Colorimetric Sensing of Silver (I) and Mercury (II) Ions Based on an

Jul 16, 2010 - We have developed a rapid and homogeneous method for the highly selective detection of Hg2+ and Ag+ using Tween 20-modified gold ...
1 downloads 0 Views 2MB Size
Download PDF
Anal. Chem. 2010, 82, 6830–6837

Colorimetric Sensing of Silver(I) and Mercury(II) Ions Based on an Assembly of Tween 20-Stabilized Gold Nanoparticles Cheng-Yan Lin,† Cheng-Ju Yu,† Yen-Hsiu Lin,† and Wei-Lung Tseng*,†,‡

Anal. Chem. 2010.82:6830-6837. Downloaded from pubs.acs.org by UNIV OF EDINBURGH on 01/29/19. For personal use only.

Department of Chemistry, National Sun Yat-sen University, Taiwan, and National Sun Yat-sen University-Kaohsiung Medical University Joint Research Center, Kaohsiung, Taiwan We have developed a rapid and homogeneous method for the highly selective detection of Hg2+ and Ag+ using Tween 20-modified gold nanoparticles (AuNPs). Citrate ions were found to still be adsorbed on the Au surface when citrate-capped AuNPs were modified with Tween 20, which stabilizes the citrate-capped AuNPs against conditions of high ionic strength. When citrate ions had reduced Hg2+ and Ag+ to form Hg-Au alloys and Ag on the surface of the AuNPs, Tween 20 was removed from the NP surface. As a result, the AuNPs were unstable under a high-ionic-strength solution, resulting in NP aggregation. The formation of Hg-Au alloys or Ag on the surface of the AuNPs was demonstrated by means of inductively coupled plasma mass spectroscopy and energy-dispersive X-ray spectroscopy. Tween 20-AuNPs could selectively detect Hg2+ and Ag+ at concentrations as low as 0.1 and 0.1 µM in the presence of NaCl and EDTA, respectively. Moreover, the probe enables the analysis of AgNPs with a minimum detectable concentration that corresponds to 1 pM. This probe was successfully applied to detect Hg2+ in drinking water and seawater, Ag+ in drinking water, and AgNPs in drinking water. Interest in monitoring toxic metal ions in aquatic ecosystems continues because these contaminants adversely affect the environment and have serious medical effects.1 Silver and mercury are two of the most hazardous metal pollutants, and they are widely distributed in ambient air, water, soil, and even food.2,3 For example, silver can inactivate sulfhydryl enzymes and accumulate in the body,4 and mercury exposure can damage a variety of organs and the immune system.5 Current approaches to detecting these two metal ions include inductively coupled * To whom correspondence should be addressed. Fax: 011-886-7-3684046. E-mail: [email protected]. † Department of Chemistry, National Sun Yat-sen University. ‡ National Sun Yat-sen University-Kaohsiung Medical University Joint Research Center. (1) Campbell, L.; Dixon, D. G.; Hecky, R. E. J. Toxicol. Environ. Health, Part B 2003, 6, 325–356. (2) Wood, C. M.; McDonald, M. D.; Walker, P.; Grosell, M.; Barimo, J. F.; Playle, R. C.; Walsh, P. J. Aquat. Toxicol. 2004, 70, 137–157. (3) Boening, D. W. Chemosphere. 2000, 40, 1335–1351. (4) Ratte, H. T. Environ. Toxicol. Chem. 1999, 18, 89–108. (5) Holmes, P.; James, K. A.; Levy, L. S. Sci. Total Environ. 2009, 408, 171– 182.

