Anal. Chem. 2010, 82, 8566–8572
Detection of Copper Ions Through Recovery of the Fluorescence of DNA-Templated Copper/Silver Nanoclusters in the Presence of Mercaptopropionic Acid Yu-Ting Su, Guo-Yu Lan, Wei-Yu Chen, and Huan-Tsung Chang* Department of Chemistry, National Taiwan University, 1, Section 4, Roosevelt Road, Taipei 106, Taiwan We have developed a simple and homogeneous fluorescence assay, comprised of 3-mercaptopropionic acid (MPA) and DNA-Cu/Ag nanoclusters (NCs) in aqueous solution, for the detection of Cu2+ ions. The fluorescence of the DNA-Cu/Ag NCs was quenched by MPA, which was recovered in the presence of Cu2+ ions. This MPAinduced fluorescence quenching arises through changes in the DNA conformation that occur after interactions between MPA and the Cu/Ag clusters. The MPAinduced fluorescence quenching displayed typical characteristics in Stern-Volmer plots; it followed a static quenching mechanism. The presence of Cu2+ ions resulted in the oxidation of MPA to form a disulfide compound, leading to recovery of the fluorescence of the DNA-Cu/Ag NCs. The fluorescence of the DNA-Cu/ Ag NCs in the presence of MPA increased upon increasing the concentration of Cu2+ ions over the range from 5 to 200 nM. The DNA-Cu/Ag NC probe provided the limit of detection (at a signal-to-noise ratio of 3) for Cu2+ ions of 2.7 nM, with high selectivity (by at least 2300-fold over other tested metal ions). We validated the practicality of using this probe for the detection of Cu2+ ions in environmental samples through analyses of Montana soil and pond water samples. Gold nanodots (Au NDs) and silver nanoclusters (Ag NCs) consisting of a few to tens of atoms possess fluorescence properties because of strong quantum-confinement effects, and they comprise a new class of fluorophores for the sensitive detection of several important analytes, including Hg2+ ions, H2O2, glucose, and concanavalin A.1 These fluorescent NCs and NDs provide high emission rates and large Stokes shifts.2 In addition, the emissions of Au NDs are readily tuned from the UV to the near-IR region by varying their sizes, through control * To whom correspondence should be addressed. Phone and fax: 011-886-233661171. E-mail:
[email protected]. (1) (a) Huang, C.-C.; Yang, Z.; Lee, K.-H.; Chang, H.-T. Angew. Chem., Int. Ed. 2007, 119, 6948–6952. (b) Huang, C.-C.; Chen, C.-T.; Shiang, Y.-C.; Lin, Z.-H.; Chang, H.-T. Anal. Chem. 2009, 81, 875–882. (c) Shiang, Y.-C.; Huang, C.-C.; Chang, H.-T. Chem. Commun. 2009, 23, 3437–3439. (2) (a) Vosch, T.; Antoku, Y.; Hsiang, J.-C.; Richards, C. I.; Gonzalez, J. I.; Dickson, R. M. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 12616–12621. (b) Zheng, J.; Nicovich, P. R.; Dickson, R. M. Annu. Rev. Phys. Chem. 2007, 58, 409–431.
8566
Analytical Chemistry, Vol. 82, No. 20, October 15, 2010
of the molar ratios of the reducing/capping agents to the metal ions and the nature of the capping agents.3 On the other hand, with the control of the sequence and length of oligonucleotides, Ag NCs of various sizes that emit colors from blue to red have been prepared through NaBH4-mediated reduction of AgNO3.4 By taking advantage of Hg2+-induced fluorescence quenching, oligonucleotide-stabilized Ag NCs have been employed to detect Hg2+ ions at concentrations as low as 5 nM.5 Proteins, such as bovine serum albumin, and polymers, such as poly(methacrylic acid), have also been used as templates for the preparation of fluorescent Au and Ag NCs.6 Unlike nonfluorescent Au and Ag NPs, Ag NCs and Au NDs have yet to become popular sensing materials, mainly because they are more difficult to prepare on large scales and because of their poor stability. In a previous study, we prepared DNA-Cu/Ag NCs from Cu2+ ions and DNA-Ag NCs, which we had synthesized from AgNO3 and NaBH4 in the presence of DNA (sequence: 5′-CCCTTAATCCCC-3′). The fluorescence intensity of the DNA-Cu/ Ag NCs (quantum yield, 51.2%) was ∼4.5-fold higher than that of the DNA-Ag NCs. The fluorescence of the DNA-Ag NCs increased upon increasing the Cu2+ ion concentration, allowing the detection of Cu2+ ions at concentrations as low as 10 nM.7 Detection of Cu2+ ions is important because short- or longterm exposure to high levels of Cu2+ ions can lead to gastrointestinal disturbances or damage to the liver and kidneys, respectively.8 The U.S. Environmental Protection Agency (EPA) has set the safe limit of Cu2+ ions in drinking water at 1.3 ppm (∼20 µM). Many current techniques for the (3) (a) Huang, C.-C.; Liao, H.-Y.; Shiang, Y.-C.; Lin, Z.-H.; Yang, Z.; Chang, H.