Polyethyleneimine-Templated Ag Nanoclusters: A New Fluorescent

Nov 8, 2012 - Ag nanoclusters functioned by hyperbranched polyethyleneimine have been developed as a new fluorescent and colorimetric platform for ...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/ac

Polyethyleneimine-Templated Ag Nanoclusters: A New Fluorescent and Colorimetric Platform for Sensitive and Selective Sensing Halide Ions and High Disturbance-Tolerant Recognitions of Iodide and Bromide in Coexistence with Chloride under Condition of High Ionic Strength Fei Qu, Nian Bing Li,* and Hong Qun Luo* Key Laboratory on Luminescence and Real−Time Analysis, Ministry of Education, School of Chemistry and Chemical Engineering, Southwest University, Chongqing 400715, P. R. China S Supporting Information *

ABSTRACT: Ag nanoclusters functioned by hyperbranched polyethyleneimine have been developed as a new fluorescent and colorimetric platform for sensitive and selective recognition of halide ions (e.g., Cl−, Br−, and I−). The recognition mechanism is based on the unique reactions between halide ions and the silver atoms. In particular, halide-induced oxidative etching and aggregation can produce a strong fluorescence quenching of Ag nanoclusters. This sensing system exhibits a remarkably high selectivity toward halide ions over most of anions and cations and shows good linear ranges and lower detection limits: the linear ranges are 0.5−80 μM for Cl−, 0.1−14 μM for Br−, and 0.05−6 μM for I−, respectively; the limits of detection for Cl−, Br−, and I−, at a signal-to-noise ratio of 3, are estimated to be 200, 65, and 40 nM, respectively. Specifically, Br− and I− could be recognized selectively in the coexistence with Cl− under the condition of higher ionic strength, which is a significant advantage in the detection of Br− and I− in real samples. In addition, the recognition of halide could be performed by the colorimetric method, which is also attractive and promising because of its simplicity, rapidity, reliability, and low cost. Furthermore, this sensing system has been applied successfully to the detection of Cl− in real water samples.

H

methods to recognize and monitor the halide ions via various approaches, such as design of selective receptors modifying the surface of nanoparticles or quantum dots15−17 and taking advantage of the macroscopic physical responses for Au and Ag nanoparticles.18−22 In order to distinguish those ultrasmall particles, nanoparticles smaller than 2 nm are usually called nanoclusters. In this size range, metal nanoclusters become molecular species and size-dependent strong fluorescent emission can often be observed upon photoexcitation in the UV−vis range.23−25 Nevertheless, little literature was reported on the application of nanoclusters in fluorescence or colorimetric detection of halide ions. Especially for Ag nanoclusters, they are supposed to react with halide ions easily due to the unique reactions between halide ions and the silver atoms of Ag nanoclusters, but it is interesting that Ag nanoclusters, encapsulated in different scaffolds including small molecules,26 protein,27 and DNA,28 have been verified to be stable in the medium of high concentrations of chloride sodium by testing ionic strength or

alide ions are ubiquitous and play essential roles in industrial, medical, and environmental processes. The specific determination of halide concentrations in natural resources is very important for both monitoring excessive and deficient halide levels, such as in soils and drinking water, which can, if left unchecked, result in several kinds of pollutions and physiological disorders. For example, chloride is a significant component in the characterization of the quality and extent of drinking water, and selective measurement of chloride in biological systems is equally as useful for the diagnosis and treatment of some diseases (e.g., cystic fibrosis and acid−base disorders).1,2 Now, iodide is often added to table salt as a source of iodine for preventing iodine deficiency disorders, but excess of iodine or iodide ingestion could also result in goiter and hypothyroidism as well as hyperthyroidism.3 However, much more is known about the toxicity of bromide, and hence, its determination is deemed equally as important.4 In view of the importance of the detection of halide ions in the environmental and physiological systems, many strategies have been proposed in the past decade, such as the traditional ion chromatography,5−7 spectrometry by artificial chromophore and fluorophore,8−10 electrochemical analysis,11−14 and so on. Benefiting from the advance of nanotechnology, nanometerscale sensors have arisen to be the most popular analytical © 2012 American Chemical Society

Received: September 5, 2012 Accepted: November 8, 2012 Published: November 8, 2012 10373

dx.doi.org/10.1021/ac3024526 | Anal. Chem. 2012, 84, 10373−10379

Analytical Chemistry

Article

silver mirror reaction. The formation of silver nanoclusters was characterized using photoluminescence spectroscopy, UV−vis spectroscopy, high resolution transmission microscopy (HRTEM), and X-ray diffraction (XRD) techniques. The UV−vis absorption spectrum shows that the PEI-capped Ag nanoclusters have two absorbance bands centered at 268 and 354 nm as shown in Figure 1, in which the absorbance band at