6830

Analytical Chemistry, Vol. 82, No. 16, August 15, 2010

plasma mass spectrometry (ICP-MS),6,7 atomic absorption spectrometry,8,9 and stripping voltammetry.10,11 Although these methods offer excellent sensitivity and multielement analysis, they are rather costly, time-consuming, complex, and nonportable. In response to these shortcomings, various sensors using small organic molecules,12,13 oligonucleotides,14,15 DNAzymes,16,17 and semiconductor quantum dots18,19 have been investigated for the selective detection of Ag+ or Hg2+ in aqueous solutions. Unfortunately, most of these methods suffer from low water solubility, a complex synthesis procedure, and time-consuming DNA probe preparation. Recently, gold nanoparticles (AuNPs) have become another emerging material for sensing Hg2+ or Ag+, because they have a high extinction coefficient in the visible region and behavior that depends on the interparticle distance. When the distances between the AuNPs become less than the average particle diameter, the color of the AuNPs changes from red to purple. Because of the coordination between the carboxyl groups of thiols and Hg2+, thiol-capped AuNPs have been used for colorimetric sensing of Hg2+.20-22 Also, Hg2+ can selectively coordinate thymine (T) bases and forms stable T-Hg2+-T complexes. The melting temperature of cDNA containing T-Hg2+-T complexes is higher than that (6) Karunasagar, D.; Arunachalam, J.; Gangadharan, S. J. Anal. At. Spectrom. 1998, 13, 679–682. (7) Barriada, J. L.; Tappin, A. D.; Evans, E. H.; Achterberg, E. P. TrAC, Trends Anal. Chem. 2007, 26, 809–817. (8) Li, Y.; Chen, C.; Li, B.; Sun, J.; Wang, J.; Gao, Y.; Zhao, Y.; Chai, Z. J. Anal. At. Spectrom. 2006, 21, 94–96. (9) Chamsaz, M.; Arbab-Zavar, M. H.; Akhondzadeh, J. Anal. Sci. 2008, 24, 799–801. (10) Kim, H. J.; Park, D. S.; Hyun, M. H.; Shim, Y. B. Electroanalysis 1998, 10, 303–306. (11) Mikelova, R.; Baloun, J.; Petrlova, J.; Adam, V.; Havel, L.; Petrek, J.; Horna, A.; Kizek, R. Bioelectrochemistry 2007, 70, 508–518. (12) Chatterjee, A.; Santra, M.; Won, N.; Kim, S.; Kim, J. K.; Kim, S. B.; Ahn, K. H. J. Am. Chem. Soc. 2009, 131, 2040–2041. (13) Zhan, X. Q.; Qian, Z. H.; Zheng, H.; Su, B. Y.; Lan, Z.; Xu, J. G. Chem. Commun. 2008, 1859–1861. (14) Lin, Y.-H.; Tseng, W.-L. Chem. Commun. 2009, 6619–6621. (15) Wang, J.; Liu, B. Chem. Commun. 2008, 4759–4761. (16) Li, T.; Shi, L.; Wang, E.; Dong, S. Chemistry 2009, 15, 3347–3350. (17) Hollenstein, M.; Hipolito, C.; Lam, C.; Dietrich, D.; Perrin, D. M. Angew. Chem., Int. Ed. 2008, 47, 4346–4350. (18) Koneswaran, M.; Narayanaswamy, R. Sens. Actuators, B 2009, 139, 91– 96. (19) Chen, J.-L.; Zhu, C.-Q. Anal. Chim. Acta 2005, 546, 147–153. (20) Huang, C.-C.; Chang, H.-T. Anal. Chem. 2006, 78, 8332–8338. (21) Yu, C.-J.; Tseng, W.-L. Langmuir 2008, 24, 12717–12722. (22) Darbha, G. K.; Singh, A. K.; Rai, U. S.; Yu, E.; Yu, H.; Chandra Ray, P. J. Am. Chem. Soc. 2008, 130, 8038–8043. 10.1021/ac1007909  2010 American Chemical Society Published on Web 07/16/2010

containing T-T mismatches.23 Based on this concept, two strands of DNA, which are designed to be complementary except for a single T-T mismatch, are used to modify the surface of the AuNPs. The resulting two types of DNAfunctionalized AuNPs are selectively aggregated in the presence of Hg2+ based on T-Hg2+-T coordination and temperature control.24 Similarly, Hg2+ was selectively detected using two types of T-rich DNA-modified AuNPs and a T-rich DNA linker at room temperature.25 Citrate-capped AuNPs interacting with single strands of T-rich oligonucleotide were found to be stable in a high-salt solution.26-28 When Hg2+ causes the conformation of T-rich DNA into a folded structure, this folded DNA cannot be adsorbed onto the AuNP surface. Salt-induced NP aggregation occurs because T-rich DNA is removed. In addition, it is well-known that a Au surface exhibits a strong affinity for Hg2+.29-32 Thus, after the reduction of Hg2+ with NaBH4, the Hg(0) thus generated is strongly bonded onto the surface of Au-based nanomaterials to form a solid amalgam-like structure. The surface plasmon resonance (SPR) band of Au nanorods and NPs in an excess of NaBH4 has been found to undergo a blue shift and a decrease in intensity after Hg2+ was added.29,31 The only work on the detection of Ag+ reported that AuNPs functionalized with cytosine-(C)-rich oligonucleotide selectively aggregated in the presence of Ag+ based on the formation of C-Ag+-C complexes.33 Although these methods all show good sensitivity and selectivity to Hg2+ or Ag+, analysis of these two metal ions using a single type of AuNPs remains a challenge. In this study, we present a label-free, rapid, and homogeneous method for sensing both Hg2+ and Ag+ using Tween 20stabilized AuNPs (Tween 20-AuNPs). Because the surfaces of Tween 20-AuNPs still had citrate ions, the reduction of Hg2+ or Ag+ with citrate resulted in the formation of Hg-Au alloy and Ag on the Au surface. When the Tween 20 was removed, NP aggregation occurred. We also investigated the effect of masking agents on the selectivity of this probe. To demonstrate its practicality, the present method was further applied to the determination of Hg2+, Ag+, and AgNPs in complex matrices. EXPERIMENTAL SECTION Chemicals. Hydrogen tetrachloroaurate (III) dehydrate, Na2HPO4, and Na3PO4 were purchased from Alfa Aesar (Ward Hill, MA). Trisodium citrate, ethylenediaminetetraacetic acid (EDTA), Tween 20, Tween 40, Tween 60, Tween 80, NaBH4, (23) Tanaka, Y.; Oda, S.; Yamaguchi, H.; Kondo, Y.; Kojima, C.; Ono, A. J. Am. Chem. Soc. 2007, 129, 244–245. (24) Lee, J.-S.; Han, M.-S.; Mirkin, C. A. Angew. Chem., Int. Ed. 2007, 46, 4093– 4096. (25) Xue, X.; Wang, F.; Liu, X. J. Am. Chem. Soc. 2008, 130, 3244–3245. (26) Yu, C.-J.; Cheng, T.-L.; Tseng, W.-L. Biosens. Bioelectron. 2009, 25, 204– 210. (27) Li, D.; Wieckowska, A.; Willner, I. Angew. Chem., Int. Ed. 2008, 47, 3927– 3931. (28) Liu, C.-W.; Hsieh, Y.-T.; Huang, C.-C.; Lin, Z.-H.; Chang, H.-T. Chem. Commun. 2008, 2242–2244. (29) Rex, M.; Hernandez, F. E.; Campiglia, A. D. Anal. Chem. 2006, 78, 445– 451. (30) Leopold, K.; Foulkes, M.; Worsfold, P. J. Anal. Chem. 2009, 81, 3421– 3428. (31) Lisha, K. P.; Anshup; Pradeep, T. Gold Bull. 2009, 42, 144–152. (32) Barrosse-Antle, L. E.; Xiao, L.; Wildgoose, G. G.; Baron, R.; Salter, C. J.; Crossley, A.; Compton, R. G. New J. Chem. 2007, 31, 2071–2075. (33) Li, B.; Du, Y.; Dong, S. Anal. Chim. Acta 2009, 644, 78–82.