-T. J. Mater. Chem. 2009, 19, 755–759. (b) Zheng, J.; Zhang, C.; Dickson, R. M. Phys. Rev. Lett. 2004, 93, 077402. (4) (a) Ritchie, C. M.; Johnsen, K. R.; Kiser, J. R.; Antoku, Y.; Dickson, R. M.; Petty, J. T. J. Phys. Chem. C 2007, 111, 175–181. (b) Richards, C. I.; Choi, S.; Hsiang, J.-C.; Antoku, Y.; Vosch, T.; Bongiorno, A.; Tzeng, Y.-L.; Dickson, R. M. J. Am. Chem. Soc. 2008, 130, 5038–5039. (c) Sengupta, B.; Ritchie, C. M.; Buckman, J. G.; Johnsen, K. R.; Goodwin, P. M.; Petty, J. T. J. Phys. Chem. C 2008, 112, 18776–18782. (d) Gwinn, E. G.; O’Neill, P.; Guerrero, A. J.; Bouwmeester, D.; Fygenson, D. K. Adv. Mater. 2008, 20, 279–283. (5) Guo, W.; Yuan, J.; Wang, E. Chem. Commun. 2009, 3395–3397. (6) (a) Xie, J.; Zheng, Y.; Ying, J. Y. J. Am. Chem. Soc. 2009, 131, 888–889. (b) Shang, L.; Dong, S. Chem. Commun. 2008, 9, 1088–1090. (7) Lan, G.-Y.; Huang, C.-C.; Chang, H.-T. Chem. Commun. 2010, 46, 1257– 1259. (8) (a) Georgopoulos, P. G.; Roy, A.; Yonone-Lioy, M. J.; Opiekun, R. E.; Lioy, P. J. J. Toxicol. Environ. Health, Part B 2001, 4, 341–394. (b) Judith, R. T. Am. J. Clin. Nutr. 1998, 67, 960–964. 10.1021/ac101659d 2010 American Chemical Society Published on Web 09/28/2010
detection of Cu2+ ions, involving inductively coupled plasma mass spectrometry (ICPMS), atomic absorption/emission spectroscopy, and atomic fluorescence spectrometry, require expensive instruments and/or complicated sample preparation processes.9 In contrast, detection with chemosensors offers the potential for simple, low-cost, and rapid tracking of Cu2+ ions in biological, toxicological, and environmental samples.10 Fluorescent poly(methacrylic acid)-templated Ag NCs have been developed for the detection of Cu2+ ions, using a “turnoff” approach, through the interactions of the Cu2+ ions with the free carboxyl groups of the polymers surrounding the emissive Ag NCs.11 Glutathione-capped Au NDs have also been used as a Cu2+ chemosensor, based on aggregation-induced fluorescence quenching.12 In this study, we used a combination of DNA-Cu/Ag NCs and 3-mercaptopropionic acid (MPA) to detect Cu2+ ions. We found that MPA-induced fluorescence quenching of DNA-Cu/Ag NCs was suppressed in the presence of Cu2+ ions, allowing detection of Cu2+ ions at concentrations as low as 2.7 nM. We investigated the roles that the nature and concentration of three thiol compounds played in inducing fluorescence quenching of the DNA-Cu/Ag NCs and, thus, in determining the sensitivity toward Cu2+ ions. Finally, we validated the practicality of this method through analyses of soil and water samples. EXPERIMENTAL SECTION Chemicals. All the metal salts, cysteine (Cys), 2-mercaptoethanol (2-ME, >99%), MPA (>99.0%), and NaBH4 (powder, 98%) were purchased from Aldrich (Milwaukee, WI). Sodium phosphate dibasic anhydrous and sodium phosphate monobasic monohydrate, used to prepare phosphate buffers (50 mM, pH 3.0-11.0), were obtained from J. T. Baker (Phillipsburg, NJ). Copper nitrate trihydrate and silver nitrate (99+%, ACS reagent) were obtained from Acros (Morris Plains, NJ). The solution of DNA (5′-CCCTTAATCCCC-3′) was purchased from Integrated DNA Technology (Coralville, IA). Montana soil (SRM 2710) was obtained from the National Institute of Standards and Technology (Gaithersburg, MD). Milli-Q ultrapure water was used in all experiments. DNA-Cu/Ag NCs. The DNA-Cu/Ag NCs were prepared using a slight modification of a literature procedure.7 Briefly, solutions of AgNO3 (1 mM, 22.5 µL) and Cu(NO3)2 (1 mM, 7.5 µL) were added to aliquots (55 µL) of the DNA solution (50 µM) in phosphate buffer (20 mM, pH 7.0). After a 15 min incubation in an ice bath, the mixture was reduced by adding freshly prepared NaBH4 (2 mM, 15 µL) under vigorous shaking. The reduced DNA-Cu/Ag NCs solution was then kept in the dark for 90 min at room temperature. For simplicity, we denote the concentration of the solution of the as-prepared DNA-Cu/Ag NCs as “1X.” The as-prepared DNA-Cu/Ag NCs were stable for at least 2 weeks when stored at 4 °C in the dark. The fluorescence spectra of the DNA-Cu/Ag NCs were recorded (9) (a) Li, Y.; Chen, C.; Li, B.; Sun, J.; Wang, J.; Gao, Y.; Zhao, Y.; Chai, Z. J. Anal. At. Spectrom. 2006, 21, 94–96. (b) Pourreza, N.; Hoveizavi, R. Anal. Chim. Acta 2005, 549, 124–128. (10) (a) Xie, J.; Me´nand, M.; Maisonneuve, S.; Me´tivier, R. J. Org. Chem. 2007, 72, 5980–5985. (b) Xu, G.-R.; Yuan, Y.; Kim, S.; Lee, J.-J. Electroanalysis 2008, 20, 1690–1695. (11) Shang, L.; Dong, S. J. Mater. Chem. 2008, 18, 4636–4640. (12) Chen, W.; Tu, X.; Guo, X. Chem. Commun. 2009, 13, 1736–1738.