interference; it is thought that the polyvalent interactions between the ligands and the metal core could provide enhanced stability for the as-prepared clusters. Although this is an advantage for the stability of Ag nanoclusters in the application, it is also a disadvantage for insensitivity of halide ions. In this paper, hyperbranched polyethyleneimine (PEI)capped Ag nanoclusters were introduced as a new fluorescent and colorimetric platform for sensing halide ions. In comparison with the other well-known Ag nanoclusters stabilized by various templates,26−28 we reported, for the first time, the selective and sensitive quenching effects of halide ions on the fluorescence of the Ag nanoclusters. On the basis of the unique reactions between halide ions and the silver atoms and the different solubility constants (Ksp) of silver compounds, both of the fluorescence and characteristic absorption peak of PEI-stabilized Ag nanoclusters could be used as good indicators for the sensitive and selective detection of halide ions (e.g., Cl−, Br−, and I−, except for F− due to the large Ksp of AgF). Specifically, we also found that Br− and I− could be recognized selectively in coexistence with Cl− under the condition of higher ionic strength, which is a significant advantage in the detection of Br− and I− in real samples. As expected, this sensing system has been applied successfully to the detection of Cl− in real water samples (e.g., tap water and mineral water samples). Moreover, PEI has been shown to be a biocompatible and efficient gene delivery vehicle both in vitro and in vivo.29−31 Therefore, our biocompatible halide probe within nanometer scale would be a promising candidate for effective and versatile sensors in medicine and biotechnology.



Figure 1. UV−vis absorption spectrum of PEI-capped Ag nanoclusters. Inset: (a) is the photographs of diluted solutions of PEIcapped Ag nanoclusters in water under visible light and UV light; (b) is the HRTEM images of PEI-capped Ag nanoclusters (inset shows a single magnified Ag nanocluster).

354 nm is attributed to the oligomeric silver species,32−36 and without any characteristic surface plasmon band of larger Ag nanoparticles at around 400−500 nm. In the inset of Figure 1, (a) is the photographs of diluted solutions of Ag clusters in water, which is nearly colorless (or a very slight yellow color) under visible light, while it emits intense blue fluorescence under a UV lamp; (b) is the HRTEM images of obtained Ag nanoclusters, which displayed an average diameter of ∼1.8 nm. The maximum fluorescence excitation and emission wavelengths of PEI-capped Ag nanoclusters are 375 and 455 nm, respectively. The as-prepared PEI-capped Ag nanoclusters exhibit a quantum yield of 3.8% in ethanol calculated by use of quinine sulfate as a reference.37−39 The formation of small clusters in the reaction is further confirmed by the X-ray diffraction pattern, which shows the absence of characteristic peaks exhibited by silver nanoparticles of size 5−10 nm (Figure S1 in the Supporting Information). No characteristic peaks were observed for silver clusters formed in the synthesis.40−42 Establishment of the Fluorescence Sensing System for Halides Ions. The fluorescence of PEI-capped Ag nanoclusters is sensitive to the presence of halide ions due to the lower Ksp of silver halide compounds. Before investigation of the effects of the concentrations of halide ions on the fluorescence quenching, many tests were carried out to optimize the sensing conditions such as probe concentration, reaction time, and the effects of pH, temperature, and light. The fluorescence-quenched efficiencies of PEI-capped Ag nanoclusters at 455 nm upon addition of halide ions were used to optimize the sensing system. (I0 and I represent the fluorescence intensity of Ag nanoclusters in the absence and presence of halide ions, respectively; ΔI = I0 − I, ΔI is the change of fluorescence intensities of Ag nanoclusters upon addition of halide ions; ΔI/I0 is the fluorescence-quenched efficiency43 by halide ions). Fluorescence response to the concentration of probe should be chosen at first. From the relationship between the fluorescence intensity and the concentration of the PEI-capped Ag nanoclusters (Figure S2 in the Supporting Information), it

MATERIALS AND METHODS

Reagents. Silver nitrate (AgNO3), hyperbranched polyethyleneimine (PEI, Mw = 10 000, 99%), 4-(2-hydroxyethyl)-1piperazineethanesulfonic acid (HEPES), and formaldehyde (35 wt %) were purchased from Aladdin Reagent Co., Ltd., China. KCl, KBr, and KI were used in this work. The stock solutions of 1 mM of Cl−, Br−, and I− were prepared, respectively, from which halide ions were diluted to serial concentrations. Other reagents of analytical reagent grade were purchased from Chengdu Kelong Chemical Reagents Factory (China) and used as received. All solutions were freshly prepared before use, and Milli-Q water (18 MΩ cm) was used throughout the experiments. Synthesis Procedure. Typically, 100 μL of 0.094 g mL−1 PEI and 50 μL of 1 mM HEPES solution were first dissolved in Milli-Q water (95 μL) by stirring for ∼2 min, and then 250 μL of 100 mM AgNO3 was added and homogenized by stirring for ∼2 min. Next, 5 μL of formaldehyde solution (35 wt %) was added under vigorous stirring, and the mixture was heated at 70 °C for 10 min. The final solution was stocked at ambient environment for at least 48 h before its further application. Without any centrifugation or purification, the PEI-capped Ag nanoclusters were diluted 10-fold in Milli-Q water, and the diluted Ag nanoclusters were applied as the probe to study the fluorescence quenching effects of halide ions in this assay. Apparatus and more detailed experimental protocols were given in the Supporting Information section.



