presence of sulfide ions in an aqueous medium. This selective as well as sulfide ion concentration-dependent quenching of the fluorescence of gold–silver nanoclusters paves the way for sulfide sensing. A linear detection range of 0.17 mM to 6.7 μM has been achieved. Hence, our prescribed method of synthesis and application of gold–silver nanoclusters are of academic as well as of practical use.

Experimental Section

Materials and Instruments

All reagents were of AR grade. Triple distilled water was employed throughout the experiment. Chloroauric acid (HAuCl4), silver nitrate (AgNO3), DA, and sodium salt of anions were obtained from Sigma-Aldrich. All glassware were cleaned with freshly prepared aqua regia, subsequently rinsed with distilled water, and dried well before use.

All ultraviolet–visible (UV–vis) absorption spectra were recorded in an Evolution 201 spectrophotometer (Thermo Scientific). The absorbance was measured using a glass cuvette. At room temperature (25 °C), the fluorescence measurement was done with a PerkinElmer LS55 fluorescence spectrometer. Fluorescence lifetimes were measured with an Easy life V fluorometer (Optical Building Blocks Corporation) equipped with a 380 nm LED excitation source. A nonlinear least squares (χ2) fit was tested to determine the fit of the decay rate to a sum of exponentials, and a visual inspection of the residuals and the autocorrelation function were used to determine the quality of the fit. The sample was taken in a quartz cuvette with a path length of 1 cm for fluorescence measurement. VG Scientific ESCALAB MK II spectrometer (UK) equipped with a Mg Kα excitation source (1253.6 eV) and a five-channeltron detection system was used for the X-ray photoelectron spectroscopy (XPS) analysis. The samples were fridge-dried before the XPS measurement. Transmission electron microscopy (TEM) analyses were done with an H-9000 NAR instrument (Hitachi) having an accelerating voltage of 300 kV. The samples were drop-casted onto a carbon-coated copper grid, and the grid was vacuum-dried before loading into the microscope. Fourier transform infrared spectroscopy (FTIR) of the samples was done using a Thermo-Nicolet continuum FTIR microscope. All experiments were repeated three times at room temperature in an almost neutral medium.

Synthesis of DAAuAg Fluorescent Cluster

DA-mediated AuAg fluorescent assembly was synthesized under UV irradiation of a mixture of AgNO3, DA, and HAuCl4 solutions. Typically, 1.2 mL of 10–2 M HAuCl4 solution, 0.4 mL of 10–2 M AgNO3 solution, and 2 mL of 2.3 × 10–3 M aqueous solution of DA were mixed in a 20 mL beaker, and the final volume of the solution was made 16 mL with distilled water. This mixture was kept under UV light of wavelength 365 nm under stirring for 1 h. After 1 h, a yellow solution with a brown precipitate was obtained. This solution was subjected to centrifugation to separate the solution and the solid product. The solution shows fluorescence and hence used for further characterization. The solvent was removed under vacuum, and the as-obtained solid mass upon redispersion in the required solvents was used for further experimentations.

Fluorescence Detection of S2– Ion in Aqueous Medium

An aliquot of the DAAuAg cluster (0.50 mg/1 mL) was added individually to the aqueous solution containing different concentrations of the S2– ion. The solution was mixed homogeneously and kept undisturbed for 6 h at room temperature. After 6 h, the solutions were taken for spectroscopic measurement. The interfering effects of other metal ions were investigated individually in the presence of the fluorescent cluster solution. During the study of the interference, other anions were added at a concentration of 7 × 10–4 M.

Water samples were obtained from the nearby locality and centrifuged to remove any heavy particles. Then, a series of samples was prepared by spiking standard solutions containing various concentrations of S2– in water samples. The resulting solutions were further mixed with fluorescent cluster solutions. After 6 h of mixing, the spectroscopic data were collected.