using a Synergy 4 hybrid multimode microplate reader (Winooski, VT). When excited at 480 nm, the DNA-Cu/Ag NCs exhibited fluorescence centered at a wavelength of 576 nm. The fluorescence quantum yield (Φf) for the NCs was calculated using fluorescein in 0.1 N NaOH (Φf ) 0.95) as a reference chromophore.13 Prior to conducting electrospray ionization-mass spectrometry (ESI-MS), transmission electron microscopy with energy dispersive spectroscopy (TEM-EDS), and X-ray photoelectron spectroscopy (XPS) analyses, we purified the fluorescent DNA-Ag NCs and DNA-Cu/Ag NCs (500 µL) by conducting centrifugal filtration (13 500g) for 15 min through a filter having a cutoff of 3 kDa and washed with ultrapure water (450 µL). Samples were prepared by mixing a solution of the DNA-Cu/Ag NCs (50 µL) with isopropyl alcohol (50 µL) to enhance the ionization efficiency in ESI-MS.14 Prior to TEM and XPS measurements, drops of the DNA-Cu/Ag NCs solution were carefully deposited onto 400-mesh nickle-coated grids and a silicon wafer, respectively. We used 5 mW solid-state lasers with outputs at 375 and 475 nm from Uniphase (Mantence, CA) to irradiate Cu/Ag NCs solutions under homogeneous stirring to check whether the absorbance bands centered at 425 and 480 nm were from the same absorbing species. Thiol-Induced Fluorescence Quenching. Mixtures (500 µL) of the DNA-Cu/Ag NCs (0.01X), phosphate buffer (20 mM, pH 8.0), and a thiol (MPA, 2-ME, or Cys; concentration, 0-5.0 µM) were equilibrated in the dark at ambient temperature for 30 min. Subsequently, the mixtures were subjected to fluorescence measurement with excitation at 480 nm. The fluorescence lifetimes of the DNA-Cu/Ag NC solutions in the presence and absence of each thiol were recorded using a photocounting system from Jobin Yvon (Edison, NJ), with laser excitation at 475 nm. A J500A circular dichroism (CD) spectropolarimeter (Jasco, Park City, KY) was employed to study the conformational changes of the DNA in the DNA-Cu/Ag NCs in the absence or presence of the thiols. An ESCALAB 250 X-ray photoelectron spectroscope (Thermo VG Scientific, West Sussex, U.K.), with Al KR X-ray radiation as the excitation X-ray source, was used to determine the oxidation states of the Ag atoms in the DNA-Cu/Ag NCs in the absence and presence of MPA. DNA-Cu/Ag NC Probe for Cu2+ Ions. A stock Cu(NO3)2 solution (10 mM) was prepared in 1 mM HNO3. Phosphate solutions (20 mM, pH 8.0) containing MPA (2.5 µM) and various amounts of Cu2+ (0-2.5 µM) were also prepared. The standard solutions were incubated at room temperature for 30 min and then aliquots of the as-prepared DNA-Cu/Ag NCs (0.1X, 50 µL) were added to the mixtures (450 µL). After 30 min at ambient temperature, aliquots of the mixtures were transferred into a microplate (well volume, 0.2 mL) for fluorescence measurement (excitation wavelength, 480 nm). Analysis of Soil and Pond Water Samples. Acidic digestion of soil samples (1.000 g) was preformed according to EPA Method 305B.14 Aliquots (50 µL) of the diluted soil samples (0.01 mg/ mL) were spiked with standard Cu2+ solutions (final concentration, 25-125 nM). Prior to analysis, the spiked samples were diluted to 450 µL with phosphate solution (20 mM, pH 8.0) (13) Lakowicz, J.; Masters, B. J. Biomed. Opt. 2008, 13, 029901-029902. (14) Brown, T. L.; Rice, J. A. Anal. Chem. 2000, 72, 384–390.