RESULTS AND DISCUSSION Synthesis and Characterization of PEI-Capped Ag Nanoclusters. The water-soluble and blue-emitting Ag nanoclusters were synthesized on the basis of a PEI-modified 10374

dx.doi.org/10.1021/ac3024526 | Anal. Chem. 2012, 84, 10373−10379

Analytical Chemistry

Article

(Figure S7a in the Supporting Information), all experiments were performed under dark condition. Selectivity of the Sensing System. To investigate whether this sensing system is specific for halide ions, the effects of 14 other kinds of anions, including benzoate, citrate, PO43−, B4O72−, SO42−, CO32−, SiO32−, ClO4−, CH3COO−, NO2−, NO3−, OH−, F−, S2−, and 13 common cations, such as Na+, K+, Mg2+, Zn2+, Ca2+, Bi2+, Pb2+, Co2+, Sr2+, Ce2+, Cu2+, Al3+, and Hg2+, on the fluorescence response of PEI-capped Ag nanoclusters were investigated under the optimum conditions (Figure 2). Most anions could not lead to any significant

can be known that the concentrations of the probe determine the fluorescence intensities but which become relatively insensitive when the concentrations of the diluted PEI-capped Ag nanoclusters are higher than 50 μL mL−1. On the other hand, the sensitivity of the detection of halide ions also depends on the concentration of probe. Generally, in the presence of quencher of a given concentration, the lower concentration of fluorescence probe, the more efficient is the fluorescence quenching and thus the higher the sensitivity found,44 which is illustrated clearly in Figure S3 in the Supporting Information. However, the linear range of the detection would be narrow when too low of a concentration of fluorescence probe was used. Finally, the concentration of 10 μL mL−1 PEI-capped Ag nanoclusters was eventually chosen in the detection of halide ions for obtaining the widest linear response and the lowest detection limit. The response rate of the fluorescence signal of the PEIcapped Ag nanoclusters upon addition of different halide ions were monitored subsequently, as shown in Figure S4 in the Supporting Information. After 30 min, the fluorescence intensities are quenched by 51%, 53%, and 42% upon addition of 50 μM Cl−, 10 μM Br−, and 5 μM I−, respectively, and remain constant with time, which indicates that the reactions are complete within 30 min. Another key factor is the pH value of the sensing system, because both of the initial fluorescence intensity (in the absence of halide ions) and the fluorescence quenching (in the presence of halide ions) of the PEI-capped Ag nanoclusters are pH dependent (Figure S5 in the Supporting Information). The pH values of the solutions were controlled by the addition of NaOH or HNO3 in the experiments in order to avoid the electrolyte-induced perturbations by buffer solutions. The charge distribution of different amino groups of PEI is strongly dependent on the variation of pH,45,46 which imparts an intrinsic characteristic to PEI-capped Ag nanoclusters as a pH sensor. The PEI-capped Ag nanoclusters exhibit lower fluorescence intensities in acid medium but higher fluorescence activities in neutral and basic medium. However, the fluorescence-quenched efficiencies are higher at pH 7.0−8.0 than that at pH > 8.0, which may result from two aspects: on the one hand, OH− could increase competition with halide ions for Ag nanoclusters in alkaline solutions; on the other hand, the halides induced oxidative etching of Ag nanoclusters can be enhanced by the introduction of a protonic acid.47 The quenching efficiencies keep similarly high values for Cl−, Br−, and I− at pH 7.0−8.0, indicating that a neutral media would ensure the sensitive detection of halide ions. The influence of temperature was then investigated. As shown in Figure S6 in the Supporting Information, the fluorescence intensity of the PEI-capped Ag nanoclusters keep stable at a temperature below 40 °C. At a temperature >20 °C, the response rates of the fluorescence signal upon addition of halide ions are almost not affected practically by temperature, so room temperature could be chosen for the effective recognition of halide ions. Owing to its prominent photosensitivity of silver halides, the influence of light on the reaction was also studied (Figure S7 in the Supporting Information). The fluorescence-quenched efficiencies by halide ions are almost not influenced by light, so there is no specific limitation of the reaction and detection under light or dark conditions. However, considering a slight instability of diluted PEI-capped Ag nanoclusters under light

Figure 2. Selectivity of PEI-capped Ag nanoclusters for the detection of halide ions over other ions: the concentrations of Cl−, Br−, and I− are 50, 10, and 5 μM, respectively; the concentration of each anion is 4 mM, except that NO2−, PO43−, and SO42− are 0.4 mM, and S2− is 10 μM; the concentration of each cation is 1 mM, except for 0.5 mM Cu2+ and 2 μM Hg2+.