Results and Discussion

Formation and Characterization of Fluorescent Bimetallic Nanoclusters

Coinage metals in their sub-nanometer dimension become fluorescent because of interband transition. It is found that bimetallic nanoclusters show furthermore intense fluorescence. Again, it is reported that bimetallic alloy particles are more fluorescent compared to the bimetallic core–shell structure. Hence, we ventured to synthesize a bimetallic nanocluster with intriguing fluorescence. Intriguing 6-fold enhancement of the fluorescence property of the bimetallic nanoclusters, compared to gold nanoclusters, is reported here. A stable fluorescent bimetallic AuAg cluster has been prepared in an aqueous medium using HAuCl4, AgNO3, and DA under UV irradiation for 1 h. Au(III) and Ag(I) salts are taken in a 3:1 molar ratio. During the course of the reaction, simultaneous generation of large Au particles and tiny Au–Ag clusters occurs. The large Au particles are deposited as a precipitate. From X-ray diffraction (Figure S1, Supporting Information), the presence of Au(0) has been confirmed while Ag(0) is absent in the precipitate. The precipitate does not show any fluorescence and henceforth is not considered for further experimentation. The pale yellow supernatant shows fluorescence at 440 nm when it is excited at 360 nm (Figure ). The fluorescent nanoclusters are termed as DAAuAg. This fluorescence remains undisturbed in terms of intensity as well as peak position even after 2 months of preparation (Figure S2, Supporting Information). This states that the as-synthesized particles are thermodynamically stable and kinetically inert. In the UV–vis spectra, a peak is observed at 271 nm, but any characteristic plasmonic peak for Au or Ag is absent. This indicates the absence of large Au or Ag nanoparticles that show a plasmonic peak. If the particle size is in sub-nanodimension, the characteristic plasmonic band is not observed for plasmonic metal nanoparticles.[28]

Figure 1

Absorption and fluorescence spectral profile of DAAuAg at room temperature in an aqueous medium. λex = 360 nm.

However, the reducing agent DA itself is fluorescent. So, the question may arise whether the fluorescence comes from DA or from the produced metallic nanoparticles. To ensure the origin of fluorescence, the same reaction was done with NaOH and K2Cr2O7 instead of the Au(III)–Ag(I) salt mixture under the same experimental conditions. The resultant product does not show any fluorescence, rather the inherent fluorescence of DA is quenched. Even with extended reaction time, no significantly fluorescent product was obtained (Figure S3, Supporting Information). Both NaOH and K2Cr2O7 can oxidize DA. Also, the control experiment was performed by treating DA under UV irradiation for 1 h in the absence of any metal salt. The product of photoactivation is also nonfluorescent and shows a broad peak at ∼274 nm in the UV–vis spectral profile (Figure S3, Supporting Information). Only when both Au(III) and Ag(I) are present at a 3:1 molar ratio, the immensely fluorescent particles are obtained. Two phenolic groups of DA after two-electron transfer reduce both Au(III) and Ag(I) to Au(0) and Ag(0) and itself gets oxidized into dopamine-quinone (DQ).[29] Theoretically calculated oxidation potential of DA is +0.63 eV.[30] The redox potential value indicates facile reduction of Au(III) [EAu(III)/Au(0)0 = +1.50 eV] compared to Ag(I) [EAg(I)/Ag(0)0 = +0.79 eV]. One of the plausible pathways of this reaction is DA mediated. Initially, DA reduces Au(III) to Au(0). Because of the facile interaction between Ag and N,[31] Ag(I) ions get attached to the Au(0)–DA/DQ surface and catalyzes the Au(0) formation process. Later, it gets reduced to Ag(0) and prompts bimetallic cluster formation. However, Sun et al.[28] have suggested that in such a situation, the reduction of Ag(I) to Ag(0) can occur, and Au(III) may partially get reduced to Au(I). This supposition is further supported by the report of Duo et al.[33] and Mohanty et al.[32] They had concluded that the added Ag(I) helps in bridging of the Au(I)-ligand motif during the progress of the reaction, and the preformed Au nanoclusters are incorporated into the Au(I)–Ag(I) ligand network, resulting in the formation of fluorescent Au–Ag bimetallic clusters. However, XPS studies (discussion regarding the nature of this bimetallic cluster has been discussed in the trailing part) reveal that Au and Ag are present in their zero oxidation states. From the redox potential values, it is clear that the formation of Ag(0) in the presence of Au(III) is not thermodynamically facile. The formation of Ag(0) is an example of the antigalvanic reaction[34] of reduction of Ag(I) to Ag(0) in the presence of the Au–DQ system. Sun et al.[28] have suggested that the antigalvanic reaction between Ag(I) and Au(0) happens probably because of the electrochemical potential difference in the fine nanocluster system.

The presence of excess ligand may cause dissolution of Au atoms from the preformed Au nanocluster surfaces and form Au(I)–N complexes. These Au(0)–N or Au(I)–N complexes can tune the thermodynamically unfavorable reduction of Ag(I). The silver-induced strong metallophilic interaction may be the cause of initiation of Au–Ag alloy formation.[35,36] However, the true mechanism of cluster formation is still up for debate.