Analytical Chemistry, Vol. 82, No. 20, October 15, 2010
8567
Scheme 1. Schematic Representation of the Operation of the DNA-Cu/Ag NC Probe in the Presence of Thiols for the Detection of Cu2+ Ions: (a) Thiol-Induced Fluorescence Quenching of the DNA-Cu/Ag NCs and (b) Cu2+ Ions Suppressing the Thiol-Induced Fluorescence Quenching of the DNA-Cu/Ag NCs
RESULTS AND DISCUSSION Characterization of the DNA-Cu/Ag NCs. In order to confirm the absorbing species of the as-prepared DNA-Cu/Ag NCs, we monitored the absorption spectra of DNA-Cu/Ag NCs after they were separately irradiated using intense lasers (5 mW; output at 375 or 475 nm) for different periods (0-180 min). We found that the two bands (centered at 425 and 480 nm) of the NCs bleached at the same rate, which is independent of the irradiation wavelength (Figure S1 in the Supporting Information). The result revealed that the two bands of the Cu/Ag NCs are from one absorbing species.16 TEM-EDS was used to confirm the formation of Cu/Ag NCs. The spectra depicted in Figure S2 in the Supporting Information reveal the presence of Cu and Ag in DNA-Cu/Ag NCs. We further mapped the DNA-Cu/Ag NCs in different areas of the same nickel grid. The inset to Figure S2
in the Supporting Information reveals that the DNA-Cu/Ag NCs in a different area have the uniform bimetal composition (Cu/Ag ) 1/2) instead of a single component of Ag or Cu. The structures of the DNA-Cu/Ag NCs were further determined by ESI-MS measurements. The most dominant peak for the DNA-Cu/Ag NCs occurred at m/z 573.7309 Da, which was assigned to the [DNA s 18H + 11Na + 2Ag + Cu]7- anion, revealing the presence of two Ag and one Cu atoms per DNA strand in this case, which was in good agreement with that of TEM-EDS. Thiol-Induced Fluorescence Quenching of the DNA-Cu/ Ag NCs. Several metal ions, including Cu2+ and Ag+, form complexes with oligonucleotides through the N7 atoms of guanine bases and the N3 atoms of cytosine bases.4a,17 As a result, our Cu/Ag NCs were protected by the DNA template, as displayed in Scheme 1. Thiol compounds (RSH) associated with the Cu/Ag NCs weaken the interaction between the DNA templates and the metal clusters, as depicted in Scheme 1a. Because the NCs were only poorly stabilized by the DNA templates, thiol-induced fluorescence quenching of the DNA-Cu/ Ag NCs occurred. Once thiol compounds such as MPA interacted with the DNA-Cu/Ag NCs, charge transfer occurred between the Cu/Ag NCs and the thiols, leading to partial oxidation of the metal atoms and reduction of the thiols.18 In the presence of Cu2+ ions, the thiols formed Cu-(SR)2 complexes (intermediate), which were further oxidized to form disulfide compounds (final
(15) Test Methods for Evaluating Soild Waste, Physical/Chemical Methods, 3rd ed.; USEPA SW-846, U.S. Government Printing Office: Washington, DC, 1996. (16) Bakr, O. M.; Amendola, V.; Aikens, C. M.; Wenseleers, W.; Li, R.; Dal Negro, L.; Schatz, G. C.; Stellacci, F. Angew. Chem., Int. Ed. 2009, 48, 5921–5926.
(17) Berti, L.; Burley, G. A. Nat. Nanotechnol. 2008, 3, 81–87. (18) (a) Guo, J.-Z.; Cui, H. J. Phys. Chem. C 2007, 111, 12254–12259. (b) Sung, M. M.; Sung, K.; Kim, C. G.; Lee, S. S.; Kim, Y. J. Phys. Chem. B 2000, 104, 2273–2277.
containing 2.5 µM MPA. The mixtures were then used to determine the concentration of Cu2+ ions in the soil sample, using the procedure described above for the analyses of the standard solutions. A water sample collected from a pond on the campus of National Taiwan University was filtered through a 0.2 µm membrane. Aliquots (300 µL) of this pond water were spiked with standard Cu2+ solutions (10 µL; final concentration, 5-200 nM). The spiked samples were then diluted to 450 µL with phosphate solution (20 mM, pH 8.0) containing 2.5 µM MPA and then analyzed separately using both ICPMS and the developed sensing technique.
8568
Analytical Chemistry, Vol. 82, No. 20, October 15, 2010
Figure 1. Fluorescence spectra of the DNA-Cu/Ag NCs in the (a) absence of MPA and Cu2+ ions, (b) presence of 2.5 µM MPA, and (c) presence of 2.5 µM MPA and 150 nM Cu2+ ions. DNA-Cu/Ag NCs (0.01X) were prepared in 20 mM phosphate buffer (pH 8.0); incubation time, 30 min.