ascorbic acid, and NaCl were ordered from Sigma-Aldrich (Louis, MO). LiCl, KCl, MgCl2, CaCl2, SrCl2, BaCl2, CrCl3, MnCl2, FeCl2, FeCl3, CoCl2, NiCl2, CuCl2, ZnCl2, Cd(ClO4)2, AlCl3, Pb(NO3)2, HgCl2, and AgNO3 were purchased from Acros (Geel, Belgium). Water used in all experiments was doubly distilled and purified by a Milli-Q system (Millipore, Milford, MA). Characterization of the AuNPs. Extinction spectra of the AuNPs were measured using a double-beam UV-visible spectrophotometer (Cintra 10e; GBC, Victoria, Australia). High-resolution transmission electron microscopy (HRTEM, FEI Tecnai G2 F20 S-Twin working at 200 kV) was used to collect HRTEM images of dispersed and aggregated AuNPs. Energy-dispersive X-ray (EDX) spectra were obtained using a HRTEM microscope. The zeta potential and size distribution of the AuNPs were measured using Delsa nano zeta potential and submicrometer particle size analyzer (Beckman Coulter Inc., U.S.). The hydrodynamic size of the AuNPs was measured using dynamic light scattering (DLS) (N5 Submicrometer Particle Size Analyzer, Beckman Coulter Inc., U.S.). To understand the sensing mechanism, we equilibrated aliquots (1.0 mL) of 0.48 nM Tween 20-AuNPs in the presence of Hg2+ (0-10 µM) or Ag+ (0-10 µM) for 5 min at ambient temperature. The resulting mixture was subjected to centrifugation at 17 000 rpm for 10 min. Following removal of the supernatants, the precipitates were washed with water. After five centrifugation/washing cycles, the pellets were resuspended in water. A portion of the samples (∼200 µL) was diluted to 50-fold and then measured by ICP-MS (Perkin-ElmerSCIEX, Thornhill, ON, Canada). Additionally, the composition of the obtained pellets was analyzed by EDX spectroscopy. For surface-assisted laser desorption/ionization time-of-flight ionization mass spectrometry (SALDI-TOF MS) (Autoflex, Bruker) measurements, citrate-capped AuNPs and Tween 20-AuNPs were separately pipetted into a stainless steel 384-well target (Bruker Daltonics) and dried under ambient temperature. Desorption/ ionization was obtained by using a 337-nm-diameter nitrogen laser with a 3 ns pulse width. MS experiments were performed in the positive-ion mode on a reflectron-type TOF MS equipped with a 3 m flight tube. To obtain good resolution and signal-to-noise ratios, the laser power was adjusted to slightly above the threshold, and each mass spectrum was generated by averaging 500 laser pulses. Nanoparticle Synthesis. We prepared citrate-capped AuNPs by means of the chemical reduction of a metal salt precursor (hydrogen tetrachloroaurate, HAuCl4) in the liquid phase. To achieve this, we rapidly added HAuCl4 (0.35 M, 54 µL) to a solution of sodium citrate (2.55 mM, 60 mL) that was heated under reflux. This heating continued for an additional 15 min, during which time the color of the solution changed to a deep red. The size of citrate-capped AuNPs determined by TEM images was 13 ± 1 nm. The SPR wavelength of citrate-capped AuNPs located at 520 nm. The particle concentration of the AuNP solution was estimated to be 4.8 nM by Beer’s law; the extinction coefficient of 13 nm AuNPs at 520 nm is 2.7 × 108 M-1 cm-1. Tween 20-AuNPs were synthesized by adding Tween 20 (10% v/v, 240 µL) to a solution of citrate-capped AuNPs (4.8 Analytical Chemistry, Vol. 82, No. 16, August 15, 2010