fluorescence decrease of Ag nanoclusters due to their relatively larger solubility21,22 (Table S1 in the Supporting Information) than that of silver halides in water (the Ksp of AgCl, AgBr, and AgI was 1.76 × 10−10, 5.32 × 10−13, and 8.49 × 10−17, respectively), suggesting that the interaction between the silver atoms and the anions plays a dominant role in the fluorescence quenching. Thus, the recognition would be greatly interfered by S2−, which has a strong interaction tendency in forming silver compound and the Ksp of Ag2S is 1.6 × 10−49. The linear response range of detection of S2− is given in Figure S8 in the Supporting Information, implying that this sensing system could also be used to detect S2− in the absence of halide ions. Meanwhile, the fluorescence of PEI-capped Ag nanoclusters was little affected by most of metallic ions, except for Hg2+. It is known that Ag nanoclusters could be used as a probe for Hg2+ based on the 5d10 (Hg2+) − 4d10 (Ag+) metallophilic interaction-induced fluorescence quenching of Ag clusters.27 The linearity of PEI-capped Ag nanoclusters−Hg2+ system is shown in Figure S9 in the Supporting Information. Except for the interferences of S2− and Hg2+, this sensing system manifests a remarkably high selectivity toward halide ions over most of anions and cations, where it is superior to other halide nanosensors, such as Ag nanoprism can not discriminate chloride and phosphate,21 and Cu@Au nanoparticles are greatly interfered by all of the metallic ions when their concentration reached 1 mM in the recognition of iodide.19 Sensitivity of the Sensing System. On the basis of the optimized conditions discussed above, the linearity and the detection limit of this sensing system were evaluated by varying the concentration of different halide ions. In Figure 3, the fluorescence of PEI-capped Ag nanoclusters was sensitive and decreased proportionately with increasing concentrations of halide ions as noted from the relationship between the 10375

dx.doi.org/10.1021/ac3024526 | Anal. Chem. 2012, 84, 10373−10379

Analytical Chemistry

Article

Figure 3. Fluorescence emission spectra (a) and the fluorescence-quenched efficiencies (ΔI/I0) (b) of PEI-capped Ag nanoclusters upon addition of different concentrations of halide ions (a and b for Cl−; c and d for Br−; e and f for I−). The inset of b displays the two linear relationships between ΔI/I0 and the concentrations of Cl− from 0.5 to 25 μM and from 25 to 80 μM, respectively; the insets of d and f show the linear range for 0.1−14 μM Br− and 0.05−6 μM I−, respectively.

Figure 4. UV−vis spectra, the absorbance changes at 454 nm (ΔA), and photographs for recording the color changes of PEI-capped Ag nanoclusters upon addition of different concentrations of halide ions (a, b, c, and d for Cl−; e, f, g, and h for Br−; i, j, k, and l for I−; c, g, and k under visible light; d, h, and l under UV light). The insets of b, f, and j show the linear range for 50−400 μM Cl−, 10−300 μM Br−, and 10−200 μM I−, respectively.

I−, at a signal-to-noise ratio of 3, were estimated to be 200, 65, and 40 nM, respectively. Thus, the sensitivity of detection is dependent on the identity of the halide ion and follows the ordering, Cl−< Br−< I−. In addition, we compared the linear ranges and LOD with other nanoscaled halide sensors (e.g.,

fluorescence-quenched efficiencies and the concentrations of Cl−, Br−, and I−. Good linear correlations were found over the concentration range from 0.5 to 80 μM for Cl−, 0.1 to 14 μM for Br−, and 0.05 to 6 μM for I−, respectively (see the inset of Figure 3b,d,f). The limits of detection (LOD) for Cl−, Br−, and 10376