products).19 As a result, the strength of the interactions of thiols with the Cu/Ag NCs decreased, and thus the efficiency of thiolinduced fluorescence quenching (see Scheme 1b) was suppressed. We conducted proof-of-concept experiments by measuring the fluorescence of the DNA-Cu/Ag NCs (0.01X) in the absence and presence of MPA and Cu2+ ions. Spectra a and b in Figure 1 reveal that the fluorescence of the DNA-Cu/Ag NCs was quenched by 95% by MPA (2.5 µM) when excited at 480 nm. Figure 1 also reveals that the fluorescence quenched by MPA (2.5 µM) was recovered in the presence of 150 nM Cu2+ ions (spectrum c); the quenching efficiency was 28%. The quenching efficiency is defined as the ratio ((IF0 - IF)/IF0), in which IF and IF0 are the fluorescence intensities of the DNA-Cu/Ag NCs in the presence of MPA (2.5 µM) with/without containing Cu2+ ions, respectively. We point out that the fluorescence of the DNA-Cu/Ag NCs was recovered completely in the presence of 0.5 µM Cu2+ ions. Next, we investigated the role that the nature and concentration of the thiols played in inducing fluorescence quenching of the DNA-Cu/Ag NCs. In addition to MPA [pKa ) 3.7 (CO2H), 10.3 (SH)], we selected 2-ME [pKa ) 9.7 (SH)] and Cys [pKa ) 1.8 (CO2H), 8.3 (SH), 10.8 (NH3+)] for this study. The fluorescence of the DNA-Cu/Ag NCs decreased upon increasing the concentration of each of the thiols. Among the three, MPA has the highest quenching efficiency. To gain more insight into the quenching mechanism, we used the SternsVolmer equation to determine the quenching (IF0/IF) dependence of the quencher concentration. From the Stern-Volmer plots in Figure 2, we estimated the Stern-Volmer quenching constants for MPA, 2-Me, and Cys to be 5.1 × 106, 3.9 × 106, and 2.3 × 106 M-1, respectively. These values revealed that the quenching efficiency decreased in the order MPA > 2-ME > Cys, consistent with the order of the formation constants for the Cu- and Ag-thiol complexes. The conditional formation constants are defined as the formation constants at pH 8.0, which are 11.7, 12.0, and 12.4 for Cu(MPA)22-, Cu(2-ME)2, and (19) (a) Pecci, L.; Montefoschi, C.; Musci, G.; Cavallini, D. Amino Acid 1997, 13, 355–367. (b) Hsu-Kim, H. Environ. Sci. Technol. 2007, 41, 2338–2342. (c) Kachur, A. V.; Koch, C. J.; Biaglow, J. E. Free Radic. Res. 1999, 31, 23–24.
Figure 2. Stern-Volmer plots representing the quenching efficiency of (9) MPA, (b) 2-ME, and (2) Cys on the fluorescence of the DNACu/Ag NCs. Other conditions were the same as those described in Figure 1.
Cu(Cys)2 complexes, respectively.20 The conditional formation constants for Ag(MPA)-, Ag(2-ME), and Ag(Cys) complexes are 16.7, 13.0, and 11.5, respectively.21 Next, we measured the fluorescence lifetimes of the DNA-Cu/Ag NCs in the presence of MPA, 2-ME, and Cys. The fluorescence of the DNA-Cu/Ag NCs in the absence and presence of each thiol decreased over time; the data followed the biexponential fitting equation F(t) ) A1 e(-t/τ1) + A2 e(-t/τ2) allowing us to determine of the values of the lifetimes τ1 and τ2 (5.92 and 1.86 ns, respectively). Electron charge transfer from metal to DNA scaffolds occurred in the metal-DNA complex.22a The decay exhibited biexpoential phenomenon, which is possibly attributed to the differential distribution of complicated luminescent pathways of polynuclear metalsDNA complexes, ligand-to-metal charge transfer (LMCT), or LMCT/ metal centered (MC) triplets.22b The lifetimes of the DNA-Cu/ Ag NCs are listed in Table S1 in the Supporting Information, revealing that they all followed biexponential decays, with slight differences in the values of A1 and A2 under different conditions. After addition of 2.5 µM thiols (e.g., MPA, 2-ME, and Cys) to the DNA-Cu/Ag NCs, the slightly changed fluorescence lifetimes imply that the thiol-induced quenching of the DNACu/Ag NCs occurred through a static quenching mechanism. Cu2+ Ions Dependence of Fluorescence Recovery. It is well-known that small amounts of transition metal ions like Cu2+ catalyze O2-oxidations of thiols (RSH) within several minutes to form disulfide (R-S-S-R′).19,23 Upon an increase in the concentration of Cu2+ ions, the thiol-induced fluorescence quenching of the DNA-Cu/Ag NC solutions decreased as a result of the formation of disulfide compounds (Figure S3A in the Supporting Information). At a constant thiol concentration, the fraction of free thiols decreased upon increasing the Cu2+ concentration. Our results reveal that free thiols, not their (20) Dryden, C.; Gordon, A.; Donat, J. J. Mar. Chem. 2007, 103, 276–288. (21) Bell, R.; Kramer, J. Environ. Toxicol. Chem. 1999, 18, 9–22. (22) (a) Patel, S. A.; Cozzuol, M.; Hales, J. M.; Richards, C. I.; Sartin, M.; Hsiang, J.-C.; Vosch, T.; Perry, J. W.; Dickson, R. M. J. Phys. Chem. C 2009, 113, 20264–20270. (b) Pyykko ¨, P. Angew. Chem., Int. Ed. 2004, 43, 4412–4456. (23) Miller, D. M.; Buettner, G. R.; Aust, S. D. Free Radic. Biol. Med. 1990, 8, 95–108.