6831

Scheme 1. Illustration of the Mechanism of Tween 20-AuNPs for Sensing Hg2+ and Ag+

nM, 60 mL).34 To investigate the effect of the kind of the surfactant on the sensing of Hg2+ and Ag+, Tween 20 was replaced by Tween 40, Tween 60, and Tween 80, once at a time. Additionally, to test the effect of the type of the AuNPs on the detection of Hg2+ and Ag+, we replaced citrate-capped AuNPs with bare AuNPs in the synthesis of Tween 20-AuNPs. The preparation of bare AuNPs (3.7 nM; 12 nm) is described in the Supporting Information (SI). Sample Preparation. Tween 20-AuNPs were prepared in 100-1000 mM sodium phosphate solution at pH 12.0. For Hg2+ sensing, metal ions (800 µL, 125-1250 nM) were added to a solution containing Tween 20-AuNPs (100 µL, 0.96-14.4 nM) and NaCl (100 µL, 1 M). For Ag+ sensing, metal ions (800 µL, 125-1250 nM) were added to a solution containing Tween 20AuNPs (100 µL, 0.96-14.4 nM) and EDTA (100 µL, 0.1 M). We equilibrated the resulting solutions at ambient temperature for the optimum incubation time, and then recorded the extinction spectra of the solutions. Analysis of Real Samples. Samples of drinking water and seawater (pH 7.9) were collected from National Sun Yat-sen University campus. We then prepared a series of samples by “spiking” them with standard solutions of Hg2+ (125-1250 nM) or Ag+ (375-1250 nM). These spiked samples (800 µL) were added either to a solution containing 100 µL of 2.4 nM Tween 20-AuNPs and 100 µL of 1 M NaCl or to a solution containing 100 of 4.8 nM Tween 20-AuNPs and 100 µL of 0.1 M EDTA. We incubated the resulting solutions for 5 min before measuring their extinction spectra. On the other hand, this proposed method was utilized to detect 10 nm AgNPs in drinking water. The preparation of citrate-capped AgNPs35 is described in the SI. Different concentrations of AgNPs (1-10 pM) present in drinking water were oxidized to Ag+ with (34) Huang, C.-C.; Tseng, W.-L. Analyst 2009, 134, 1699–1705. (35) Wei, H.; Chen, C.; Han, B.; Wang, E. Anal. Chem. 2008, 80, 7051–7055.

6832

Analytical Chemistry, Vol. 82, No. 16, August 15, 2010

a solution of 1 µM H2O2 and 1 µM H3PO4. After 10 min, the resulting solution (800 µL) was added to a solution containing 100 of 4.8 nM Tween 20-AuNPs and 100 µL of 0.1 M EDTA. The extinction spectra of the resulting solutions were recorded after 5 min incubation. RESULTS AND DISCUSSION Sensing Mechanism. SALDI-TOF MS technique is applied to the detection of small molecules that are adsorbed on the surface of the AuNPs when AuNPs are used as SALDI matrices.36,37 Thus, we utilized this technique to determine whether citrate ions were adsorbed on the surface of Tween 20-AuNPs. SI Figure S1A shows the SALDI spectrum of citrate-capped AuNPs. The peak detected at m/z 258.00 corresponded to [citrate +3Na]+. This peak was also observed in the SALDI spectrum of Tween 20AuNPs (SI Figure S1B), indicating that citrate ions are capped on the surface of Tween 20-AuNPs. Moreover, the zeta potential of Tween 20-AuNPs was found to be -14.0 ± 0.8 mV in 10 mM phosphate at pH 12.0. These results suggest that a neutral Tween 20 coating only shields the citrate ions (no displacement occurs). Accordingly, we reasoned that citrate ions adsorbed onto the NP surface can act as a reducing agent for Hg2+ and Ag+ when citrate-capped AuNPs have been modified with Tween 20. Scheme 1 shows the mechanism by which Tween 20-AuNPs sense Hg2+ and Ag+. Because of the high affinity between Au and Hg,29-32 the reduced Hg(0) is directly deposited onto the Au surface through the formation of Hg-Au alloys; meanwhile, Tween 20 molecules are desorbed from the AuNPs. The removal of the stabilizer (Tween 20) causes the AuNPs to aggregate under conditions of high ionic strength. Also, citrate ions can be used (36) Su, C.-L.; Tseng, W.-L. Anal. Chem. 2007, 79, 1626–1633. (37) Wu, H.-P.; Yu, C.-J.; Lin, C.-Y.; Lin, Y.-H.; Tseng, W.-L. J. Am. Soc. Mass Spectrom. 2009, 20, 875–882.