The measured quantum yield (QY) is 6.45% (standard is quinine sulfate in 1 M H2SO4). Wu et al.[37] reported that the high QY in Au nanoclusters is due to surface ligands, which take part in a charge-transfer interaction with core Au nanoparticles via electron-rich atoms or groups present in the ligand. Hence, we consider that the electron rich −NH2 group interacts with the as-synthesized DAAuAg nanoclusters and provides stability, and intense fluorescence is observed. The “silver effect” phenomenon is another factor on which the fluorescence and stability of a gold–silver nanocluster depends. Silver effect is reported to show both increase[38] and decrease[28] in the fluorescence of bimetallic nanoclusters compared to gold nanoclusters. In our case, the enhancement in fluorescence is observed. Silver concentration-dependent fluorescence of DAAuAg particles was compared with the emission of DAAu particles. The gold nanoclusters prepared using the DA template under UV irradiation (1 h) without silver is termed as DAAu. The synthetic procedure remained the same with the only difference being that silver is absent in DAAu. The particles show fluorescence at 430 nm when excited at 360 nm. No characteristic plasmonic peak for Au particles is found in the UV–vis spectra (Figure ). A peak at ∼277 nm is obtained for the quinine form of DA, DQ. This peak is found for the DA control solution, DAAuAg solution, and DAAu solution. However for both monometallic gold and bimetallic gold–silver particles, the characteristic plasmonic peaks of Au and Ag are absent. However, when DA is mixed with Au(III), a characteristic plasmonic peak of Au(0) is observed at ∼490 nm at room temperature, although no fluorescence signal is observed (Figure S3, Supporting Information). The Ag(I) concentration-dependent increment in the fluorescence intensity can be understood considering the QY and lifetime values. During the measurement of the QY value, it is found that the QY value is higher when Ag(I) is present in the reaction medium. The QY value of DAAu is 2.03%, whereas the QY values for the products with different Au(III)–Ag(I) concentration ratios are 6.29, 6.45, 6.1, 5.6, 5.4, and 5.01% for [Au(III)]–[Ag(I)] = 3:0.5, 3:1, 2:1, 3:2, 6:5, and 1:1, respectively. The lifetime values also vary widely for these products. The average lifetime values are found be 0.97, 3.12, 4.49, 4.05, 3.89, 4.01, and 3.67 ns for [Au(III)]–[Ag(I)] = 1:0, 3:0.5, 3:1, 2:1, 3:2, 6:5, and 1:1, respectively (Table S1). So, for different concentrations of Ag(I) in the reaction mixture, the optical property as well as lifetime values are different.

Figure 2

(A) Absorption and emission spectral profile of DAAu particles at room temperature in an aqueous medium. (B) Fluorescence spectra of the as-synthesized particles at different molar ratios of Au(III) and Ag(I) salts in the reaction medium. At a concentration ratio of 3:1, the bimetallic nanocluster (DAAuAg) shows maximum fluorescence, which is 6-fold higher than the fluorescence of gold nanoclusters (DAAu). λex = 360 nm. (C) Fluorescence emission maximum vs precursor salt concentration ratio plot. λex = 360 nm.

The fluorescence emission peak of DAAuAg (Au–Ag = 3:1) is found to be slightly red-shifted compared to DAAu nanoclusters, but the emission is 6-fold higher. This red shift indicates the alloy formation. It is observed that even at different molar ratios of Au and Ag, the fluorescence intensity is higher than that of DAAu nanoclusters (Figure ). The fluorescence of the bimetallic nanocluster is mainly based on Au nanoclusters, and a certain concentration of Ag intensifies the emission via the silver effect. With excess Ag(I) concentration, the fluorescence gradually decreases. Again, the duration of exposure is another factor. It is seen that 1 h of exposure of the reaction mixture results in maximum emissive particles (Figure S4, Supporting Information). Further exposure causes more aggregation, resulting in nonfluorescent products. Fluorescence lifetime measurement of DAAuAg shows that the decay profile can be fitted in a biexponential curve. The average lifetime is 4.49 ns (Figure ). TEM image shows that the particles are homogeneously distributed with an average diameter of 2–3 nm (Figure ). In the high-resolution TEM (HRTEM) image, no distinct lattice mismatch is observed (Figure , inset)[12] This indicates that there may be alloy formation. FTIR spectra show the peak for −C=O, which indicates that the reduction goes through the formation of a quinine derivative (Figure S4, Supporting Information). From XPS studies, it is found that Ag and Au are present in their zero oxidation states. The peaks at 368.81 and 374.4 eV stand for Ag(0) 3d5/2 and Ag(0) 3d3/2, respectively. Peaks for Au 4f7/2 and Au 4f5/2 are found at 83.8 and 87.01 eV, respectively (Figure ). This indicates the partial presence of Au(I) in the nanoclusters. The as-synthesized nanoclusters are so robust that they can be vacuum-dried and redispersed for repetitive use.

Figure 3

(A) Fluorescence decay profile of DAAuAg at room temperature in an aqueous medium. λex = 360 nm. (B) TEM image of DAAuAg; inset: HRTEM image.