dx.doi.org/10.1021/ac3024526 | Anal. Chem. 2012, 84, 10373−10379

Analytical Chemistry

Article

this range of 400 to 600 nm exhibits an obvious red-shift by comparison with the typical surface plasmon resonance peak of Ag nanoparticles (∼400 nm).32−36 We attributed the observed red shift of the absorption peak to both of the formation of larger particles and the existence of AgX layer (X− = Cl−, Br−, and I−) on the surface of silver clusters.35,36,48,49 Correspondingly, the color of the solutions changed from nearly colorless to yellow for Cl−, to brown for Br−, and to dark brown for I− and deepened with increasing concentration of halide ions. Consequently, the absorbance changes at 454 nm (ΔA) were used as a function of halide concentrations and exhibited good linear relationships for 50−400 μM Cl− (LOD 20 μM), for 10− 300 μM Br− (LOD 5 μM), and for 10−200 μm I− (LOD 5 μM), respectively. Although the sensitivity of colorimetric detection of halide ions is lower than that of fluorescence response, the color reactions of PEI-capped Ag nanoclusters upon addition of halide ions are easy and rapid: an obvious color change would be observed less than 10 min for 1 mM Cl−, 100 μM Br−, and 60 μM I−, respectively. Thus, PEI-capped Ag nanoclusters serving as a colorimetric sensor for the recognition of halides are still attractive and promising because of their simplicity, rapidity, reliability, and low cost. Mechanism for the Recognition of Halide Ions. The sensitive and selective recognition of halide ions by PEI-capped Ag nanoclusters are based on the unique reactions between halide ions and the silver atoms. Besides the Ksp of different silver halides, the sensitivities of fluorescence and colorimetric detection of halide ions are also considered to be associated with the halide ions induced oxidative etching and aggregation of Ag nanoclusters. Previous studies have shown the interactions of halide ions with silver or other nanoparticles.19,20,35,36,50 Henglein35,36 explained systematically about the potential of halide ions to catalyze oxidative corrosion of the silver particles. Experiments by Zhang et al.19,20 later provided the mechanism of halides-induced aggregating/fusion, fragmentation, and Ostwald ripening process of Cu@Ag nanoparticles or Cu@Au nanoparticles. Recently, Espinoza et al.50 explored the kinetics of halide-induced decomposition and aggregation of silver nanoparticles. In this work, dynamic light scattering (DLS) measurements, including Zeta (ζ)-potential and hydrodynamic diameter (HDD) determinations, were performed to elucidate the recognition mechanism (Figure S17 in the Supporting Information). The PEI-capped Ag nanoclusters possess positive charges at pH 7.5 due to protonation of partial primary amines of PEI.45,46 Besides, the apparent HDD of PEI-capped Ag nanoclusters (∼3.5 nm) is larger than the TEM diameter because DLS gives the size of solvent-swollen aggregates and TEM gives the size of dry particles.51 However, the ζ-potential decreased, and in the meantime, the HDD increased gradually with increasing concentration of halide ions. On the basis of these previous illuminating studies mentioned above and our experimental data, a mechanism that halide-induced oxidative etching and aggregation leading to the fluorescence quenching of Ag nanoclusters is proposed, as represented in Scheme 1. In the first step, halide ions, as the nucleophile reagents, adsorbed on the surface of Ag nanoclusters spontaneously due to the high affinity of Ag−halide bond or the lower Ksp of silver halide compounds. The disturbances at the surface Ag atom are caused by precomplexation or preoxidization which leads to catalyzed etching in the presence of air and the formation of a monolayer of AgX (X− = Cl−, Br−, and I−) via reaction 1.35,50,52

nanoparticles and quantum dots), which were reported by the literature in recent years,15−21 and our method shows better linear ranges and lower detection limits than those above (Tale S2 in the Supporting Information). In the meanwhile, the linear range and LOD of this sensing system are also not inferior to the traditional ion chromatography,6 spectrometry by artificial chromophore and fluorophore,9 and electrochemical analysis.11−14 Selective Detection of Br− and I− in Coexistence with Cl− under Condition of Higher Ionic Strength. In the study of the influence of ionic strength (Figure S10 in the Supporting Information), we found that the sensitivity of Cl− detection decreased dramatically with increasing ionic strength, and the fluorescence-quenched efficiencies were less than 5% upon addition of 50 μM Cl− as the concentration of KNO3 was higher than 10 mM. Meanwhile, the sensitivities of recognitions of Br− and I− seemed to be not affected significantly by ionic strength. Thus, the high concentrations of KNO3 could be used to distinguish I− and Br− separately in coexistence with Cl−. As shown in Figures S11 and S12 in the Supporting Information, the quantification of 5 μM I− would not be disturbed in coexistence with 50 μM Cl− under the condition of 10 mM KNO3, as well as the detection of 10 μM Br− in coexistence with 50 μM Cl− under the condition of 20 mM KNO3. Hence, 10 and 20 mM KNO3 solutions were chosen to selective recognition of I− and Br− in coexistence with Cl−, respectively. The stability of the detection of I− and Br− in coexistence with Cl− under high ionic strength was also studied (Figure S13 in the Supporting Information). After incubation of 30 min, the fluorescence-quenched efficiencies upon addition of I− and Br− remain approximately constant with time, indicating that the reactions are complete within 30 min. Besides, the tolerance concentrations of Cl− were also investigated. Figure S14 in the Supporting Information illustrates that the maximum tolerance concentrations of Cl− are 20-fold higher than that of I− and 10fold higher than that of Br−, respectively. Consequently, the selective detection of I− and Br− in coexistence with Cl− was established under the condition of higher ionic strength (Figures S15 and S16 in the Supporting Information). The relative fluorescence-quenched efficiencies exhibited good linear relationships with the concentrations of I− (0.5−6 μM; LOD 90 nM) in 10 mM KNO3 and with the concentrations of Br− (1−14 μM; LOD 300 nM) in 20 mM KNO3. In comparison with the linear ranges and LOD of the detection of I− and Br− in the absence of Cl− and salt solutions, the sensitivities of the detection declined in the presence of Cl− and high ionic strength, but such highly specific discriminations of I− and Br− in coexistence with Cl− by simple Ag nanoclusters in high ionic strength are still exhilarating and unexpected, which is a significant advantage in the detection of I− and Br− in real samples. UV−Vis Spectroscopic Investigation and Colorimetric Detection of Halide Ions. The PEI-capped Ag nanoclusters would also display a distinct color variation and obvious changes in the UV−vis spectra upon addition of halide ions with increasing concentration of probe. The concentrationdependent optical spectra of the PEI-capped Ag nanoclusters upon addition of Cl−, Br−, and I− are shown in Figure 4, in which an absorption band in the range of 400−600 nm increases gradually with increasing concentration of halide ions. However, the addition of Br− and, particularly, I− differs distinctly from that of Cl− in having a significant optical absorption in this region. Meanwhile, the absorption band in 10377