Figure 1. Extinction spectra of solutions of (A) 0.48 nM Tween 20AuNPs and (B) 0.37 nM Tween 20-modified bare AuNPs (a) before and (b, c) after the addition of (b) 1 µM Hg2+ and (c) 1 µM Ag+. Tween 20-AuNPs are prepared in 20 mM phosphate at pH 12.0. The incubation time is 5 min.

to reduce Ag+ onto the Au surface.38,39 The formation of Agcoated AuNPs enables Tween 20 to be removed from the NP surface, thereby inducing aggregation of the AuNPs in a highionic-strength solution. Evidence for the Formation of Hg-Au Alloys and Ag Shells. To test the hypothesis mentioned above, we monitored the extinction spectra of Tween 20-AuNPs in the absence and presence of Hg2+ and Ag+. Curve a in Figure 1A shows that the SPR wavelength of Tween 20-AuNPs appears at 520 nm, indicating that they are well dispersed in 20 mM phosphate solution at pH 12.0. In other words, Tween 20 molecules can indeed protect citrate-capped AuNPs against a high-ionic-strength solution.40 After separately adding 1.0 µM Hg2+ (curve b) and 1.0 µM Ag+ (curve c), we observed a decrease in the strength of SPR band at 520 nm and the formation of a new red-shift band. These changes were characteristic of AuNP aggregation. The Hg2+- and Ag+(38) Xie, W.; Su, L.; Donfack, P.; Shen, A.; Zhou, X.; Sackmann, M.; Materny, A.; Hu, J. Chem. Commun. 2009, 5263–5265. (39) Xia, H.; Bai, S.; Hartmann, J.; Wang, D. Langmuir 2009. (40) Shen, C.-C.; Tseng, W.-L.; Hsieh, M.-M. J. Chromatogr., A 2009, 1216, 288–293.

induced NP aggregation was nearly complete after 5 min (SI Figure S2). Obviously, the deposition of Hg and Ag onto the Au surface enables Tween 20 to be removed, thereby driving NP aggregation. We ruled out the possibility that coordination between Tween 20 and metal ions induces the NP aggregation because Tween 20 does not contain any functional group to interact with Hg2+ and Ag+. To ensure the role of citrate ions in the reduction of Hg2+ and Ag+, the extinction spectra of bare AuNPs modified with Tween 20 were examined under the same conditions. The addition of Hg2+ and Ag+ to this type of AuNP resulted in a rare shift in the SPR wavelength (Figure 1B), clearly indicating that citrate ions are indispensable for detecting Hg2+ and Ag+ using Tween 20-AuNPs. On the other hand, we prepared Hg-Au alloy- and Ag-coated AuNPs and modified them with Tween 20. Compared to Tween 20-AuNPs, a blue shift in SPR and a decrease in SPR intensity were observed for Tween 20-modified Hg-Au alloy-coated AuNPs (SI Figure S3). This is attributed to the formation of Hg-Au alloy on the surface of the AuNPs. A similar phenomenon was reported when citrate-capped AuNPs were exposed in the presence of Hg vapor.41 Moreover, this kind of NPs was found to be unstable in a highionic-strength solution. This result reflects that Tween 20 molecules were not attached to the surface of Hg-Au alloy, thereby incapable of protecting NPs against a high-ionic strength solution. SI Figure S4 shows that the deposition of Ag on the surface of the AuNPs resulted in an increase SPR intensity relative to Tween 20-AuNPs.42,43 Similarly, Tween 20-modified Ag-coated AuNPs were aggregated in a high-ionic-strength solution because Tween 20 molecules were not adsorbed on the surface of Ag shell. To provide further evidence for the formation of Hg-Au alloys and Ag shells, we used ICP-MS to quantitatively determine the composition of the precipitates, which were obtained by five cycles of centrifugation of a solution of Tween 20-AuNPs and metal ions. When a series of concentrations (0-1 µM) of Hg2+ were present in a solution of 0.48 nM Tween 20-AuNPs, the molar ratio of Hg to Au in the precipitates gradually increased with increasing Hg2+ concentration (Figure 2A). A similar phenomenon was seen in the case of Ag+ (Figure 2B). If Hg-Au alloys and Ag shells did not form on the Au surface, the molar ratios of Hg to Au and Ag to Au should remain constant under these conditions. Under identical treatment conditions (five centrifugation/washing cycles), the composition of precipitates was also determined by EDX analysis. The Hg content in the precipitates increased with increasing Hg2+ concentration (Figure 2C). We observed a similar phenomenon when different concentrations of Ag+ were present in a solution of Tween 20-AuNPs (Figure 2D). These results are in agreement with those obtained by ICP-MS. These findings strongly support the idea that Hg2+- and Ag+-induced aggregation of Tween 20-AuNPs is indeed the result of the formation of Hg-Au alloys and Ag on the Au surface. Effect of Surfactant Chain Length, NP Concentration, and Ionic Strength. We next explored the effect of surfactant chain length on the metal-ion-induced aggregation of the AuNPs. The (41) Morris, T.; Copeland, H.; McLinden, E.; Wilson, S.; Szulczewski, G. Langmuir 2002, 18, 7261–7264. (42) Anandan, S.; Grieser, F.; Ashokkumar, M. J. Phys. Chem. C 2008, 112, 15102–15105. (43) Guerrini, L.; Garcia-Ramos, J. V.; Domingo, C.; Sanchez-Cortes, S. J. Raman Spectrosc. 2010, 41, 508–515.