Figure 4

(A) Broad-range XPS spectra of DAAuAg. Narrow-range XPS spectra of elemental (B) Au and (C) Ag. The spectra are taken under fridge-drying conditions.

Sensing of the Sulfide Ion in Aqueous Medium

The stability of the as-synthesized nanoclusters was tested in the presence of different anions. It is observed that the intense fluorescence of DAAuAg is quenched significantly in the presence of sulfide ions selectively (Figure ). Sulfide ion has an inherent affinity toward Au. In the presence of the sulfide ion, there occurs aggregation of the particles. From the TEM image, this aggregation is obvious (Figure ). This hampers the emission from the DAAuAg particles. Sulfide ion continues to show quenching in the presence of other anions also. Even thiosulfate ion does not interfere in the sulfide ion-induced quenched emission (Figure ). The relative fluorescence quenching for 0.67 mM and 6.7 nM is 43.3 and 92.45%, respectively. The XPS studies reveal that there is a slight deviation in the peak positions of Au and Ag because of interaction with sulfide (Figure S5, Supporting Information). On increasing the sulfide concentration from 0 to 0.67 mM, a straight regression line is obtained for a range of 0.17 mM −6.7 μM (Figure ). Interestingly, no interference is observed from other anions. This may be due to the fact that the sulfide ion is much smaller in size, so it easily interacts with tiny nanoclusters causing effective quenching. The comparative account for the detection limit of sulfide ion in an aqueous medium through different sensing techniques has been recorded in Table S2. It is interesting to note that in spite of facile interaction of Cl– with Ag(I), there occurs no interference in the presence of Cl– in the solution. Probably, the difference between the Ksp values of AgCl and Ag2S is the reason.[39]

Figure 5

(A) Fluorescence spectra of DAAuAg in the presence of different anions in an aqueous medium. λex = 360 nm. [Anions] = 7 × 10–4 M, [DAAuAg] = 0.5 mg/mL. (B) TEM image of DAAuAg in the presence of S2–.

Figure 6

(A) Effect of sulfide ion on the fluorescence spectra of DAAuAg in presence of other anions in the interference study. [Anions/S2–] = 7 × 10–4 M, [DAAuAg] = 0.5 mg/mL. (B) Bar diagram of the interference study. λex = 360 nm.

Figure 7

(A) Fluorescence spectral profile of DAAuAg for different sulfide concentrations. (B) Relative fluorescence intensity vs sulfide ion concentration plot; inset: linear detection range. λex = 360 nm. [DAAuAg] = 0.5 mg/mL.

Real Sample Analysis

The as-synthesized probe is further tested on real water samples. Water samples were collected from different sources. Initial test for the presence of sulfide was done. It was found that drain water contained a trace of sulfide. Water samples were spiked with the sulfide salt through the standard addition method. Then, the fluorescence was measured. This technique was repeated three times. The obtained relative standard deviation values allow us to conclude that the nanoclusters can be successfully used for real sample analysis (Table ).

Table 1

Determination of the S2– Concentration in Different Water Samples Using Our Proposed Strategy

water sample	amount of standard S2– added (mg/mL)	total S2– recovered (mg/mL)	% recovery (%)	relative error
tap water (collected from the laboratory tap)	0.04	0.039	97.5	2.5
	0.06	0.058	96.7	3.3
drinking water (collected from the department)	0.04	0.041	102.5	2.5
	0.06	0.059	98.3	1.7
drain water (collected from the department)	0.04	0.043	107.5	7.5
	0.06	0.062	103.3	3.3

Previously, there were few reports of the evolution of the tiny gold nanoparticle with the bioactive molecule DA, with and without photoactivation.[40,41] Stabilization of the gold nanoparticle in solution was also studied. Our work further confirms the previous report of DA-induced Au nanoparticle synthesis, and the fluorescence property of the synthesized nanoparticles was rediscovered, followed by the synthesis of bimetallic nanoclusters under photoactivation. The as-synthesized robust fluorescent nanoclusters have been employed for selective and sensitive sulfide ion detection in aqueous solution. Figure schematically shows the whole procedure.

Figure 8

Schematic representation of the synthesis and application of DAAuAg.

Conclusions

In a nutshell, we have synthesized intriguingly fluorescent AuAg bimetallic nanoclusters with the help of “silver effect” in a one-pot synthetic method. Proper molar ratio of gold and silver brings intense fluorescence in the as-synthesized nanoclusters. The bimetallic alloying has been achieved using a simple ligand DA. Also, use of the as-synthesized clusters for probing different anions in an aqueous medium shows that sulfide ion exclusively quenches the fluorescence without the interference of other anions. The use of the as-synthesized nanoclusters has been further used for real samples.