dx.doi.org/10.1021/ac3024526 | Anal. Chem. 2012, 84, 10373−10379

Analytical Chemistry

Article

of a barrier.50 Thus, although the concentration of Cl− added is 10-fold higher than that of I− and Br−, the HDD of aggregated particles (Agm−AgCl) is still smaller than that of aggregated Agm−AgI and Agm−AgBr particles, and the absorbance at 400− 600 nm upon addition of Br− or I− increases much more than that upon addition of Cl−. At higher halide concentrations, aggregation is the dominant process. It should be noted that the pathways of oxidative etching and aggregation are not isolated in a given concentration range, but one of them would dominate in one concentration. Detection of Chloride in Real Samples. The detection of chloride would be disturbed by Br−, I−, S2−, and high ionic strength, so some pretreatment methods should be introduced for real sample determination (e.g., addition of masking agents or other methods of separation and purification). However, fortunately, such interference could not be serious under certain circumstances when the concentrations of these anions are very low, for example, in tap water or mineral water. Thus, this sensing system for the detection of Cl− was further applied to the real water samples. Especially, quantitative analysis of Cl− in tap water or mineral water is very helpful for water quality monitoring. Here, tap water for three consecutive days and two mineral water samples were analyzed by PEI-capped Ag nanoclusters for the quantification of Cl− with the fluorescence method. After 10-fold dilution and adjusting pH at 7.5, the samples were directly measured without other pretreatment. Typical detection results are given in Tables S3 and S4 in the Supporting Information. The average concentrations of Cl− are 353.87 μM (12.56 mg L−1) in tap water and 48.65 μM (1.73 mg L−1) in mineral water, which conform to the national standards of drinking water in China (GB5749-2006) that the concentration of chloride should be not more than 250.0 mg L−1.

Scheme 1. Mechanism Scheme for the Recognition of Halide Ions by Ag Nanoclusters

Ag n − 1 − Ag + X− +

1 1 O2 + H 2O 4 2

→ Ag δn−− 1 − Ag δ +(X) + OH−

(1)

The slopes of linear regression equations of different halide ions could be interpreted as the rate of the oxidative etching (Figure S18 in the Supporting Information), which increases as the standard electrochemical potential (SHE) of the various silver halides becomes more negative: AgCl/Ag pair is 0.22 V (vs SHE), AgBr/Ag pair is 0.071 V (vs SHE), and AgI/Ag pair is −0.15 V (vs SHE). Besides, the rate of diffusion of reactants (O2, electrons, Ag+, and X−) also decides the rate of oxidative etching process. AgCl layer is more uniform and denser, which could prevent oxygen from reaching the surface of Ag nanoclusters,53 and the diffusion of Ag+ is substantially faster through AgI than AgBr or AgCl.54 It means that the layer of AgBr produces a more significant barrier to reaction 1 than does AgI but is not as effective as AgCl. Thus, the rate of decay upon addition of I− is much faster than that of Cl− and Br−. At lower halide concentrations, oxidative etching is likely the dominant mechanism. In this step, the particles are the halidecoated silver particles (Agm−AgX), rather than separated silver nanoclusters or nanoparticles and silver halides. In addition, the chemisorbed halide ions neutralize the surface charge of PEI-capped Ag nanoclusters and, as a result, increase the van der Waals attractive forces among these halidecoated silver particles, which would also lead to the onset of aggregation. Such halide induced ζ-potential reduction has been reported by various research.19,20,55 In the meanwhile, the rate of oxidative etching also decreases somewhat before the aggregation occurs. The lower ζ-potential triggers the aggregation process, which must accompany the changes of UV−vis spectra and the color of solutions. It is clearly seen that the absorbance is highly dependent on the quantity of halide ions added. In the second step, the obvious color change of the solution indicates that aggregation becomes the dominant mechanism at the higher concentrations. Owing to the stability of AgCl layer mentioned above, the passivating surface would thicken with increasing concentration of Cl−, which hinders the degree of aggregation. I− or Br− also seems to form a passivating layer, but it is much less effective than that formed by Cl−; especially, that AgI layer is more porous, presenting less