Analytical Chemistry, Vol. 82, No. 20, October 15, 2010
8569
oxidized forms nor metal complexes, induced the fluorescence quenching. The fluorescence enhancement factors [(IF s IF0)/ IF0] induced by the Cu2+ ions in solutions containing MPA were larger than those containing the other two thiols; here, IF0 and IF are the fluorescence intensities of the DNA-Cu/Ag NCs in the absence and presence of Cu2+ ions, respectively. We note that, relative to the other two thiols (2-ME and cysteine), the R-S- group of MPA (pKa values of MPA, 2-ME, and cysteine are 10.3, 9.7 and 8.3, respectively) is more nucleophilic toward Cu2+ ions. The oxidation rate of thiols decreases upon a decrease in their basicity.24 We further optimized the MPA concentration for the detection of Cu2+ ions. The fluorescence of the DNA-Cu/Ag NCs was almost quenched (95%) in the presence of 2.5 µM MPA molecules. With the use of 2.5 µM MPA, this probe provided a wider dynamic range of detecting Cu2+ ions. Figure S3B in the Supporting Information displays that the sensitivity was lower at high concentrations of MPA (5 µM); Cu2+ ions at 5 nM were not detected. On the other hand, using low concentrations of MPA, quenching efficiency was lower, leading to higher LODs (4.6 nM). Next, we investigated the role that pH played in determining the sensitivity of the DNA-Cu/Ag NCs/MPA system toward Cu2+ ions. Over the pH range from 3.0 to 11.0, we found that pH 8.0 provided the optimal fluorescence signal (Figure S4 in the Supporting Information). The fluorescence of the DNA-Cu/Ag NCs in the absence of MPA and Cu2+ ions reached a plateau at pH 8.0, too. We recorded CD spectra (Figure 3 and Figure S5 in the Supporting Information) of the solutions at pH 6.0, 8.0, and 10.0, respectively. CD is a common technique for monitoring changes in DNA conformation. Notably, the ellipticity of singlestranded DNA at 265 nm is more negative than that of doublestranded DNA.25 We also point out that there was no CD response over the wavelength range 350-600 nm. At pH 6.0, the ellipticity at 265 nm was more negative in the presence of 15 µM MPA than
those at pH 8.0 or 10.0. The interactions between MPA and the Cu/Ag NCs were weaker at pH 6.0, mainly due to a lesser degree of the dissociation of MPA molecules (pKa 3.7). As a result, the Cu2+-induced recovery of the fluorescence was relatively low at pH 6.0, leading to decreased sensitivity of this sensing system. At high pH values (e.g., 10.0), the CD spectra were similar to those at pH 8.0, revealing that the interactions of the NCs and DNA were similar at the two pH values. However, Cu2+ oxide and hydroxide species became dominant, reducing the reactivity of Cu2+; in addition, the fluorescence of the DNACu/Ag NCs weakened upon increasing the pH to values greater than 8.0.7,26 Evidence of the Interaction between MPA and the Cu/Ag NCs. We employed CD spectroscopy to investigate the interactions between MPA and the Cu/Ag NCs in the DNA-Cu/Ag NCs. Figure 3 displays the CD spectra of DNA and the DNA-Cu/Ag NCs, recorded under various conditions. The ellipticity at 265 nm of the DNA-Cu/Ag NCs (curve b) was less positive than that of the corresponding free DNA (curve a), due primarily to strong interactions between the DNA and the Cu/Ag NCs.27 The CD spectra of MPA-free DNA-Cu/Ag NCs solutions in the presence and absence of Cu2+ ions (18 µM) were similar (curves b and c), suggesting that the Cu2+ ions would not alter the structure of DNA-Cu/Ag NCs. In the presence of 15 µM MPA (curve d), the ellipticity at 265 nm was more positive, revealing that strong interactions existed between MPA molecules and the Cu/Ag NCs. Increasing the MPA concentration from 0 to 30 µM caused the ellipticity to increase (data not shown). In the presence of 30 µM MPA, the ellipticity at 265 nm of the DNACu/Ag NCs was close to that of the corresponding free DNA, revealing the existence of weak interactions between DNA and Cu/Ag atoms. The fluorescence of the DNA-Cu/Ag NCs was quenched by greater than 95% in the presence of 30 µM MPA. These results also reveal that DNA played an important role in determining the fluorescence of the DNA-Cu/Ag NCs. When the NCs were exposed to the bulk solution to a greater extent, the degree of quenching, due to collisions among the particles, increased. In addition, metalsligand charge transfer between the Cu/Ag NCs and DNA decreased, leading to weak fluorescence.22 Interestingly, the ellipticity at 265 nm of the DNACu/Ag NCs containing 15 µM MPA was lower in the presence of 1.2 µM Cu2+ (curve e). Upon an increase in the Cu2+ concentration from 0 to 1.2 µM, the ellipticity decreased, reaching its original value at 1.2 µM (data not shown). These results also support the notion that interactions between MPA and the Cu/Ag NCs in the DNA-Cu/Ag NCs were minimized after adding Cu2+ ions. We recorded XPS spectra of the DNA-Cu/Ag NCs in the presence and absence of MPA to provide further information supporting the existence of interactions between MPA and the Cu/Ag NCs. We assign the peak at the binding energy (BE) of 284.6 eV to the carbon atoms (C 1s) of the DNA molecules.28 The BE for Ag 3d5/2 in the DNA-Cu/Ag NCs in the absence of
(24) Forlano, P.; Olabe, J. A.; Magallanes, J. F.; Blesa, M. A. Can. J. Chem. 1997, 75, 9–13. (25) (a) Vorlı´ckova´, M.; Kejnovska´, I.; Kovanda, J.; Kypr, J. Nucleic Acids Res. 1998, 26, 1509–1514. (b) Oliver, A. W.; Kneale, G. G. Biochem. J. 1999, 339, 525–531. (c) Johnson, N. P.; Baase, W. A.; von Hippel, P. H. Proc. Natl. Acad. Sci. U.S.A 2005, 102, 7169–7133.