Analytical Chemistry, Vol. 82, No. 16, August 15, 2010

6833

Figure 2. (A, B) Effect of the concentration of (A) Hg2+ and (B) Ag+ on the composition ratio of (A) Au to Ag and (B) Au to Hg of the precipitates. A series of concentration of (A) 0-1000 nM Hg2+ and (B) 0-1000 nM Ag+ was added to 1.0 mL of 0.48 nM Tween 20-AuNPs. The precipitates were obtained by five cycles of centrifugation of the resulting solutions. (C, D) EDX spectra of the precipitates obtained after the addition of (a) 0.1, (b) 1, and (c) 10 µM (C) Hg2+ and (D) Ag+ to a solution of 0.48 nM Tween 20-AuNPs. Tween 20-AuNPs are prepared in 20 mM phosphate at pH 12.0. The incubation time is 5 min.

extinction values of the solution at 650 and 520 nm corresponded to the quantities of dispersed and aggregated AuNPs, respectively. Thus, the molar ratio of dispersed to aggregated AuNPs can be expressed by the ratio of the extinction value Ex at 650 nm to that at 520 nm (Ex650 nm/Ex520 nm). As shown in Figure 3A, the addition of both Hg2+ and Ag+ to a solution of Tween 20-AuNPs resulted in a high value of Ex650 nm/Ex520 nm. However, when we replaced Tween 20 with Tween 40, the value of Ex650 nm/ Ex520 nm became small. This suggests a relatively small amount of Hg-Au alloys or Ag on the surface of Tween 40-modified AuNPs, resulting in a small degree of NP aggregation. Similar phenomena were observed in the case of Tween 60- and Tween 80-modifed AuNPs. To further confirm our hypothesis, ICPMS was used to determine the composition of the NPs. After adding 1 µM Hg2+ to different kinds of AuNPs, the concentrations of Hg in Tween 20-, 40-, 60-, and 80-modified AuNPs were 148, 64, 58, and 51 ppb, respectively. Similarly, upon the addition of 1 µM Ag+, the concentrations of Ag in Tween 20-, 40-, 60-, and 80-modified AuNPs were 65, 19, 24, and 25 ppb, respectively. On the basis of these results, Tween 20 was selected for the following studies. Previous studies have shown that the sensitivity of a AuNPbased sensor is highly dependent on the concentration of 6834

Analytical Chemistry, Vol. 82, No. 16, August 15, 2010

AuNPs.44 At relatively high concentrations of Tween 20-AuNPs, the deposition of Hg2+ or Ag+ on the surface of a single particle decreased, reducing the degree of NP aggregation. Moreover, the aggregation rate of NPs increased with increasing NP concentration. It can be seen that the optimum concentrations of Tween 20-AuNPs for sensing Hg2+ and Ag+ were 0.24 and 0.48 nM, respectively, when the incubation time was fixed at 5 min (Figure 3B). Moreover, we investigated the effect of Na3PO4 concentration on the colorimetric sensitivity of Tween 20AuNPs to Hg2+ and Ag+. SI Figure S5 shows that the zeta potential of Tween 20-AuNPs reduced with increasing Na3PO4 concentration, implying that electrostatic repulsion between Tween 20-AuNPs decreased with increasing ionic strength of the solution. Thus, under conditions of high ionic strength, the slight electrostatic repulsion between Tween 20-AuNPs provided a low barrier for metal-ion-induced NP aggregation. As expected, the difference in Ex650 nm/Ex520 nm for the cases with and without 1.0 µM Hg2+ gradually increased with increasing Na3PO4 concentration and reached a plateau at 80 mM Na3PO4 (Figure 3C). A similar effect was found in the analysis of Ag+ (Figure 3D). Consequently, 80 mM Na3PO4 was chosen for the following studies. (44) Huang, C.-C.; Tseng, W.-L. Anal. Chem. 2008, 80, 6345–6350.

Figure 3. (A) The value of Ex650 nm/Ex520 nm of Tween 20- 40-, 60-, and 80-modified AuNPs (0.48 nM) after the addition of (a) 1 µM Hg2+ and (b) 1 µM Ag+. (B) Effect of the concentration of Tween 20-AuNPs on the ratio Ex650 nm/Ex520 nm in the presence of (a) 1 µM Hg2+ and (b) 1 µM Ag+. (C) Effect of phosphate concentration on the ratio Ex650 nm/Ex520 nm of 0.24 nM Tween 20-AuNPs in the (a) absence and (b) presence of 1 µM Hg2+. (D) Effect of phosphate concentration on the ratio Ex650 nm/Ex520 nm of 0.48 nM Tween 20-AuNPs in the (a) absence and (b) presence of 1 µM Ag+. (A, B) Tween 20-AuNPs are prepared in 20 mM phosphate at pH 12.0. (A-D) The incubation time is 5 min.