CONCLUSION In this paper, PEI-capped Ag nanoclusters, as a new fluorescent and colorimetric platform, are used for the sensitive and selective detection of halide ions. This sensing system exhibits a remarkably high selectivity toward halide ions over most anions and cations and shows better linear ranges and lower detection limits than other halide nanosensors. Especially, high disturbance−tolerant recognitions of iodide and bromide in coexistence with chloride could be achieved under the condition of high ionic strength. Besides, the recognition of halide could also be performed by the colorimetric method, which is also attractive and promising. The recognition mechanism is based on the unique reactions between halide ions and the silver atoms. In particular, halide-induced oxidative etching and aggregation would produce the fluorescence quenching of Ag nanoclusters. Furthermore, the present nanosensor has been applied successfully to the detection of Cl− in real water samples. All of the substantial advantages of this sensing system have great potential for the reliable detection of halide ions in various fields, such as water, soil, and biological samples and so on.



ASSOCIATED CONTENT

* Supporting Information S

Figures S1−S18 and Tables S1−S4. This material is available free of charge via the Internet at http://pubs.acs.org. 10378

dx.doi.org/10.1021/ac3024526 | Anal. Chem. 2012, 84, 10373−10379

Analytical Chemistry



Article

(29) Godbey, W. T.; Wu, K. K.; Hirasaki, G. J.; Mikos, A. G. Gene Ther. 1996, 6, 1380−1388. (30) Liu, Z. Z.; Zheng, M.; Meng, F. H.; Zhong, Z. Y. Biomaterials 2011, 32, 9109−9119. (31) Kim, H.; Namgung, R.; Singha, K.; Oh, I. K.; Kim, W. J. Bioconjugate Chem. 2011, 22, 2558−2567. (32) Ershov, B. G.; Janata, E.; Henglein, A.; Fojtik, A. J. Phys. Chem. 1993, 97, 4589−4594. (33) Henglein, A.; Mulvaney, P.; Linnert, T. Faraday Discuss. 1991, 92, 31−44. (34) Ershov, B. G.; Henglein, A. J. Phys. Chem. B 1998, 102, 10663− 10666. (35) Linnert, T.; Mulvaney, P.; Henglein, A.; Weller, H. J. Am. Chem. Soc. 1990, 112, 4657−4664. (36) Henglein, A. J. Phys. Chem. 1993, 97, 547−5471. (37) Fletcher., A. N. Photochem. Photobiol. 1969, 9, 439−444. (38) A Guide to Recording Fluorescence Quantum Yields; HORIBA Jobin Yvon Inc.: Middlesex, United Kingdom; http://www.jobinyvon. com/usadivisions/fluorescence/applications/quantumyieldstrad.pdf (accessed August 30, 2012). (39) Grabolle, M.; Spieles, M.; Lesnyak, V.; Gaponik, N.; Eychmüller, A.; Resch-Genger, U. Anal. Chem. 2009, 81, 6285−6294. (40) Rao, T. U. B.; Pradeep, T. Angew. Chem., Int. Ed. 2010, 49, 3925−3929. (41) Rao, T. U. B.; Nataraju, B.; Pradeep, T. J. Am. Chem. Soc. 2010, 132, 16304−16307. (42) Bootharaju, M. S.; Pradeep, T. Langmuir 2011, 27, 8134−8143. (43) Han, B. Y.; Yuan, J. P.; Wang, E. K. Anal. Chem. 2009, 81, 5569−5573. (44) Dong, Y. Q.; Wang, R. X.; Li, G. L.; Chen, C. Q.; Chi, Y. W.; Chen., G. L. Anal. Chem. 2012, 84, 6220−6224. (45) Nagaya, J.; Homma, M.; Tanioka, A.; Minakata, A. Biophys. Chem. 1996, 60, 45−51. (46) Borkovec, M.; Koper, G. J. M. Macromolecules 1997, 30, 2151− 2158. (47) Im, S. H.; Lee, Y. T.; Wiley, B.; Xia, Y. Angew. Chem. 2005, 117, 2192−2195. (48) Yin, Y. D.; Li, Z. Y.; Zhong, Z. Y.; Gates, B.; Xia, Y. N.; Venkateswaranc, S. J. Mater. Chem. 2002, 12, 522−527. (49) Li, X.; Lenhart, J. J.; Walker, H. W. Langmuir 2010, 26, 16690− 16698. (50) Espinoza, M. G.; Hinks, M. L.; Mendoza, A. M.; Pullman, D. P.; Peterson, K. I. J. Phys. Chem. C 2012, 116, 8305−8313. (51) Song, T.; Dai, S.; Tam, K. C.; Lee, S. Y.; Goh, S. H. Langmuir 2003, 19, 4798−4803. (52) Kapoor, S. Langmuir 1998, 14, 1021−1025. (53) Iski, E. V.; El-Kouedi, M.; Calderon, C.; Wang, F.; Bellisario, D. O.; Ye, T.; Sykes, E. C. H. Electrochim. Acta 2011, 56, 1652−1661. (54) Berry, C. R. J. Opt. Soc. Am. 1950, 40, 615−617. (55) Cheng, W. L.; Dong, S. J.; Wang, E. K. Angew. Chem. 2003, 115, 465−468.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (N.B.L.); [email protected] (H.Q.L.). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (No. 21273174, 20975083) and the Municipal Science Foundation of Chongqing City (No. CSTC−2008BB 4013).