(26) Mun ˜oz-Rojas, D.; Subı´as, G.; Fraxedas, J.; Go´mez-Romero, P.; Casan ˜-Pastor, N. J. Phys. Chem. B 2005, 109, 6193–6203. (27) (a) Arakawa, H.; Neault, J. F.; Tajmir-Riahi, H. A. Biophys. J. 2001, 81, 1580–1587. (b) Shemer, G.; Krichevski, O.; Markovich, G.; Molotsky, T.; Lubitz, I.; Kotlyar, A. B. J. Am. Chem. Soc. 2006, 128, 11006–1107. (28) Li, S.; Dang, Z. Appl. Surf. Sci. 2005, 249, 346–353.
Figure 3. CD spectra of solutions containing (a) DNA (3 µM), (b) DNA-Cu/Ag NCs (0.06X), (c) DNA-Cu/Ag NCs (0.06X) and Cu2+ ions (18 µM), (d) DNA-Cu/Ag NCs (0.06X) and MPA (15 µM), and (e) DNACu/Ag NCs (0.06X), MPA (15 µM), and Cu2+ ions (1.2 µM). Other conditions were the same as those described in Figure 1.
8570
Analytical Chemistry, Vol. 82, No. 20, October 15, 2010
Figure 4. (A) Selectivity and (B) sensitivity of the DNA-Cu/Ag NCs/ 2.5 µM MPA probe toward Cu2+ ions over other metal ions. (A) Concentrations of 0.5 and 50 µM for Cu2+ ions and each of the other metal ions, respectively. (B) Error bars represent standard deviations from three repeated measurements. IF0 and IF are the fluorescence intensities of the DNA-Cu/Ag NCs in the absence and presence of Cu2+ ions, respectively. Inset: Linear range of the plot of (IF - IF0)/IF0 against the Cu2+ ion concentration (0-0.2 µM). Other conditions were the same as those described in Figure 1.
MPA was 367.8 eV (Figure S6a in the Supporting Information); it had shifted to lower energy (by ∼0.4 eV) relative to that of pure Ag atoms, revealing the existence of ionic Ag species. The BE for Ag 3d5/2 in the DNA-Cu/Ag NCs in the presence of MPA was 368.4 eV (Figure S6b in the Supporting Information), presumably because of sulfonate groups (alkanethiols are readily oxidized to form sulfonates on Ag surfaces) coordinated to the Cu/Ag NCs.29,30 Selectivity and Sensitivity. Under the optimal conditions, we tested the selectivity of the DNA-Cu/Ag NCs (0.01 X)/2.5 µM MPA probe toward Cu2+ ions (0.5 µM), relative to other metal ions (Zn2+, Cd2+, Pb2+, Mg2+, Mn2+, Sr2+, Hg2+, Ni2+, Co2+, Ca2+, Ag+, Fe3+, Al3+, and Cr3+; each 50 µM). Figure 4A reveals that our probes responded selectively toward Cu2+ ions over the other tested metal ions by factors of at least 2300-fold. Although Fe3+ ions can also catalyze the oxidation of the thiols, they did not cause interference in determining Cu2+ ions, mainly (29) Gole, A.; Sainkar, S. R.; Sastry, M. Chem. Mater. 2000, 12, 1234–1239. (30) Schoenfisch, M. H.; Pemberton, J. E. J. Am. Chem. Soc. 1998, 120, 4502– 4513.