Selectivity, Sensitivity, and Application. To realize the selectivity of 0.24 and 0.48 nM Tween 20-AuNPs toward Hg2+ and Ag+, respectively, other metal ionssincluding Li+, Na+, K+, Mg2+, Ca2+, Sr2+, Ba2+, Cr3+, Mn2+, Fe2+, Fe3+, Co2+, Ni2+, Cu2+, Zn2+, Cd2+, Al3+, Pb2+, Hg2+, and Ag+swere examined under identical conditions. Only Hg2+ and Ag+ caused the aggregation of both 0.24 and 0.48 nM Tween 20-AuNPs (SI Figure S6), revealing that this probe is selective for Hg2+ and Ag+. To circumvent this problem, we tested the effect of masking agents, including NaCl and EDTA. It is well-known that the solubility product of AgCl is 1.8 × 10-10, whereas HgCl2 is soluble in water (70 g/L). Accordingly, we tested the ability of NaCl to mask the interfering metal ions in our sensing system. In the presence of 0.1 M NaCl, the addition of 1 µM Hg2+ to a solution of 0.24 nM Tween 20-AuNPs resulted in an apparent change in Ex650 nm/Ex520 nm, whereas the remaining metal ions (100 µM) had negligible effects on the same system (Figure 4A). The selectivity of this probe is more than 100-fold for Hg2+ over all other tested metal ions. Additionally, to improve the selectivity of 0.48 nM Tween 20-AuNPs for Ag+, we chose

EDTA as a masking agent, since it forms a more stable complex with Hg2+ than with Ag+. As expected, the presence of 0.01 M EDTA masked Tween 20-AuNPs toward Hg2+ (Figure 4B). As a result, Tween 20-AuNPs provided high selectivity (>100-fold) toward Ag+ over all other tested metal ions. Under optimum conditions (for sensing Hg2+: 0.24 nM Tween 20-AuNPs, 80 mM Na3PO4, and 0.1 M NaCl; for sensing Ag+: 0.48 nM Tween 20-AuNPs, 80 mM Na3PO4, and 0.01 M EDTA), we evaluated the sensitivity of this probe toward Hg2+ and Ag+. When the concentrations of Hg2+ varied from 0 to 1000 nM, the extinction spectra of 0.24 nM Tween 20-AuNPs showed a gradual increase in extinction at 650 nm and their color gradually changed from red to purple (Figure 5A). A solution of 0.48 nM Tween 20-AuNPs showed a similar response to Ag+ (Figure 5B). These spectra showed clear isosbestic points at 556 and 549 nm upon the addition of Hg2+ and Ag+, respectively. This observation reveals that the aggregation of Tween 20-AuNPs is directly related to the concentration of Hg2+ or Ag+. This result was further confirmed by HRTEM images and DLS measureAnalytical Chemistry, Vol. 82, No. 16, August 15, 2010

6835

Figure 4. The value of Ex650 nm/Ex520 nm of a solution of 80 mM Na3PO4 containing (A) 0.24 nM Tween 20-AuNPs and 0.1 M NaCl and (B) 0.48 nM Tween 20-AuNPs and 0.01 M EDTA upon the addition of (A) 1 µM Hg2+ and 100 µM other metal ions and (B) 1 µM Ag+ and 100 µM other metal ions. The incubation time is 5 min.

ments. SI Figure S7 shows the concentration-dependent TEM images of the aggregated AuNPs after adding different concentrations (0, 0.1, 1, and 10 µM) of Hg2+ or Ag+ to a solution of Tween 20-AuNPs. Moreover, the hydrodynamic size of the aggregated AuNPs increased with an increase in the concentration of Hg2+ or Ag+ (SI Figure S8). These results provide clear evidence that the aggregation degree of Tween 20-AuNPs is highly dependent on the concentration of Hg2+ or Ag+. We observed that the ratio Ex650 nm/Ex520 nm increased linearly with increasing Hg2+ and Ag+ concentration over the range of 200 to 800 nM (R2 ) 0.9943) and 400 to 1000 nM (R2 ) 0.9935), respectively (Figure 5C and D). A difference in linearity between two metal ions could be due to that the sensing mechanism of Tween 20-AuNPs for Hg2+ is different from that for Ag+. This probe could detect Hg2+ and Ag+ at concentrations as low as 100 and 100 nM, respectively. The result is useful for detecting Ag+ in drinking water, because the maximum level of silver in drinking water permitted by the United States Environmental Protection Agency (EPA) is 50 µg/L (∼460 nM). To test the practicality of the present approach, a solution of 0.24 nM Tween 20-AuNPs was used to analyze Hg2+ in drinking water and seawater. As shown in SI Figures S9 and S10, Ex650 nm/Ex520 nm increased linearly upon increasing the spiked concentration of Hg2+ in drinking water over the range of 200-600 nM (R2 ) 0.9944) and in seawater over the range of 300-1000 nM (R2 ) 0.9977). Evidence of the Hg2+-induced aggregation of Tween 20-AuNPs in seawater can be seen in the HRTEM images (SI Figure S11). The lowest detectable concentrations of Hg2+ in drinking water and seawater were 6836