REFERENCES

(1) Cohen, T. S.; Prince, A. Nat. Med. 2012, 18, 509−519. (2) Nitu, M.; Montgomery, G.; Eigen, H. Pediatr. Rev. 2011, 32, 240−251. (3) Azizi, F.; Hedayati, M.; Rahmani, M.; Sheikholeslam, R.; Allahverdian, S.; Salarkia, N. J Endocrinol Invest. 2005, 28, 23−29. (4) Leeuwen, F. X. R.; Sangster, B; Hildebrandt, A. G. Crit. Rev. Toxicol. 1987, 18, 189−213. (5) Villagrán, C.; Deetlefs, M.; Pitner, W. R.; Hardacre, C. Anal. Chem. 2004, 76, 2118−2123. (6) Hao, F. P.; Haddad, P. R.; Ruther, T. Chromatographia 2008, 67, 495−498. (7) Malongo, T. K.; Patris, S.; Macours, P.; Cotton, F.; Nsangu, J.; Kauffmann, J. Talanta 2008, 76, 540−547. (8) Gilday, L. C.; White, N. G.; Beer, P. D. Dalton Trans. 2012, 41, 7092−7097. (9) Li, E. Y.; Lin, L.; Wang, L.; Pei, M. S.; Xu, J. K.; Zhang, G. Y. Macromol. Chem. Phys. 2012, 213, 887−892. (10) Caballero, A.; Zapata, F.; White, N. G.; Costa, P. J.; Félix, V.; Bee, P. D. Angew. Chem. 2012, 124, 1912−1916. (11) Trnkova, L.; Adam, V.; Hubalek, J.; Babula, P.; Kizek, R. Sensors 2008, 8, 5619−5636. (12) Zahran, E. M.; Hua, Y.; Lee, S.; Flood, A. H.; Bachas, L. G. Anal. Chem. 2011, 83, 3455−3461. (13) Chiu, M. H.; Cheng, W. L.; Muthuraman, G.; Hsu, C. T.; Chung, H. H.; Zen, J. M. Biosens. Bioelectron. 2009, 24, 3008−3013. (14) Qin, X.; Wang, H. C.; Miao, Z. Y.; Wang, X. S.; Fang, Y. X.; Chen, Q.; Shao, X. G. Talanta 2011, 84, 673−678. (15) Kumar, A.; Chhatra, R. K.; Pandey, P. S. Org. Lett. 2010, 12, 24− 27. (16) Graefe, A.; Stanca, S. E.; Nietzsche, S.; Kubicova, L.; Beckert, R.; Biskup, C.; Mohr, G. J. Anal. Chem. 2008, 80, 6526−6531. (17) Ruedas-Rama, M. J.; Orte, A.; Hall, E. A. H.; Alvarez-Pez, J. M.; Talavera, E. M. Analyst 2012, 137, 1500−1508. (18) Vivek, J. P.; Burgess, I. J. J. Phys. Chem. C 2008, 112, 2872− 2880. (19) Zhang, J.; Xu, X. W.; Yang, C.; Yang, F.; Yang, X. R. Anal. Chem. 2011, 83, 3911−3917. (20) Zhang, J.; Yuan, Y.; Xu, X. W.; Wang, X. L.; Yang, X. R. ACS Appl. Mater. Interfaces 2011, 3, 4092−4100. (21) Jiang, X. C.; Yu, A. B. Langmuir 2008, 24, 4300−4309. (22) Jin, J. Y.; Ouyang, X. Y.; Li, J. S.; Jiang, J. H; Wang, H.; Wang, Y. X.; Yang, R. H. Analyst 2011, 136, 3629−3634. (23) Empedocles, S.; Bawendi, M. Acc. Chem. Res. 1999, 32, 389− 396. (24) Schaaff, T. G.; Knight, G.; Shafigullinet, M. N.; Borkman, R. F.; Whetten, R. L. J. Phys. Chem. B 1998, 102, 10643−10646. (25) Lee, T. H.; Gonzalez, J. I.; Zheng, J.; Dickson, R. M. Acc. Chem. Res. 2005, 38, 534−541. (26) Zhou, T. Y.; Rong, M. C.; Cai, Z. M.; Yang, C. J.; Chen, X. Nanoscale 2012, 4, 4103−4106. (27) Guo, C. L.; Irudayaraj, J. Anal. Chem. 2011, 83, 2883−2889. (28) Chen, W. Y.; Lan, G. Y.; Chang, H. T. Anal. Chem. 2011, 83, 9450−9455. 10379

dx.doi.org/10.1021/ac3024526 | Anal. Chem. 2012, 84, 10373−10379