because of the formation of Fe3+-phosphate species (e.g., Fe3(PO4)2 · 8H2O, pKsp 36.0, or FePO4 · 2H2O, pKsp 26.4).23,31 Compared with our previous DNA-Ag NC probe for Cu2+ ions, the present probe is more selective toward Cu2+ over the tested interference ions.7 The fluorescence signals from solutions of DNA-Cu/Ag NCs, MPA, and Cu2+ ions were almost unaffected by the presence of interfering metal ions (Na+, K+, Mg2+, Ca2+, Zn2+, Pb2+, Cd2+, Ni2+, Cr2+, and Fe3+; 0.3 µM each). The selectivity of our system is much better than those of other sensors for Cu2+ ions.11,12,32 Figure 4B reveals that the fluorescence enhancement factors [(IF - IF0)/IF0] induced by Cu2+ ions increased upon increasing the concentration of Cu2+ ions up to 500 nM. A linear correlation (R2 ) 0.988) existed between the value of [(IF - IF0)/IF0] and the concentration of Cu2+ ions over the range 5-200 nM. The limit of detection (LOD) for Cu2+ ions, at a signal-to-noise ratio of 3, was 2.7 nM; this value is lower than those obtained when using complicated fluorescent chemosensors, DNA-Ag NCs, or GSH-capped fluorescent Au nanoparticles.7,11,12,32 Detection of Cu2+ Ions in Soil and Pond Water Samples. To test the practicality of our developed approach, we used the DNA-Cu/Ag NCs/2.5 µM MPA probe to determine the concentrations of Cu2+ ions in samples of Montana soil (SRM 2710) and pond water. By application of a standard addition method, the present approach provided a linear response to Cu2+ ions in spiked samples at concentrations over the range 0-125 nM (y ) 5.78x + 0.25; R2 ) 0.976). The concentrations of Cu2+ ions in the Montana Soil sample (n ) 5) detected using this new approach and ICPMS were 2.60 (±0.23) and 2.79 (±0.11) mg/ g, respectively. The F-test value for comparison of the two methods was 4.37 (the F-test value is 6.39 at 95% confidence), revealing that our new technology and the ICPMS-based approach do not differ significantly in their precision. With the use of a t-test (the Student t value is 2.776 at 95% confidence), the interval for Cu2+ ions was 2.23-2.97 mg/g, which is not significantly different from the true value (the mean value, provided by NIST, for Cu2+ ions in the soil sample was 2.95 mg/g). We obtained a linear correlation (R2 ) 0.996) between the response and the concentration of Cu2+ ions spiked into the pond water over the range 5-200 nM. Neither our sensor nor the ICPMS-based approach detected the presence of Cu2+ ions in this pond water sample. The DNA-Cu/Ag NC probe provided recoveries of 96-117%. These results reveal the practicality of using this probe comprising DNA-Cu/Ag NCs and MPA for the determination of the concentrations of Cu2+ ions in environmental samples. CONCLUSIONS We have developed a simple, homogeneous fluorescence assay for the detection of Cu2+ ions using a combination of watersoluble DNA-Cu/Ag NCs and MPA. The remarkable sensitivity (LOD, 2.7 nM) and selectivity (2300-fold over other tested metal ions) of this sensing probe result from three facts: (1) the high quantum yield (47.8%) of the DNA-Cu/Ag NCs, (2) the efficient (31) Harris, D. C. Exploring Chemical Analysis, 3rd ed.; W. H. Freeman and Company: New York, 2005. (32) (a) Wu, Q.; Anslyn, E. J. Am. Chem. Soc. 2004, 126, 14682–14998. (b) Weng, Y.-Q.; Yue, F.; Zhong, Y.-R.; Ye, B.-H. Inorg. Chem. 2007, 46, 7749– 7755.
Analytical Chemistry, Vol. 82, No. 20, October 15, 2010
8571
fluorescence quenching ability of MPA, and (3) the highly catalytic activity of Cu2+ ions toward the oxidation of MPA. Analyses of soil and pond water samples revealed that this new approach allows the simple and sensitive detection of Cu2+ ions in real environmental samples. SUPPORTING INFORMATION AVAILABLE Absorption spectra of DNA-Cu/Ag NCs after laser irradiation under 475 and 375 nm for different periods of time (Figure S1); TEM-EDS results of the as-prepared DNA-Cu/Ag NCs and mapping the composition of Cu and Ag element in the same nickel grid of three different areas (Figure S2); fluorescence ratios of the DNA-Cu/Ag NCs in the presence of Cu2+ ions and MPA, 2-ME, or Cys (Figure S3); effect of pH on the fluorescence intensity of the DNA-Cu/Ag NCs/2.5 µM MPA probe in the absence and presence of Cu2+ ions (0.3 µM) (Figure S4); CD spectra of solutions at pH 6.0 and 10.0 at different concentrations of Cu2+ ions and MPA (Figure S5); X-ray photoelectron spectra of the Ag 3d core level of the DNA-Cu/Ag NCs in the
8572
Analytical Chemistry, Vol. 82, No. 20, October 15, 2010
absence and presence of MPA (Figure S6); and fluorescence lifetimes of the DNA-Cu/Ag NCs (0.01X) in the absence and presence of 2.5 µM MPA, 2-ME, and Cys (A1 and A2 are the relative amplitude of τ1 and τ2) (Table S1). This material is available free of charge via the Internet at http://pubs.acs.org. ACKNOWLEDGMENT This study was supported by the National Science Council of Taiwan under Contracts NSC 98-2113-M-002-011-MY3, NSC 98-2627-M-002-013, and NSC 98-2627-M-002-014. We thank to Prof. J.-R. Yang and Mr. H. -W. Yen (NTU Department of Materials Science and Engineering) for conducting the TEMEDS measurements.
Received for review June 24, 2010. Accepted August 29, 2010. AC101659D