Analytical Chemistry, Vol. 82, No. 16, August 15, 2010

estimated to be 200 and 100 nM, respectively. Although the sensitivity of this probe is insufficient to detect the maximum level of mercury (2 ppb) in drinking water permitted by the U.S. EPA, we suggest that Tween 20-AuNPs can be used as probes for solid-phase preconcentration of mercury in complex matrices prior to ICP-MS analysis.30 We also evaluated the feasibility of this approach for sensing of Ag+ and AgNPs in drinking water. We obtained a linear correlation (R2 ) 0.9963) between the ratio Ex650 nm/Ex520 nm and the concentration of Ag+ spiked into the drinking water over the range of 400-1000 nM (SI Figure S12), which includes the maximum permissible limit of silver in drinking water. Moreover, since hazardous AgNPs may pose threats to human health or the environment,45 this approach was further used to monitor AgNPs in drinking water. Under acidic conditions (1.0 µM H3PO4), AgNPs (10 ± 2 nm) are oxidized to Ag+ ions with 1.0 mM H2O2.12 The oxidation of AgNPs to Ag+ was complete after 10 min. The generated Ag+ was directly detected by 0.48 nM Tween 20-AuNPs. The degree of aggregation of Tween 20AuNPs increased when the concentration of AgNPs was increased from 1 to 10 pM (SI Figure S13). The correlation coefficient (R2) for the determination of AgNPs in the range 1-6 pM was 0.9988. These results suggest that this probe will be suitable for routine assays of AgNPs in consumer products such as cosmetics and fabrics.46 Table 1 shows the quantitative measurements of Hg2+, Ag+, and AgNPs in different matrices based on the use of Tween 20-AuNPs. (45) Lubick, N. Environ. Sci. Technol. 2008, 42, 8617. (46) Benn, T. M.; Westerhoff, P. Environ. Sci. Technol. 2008, 42, 4133–4139.

Figure 5. Extinction spectra and color changes of solutions containing (A) 0.24 nM Tween 20-AuNPs and 0.1 M NaCl and (B) 0.48 nM Tween 20-AuNPs and 0.01 M EDTA upon the addition of (A) 0-1000 nM Hg2+ and (B) 0-1000 nM Ag+. The arrows indicate the signal changes with increases in analyte concentrations (A: 0, 100, 200, 300, 400, 500, 600, 700, 800, and 1000 nM; B: 0, 100, 200, 300, 400, 500, 600, 700, 800, 900, and 1000 nM). A plot of Ex650 nm/Ex520 nm versus the concentration of (C) Hg2+ and (D) Ag+. The incubation time is 5 min. The error bars represent standard deviations based on three independent measurements. Table 1. Quantification of Hg2+, Ag+ and AgNPs in the Different Sample Matrix Based on the Use of Tween 20-AuNPs analyte

matrix

linear rang (M)

R2

MDC (M)a

Hg2+ Hg2+ Hg2+ Ag+ Ag+ AgNPs

deionized water drinking water seawater deionized water drinking water drinking water

2 × 10-7 to 8 × 10-7 2 × 10-7 to 6 × 10-7 3 × 10-7 to 1 × 10-6 4 × 10-7 to 1 × 10-6 4 × 10-7 to 1 × 10-6 6 × 10-12 to 1 × 10-11

0.9943 0.9944 0.9977 0.9935 0.9963 0.9988

1 × 10-7 2 × 10-7 1 × 10-7 1 × 10-7 3 × 10-7 1 × 10-12

a

MDC, minimum detectable concentration.

CONCLUSIONS This study reports a new assay for the selective detection of Hg2+ and Ag+ using Tween 20-AuNPs. This probe can be further applied to detecting hazardous AgNPs. We demonstrated that the aggregation of Tween 20-AuNPs results from the formation of Hg-Au alloy or Ag on the surface of the

AuNPs. Thus, in the opinion of the authors, Tween 20-AuNPs should be used as a selective probe for extracting a large volume of Hg2+ and Ag+ prior to ICP-MS analysis. Moreover, we believe that the present approach holds great potential for monitoring Ag+ and AgNPs in environmental samples. ACKNOWLEDGMENT We thank National Science Council (NSC 98-2113-M-110-009MY3) and National Sun Yat-sen University-Kaohsiung Medical University Joint Research Center for the financial support of this study. SUPPORTING INFORMATION AVAILABLE Experimental details, additional references, and Figures S1-13. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review March 27, 2010. Accepted June 25, 2010. AC1007909

Analytical Chemistry, Vol. 82, No. 16, August 15, 2010

6837