Multiplex Plasmonic Sensor for Detection of Different Metal Ions Based

Jan 21, 2013 - In this paper, a label-free multiplex plasmonic sensor has been developed to selectively determine different metal ions including Fe3+,...
5 downloads 11 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Multiplex Plasmonic Sensor for Detection of Different Metal Ions Based on a Single Type of Gold Nanorod Haowen Huang,* Shenna Chen, Fang Liu, Qian Zhao, Bo Liao, Shoujun Yi, and Yunlong Zeng Laboratory of Theoretical Chemistry and Molecular Simulation of Ministry of Education. School of Chemistry and Chemical Engineering, Hunan University of Science and Technology, Xiangtan, P. R. China S Supporting Information *

ABSTRACT: In this paper, a label-free multiplex plasmonic sensor has been developed to selectively determine different metal ions including Fe3+, Hg2+, Cu2+, and Ag+ ions based on a single type of gold nanorod (GNR). Under proper conditions, these metal ions can react with GNRs, resulting in changes of nanostructure and composition. The determination of Fe3+, Hg2+, Cu2+, and Ag+ ions is therefore readily implemented due to changes of longitudinal plasmon wavelength (LPW) of nanorods. Moreover, the GNRbased assay can not only determine all four kinds of metal ions successively but also can detect which of any one or several kinds of metal ions. This assay is sensitive to detect Fe3+, Hg2+, Cu2+, and Ag+ as low as 10−6, 10−8, 10−10, and 10−8 M, respectively. Importantly, the special nanostructure and composition of the nanorods are induced by these metal ions, which allow this sensor to maintain high selectivity to determine these metal ions. This nanosensor abrogates the need for complicated chemosensors or sophisticated equipment, providing a simple and highly selective detection platform.

G

spectrometry,25,26 inductively coupled plasma mass spectrometry,27,28 electrochemical methods,29,30 and so on, nanoparticlebased methods neither require sophisticated equipment nor complicated operations. In this paper, we developed a multiplex plasmonic sensor to selectively determine different metal ions based on single type of GNR. Under proper experimental conditions, the metal ions such as Fe3+, Cu2+, Ag+, and Hg2+ react with GNRs, resulting in the changes of nanostructure and composition. The determination of these different metal ions is therefore readily implemented along with changes of longitudinal plasmon wavelength (LPW) of nanorods, as shown in Scheme 1. Moreover, the GNR-based assay can not only determine all four kinds of metal ions successively but also can detect which of any one or several kinds of these metal ions with high selectivity. This assay abrogates the need for complicated chemosensors or sophisticated equipment, providing a simple and highly selective detection platform.

old nanorods (GNRs) are elongated gold nanoparticles with unique optical properties depending on their size and shape.1−3 GNRs are sensitive to the dielectric constant of the surrounding medium owing to localized surface plasmon resonance (LSPR).4−6 A slight change of the local refractive index around GNRs will result in an observable shift of plasmon resonance frequency. These properties make GNRs specifically suitable for biological sensing applications.7−11 On the other hand, GNR-based sensors have been developed for sensitive detection of heavy-metal ions owing to the optical, electronic, and catalytic ability of GNRs.12−14 Campiglia and co-works took advantage of the strong affinity between Au and Hg to shorten the aspect ratio of GNR and then sensitively determined the mercury in water samples.15 Cu2+ was detected by the aggregation of GNRs along with a rapid color change.16 Also, the mesporous silica coated gold nanorod might probe Hg2+.17 Ever-increasing industrial development leads to increasing the possibility for a release of pollutants into the environment. Some metal ions may be toxic pollutants that are nonbiodegradable, undergo transformations, and have great environmental, public health, and economic impacts.18−20 These contaminants adversely affect the environment and the ability to bioaccumulate in organisms, even the excess necessary metal ions are dangerous to the entire ecosystem.21,22 Thus, determination of toxic metal ions is an important issue both in environmental monitoring and clinical research. Approaches of determining metal ions constantly grow attention. Compared to the techniques usually adopted for detection of metal ions including atomic absorption spectrometry,23,24 atomic emission © 2013 American Chemical Society

■

MATERIALS AND METHODS

Materials. HAuCl4·3H2O, cetyltrimethylammonium bromide (CTAB), Na2S2O3, ascorbic acid, and silver nitrate were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All of the chemicals, unless mentioned Received: November 14, 2012 Accepted: January 21, 2013 Published: January 21, 2013 2312

dx.doi.org/10.1021/ac303305j | Anal. Chem. 2013, 85, 2312−2319

Analytical Chemistry

Article

Scheme 1. Schematic Illustration of Determination of Fe3+, Ag+, Cu2+, and Hg2+ Based on a Single Type of GNR

reaction was allowed to proceed for another 30 min at room temperature before the vis−near infrared (NIR) absorption measurement. The concentration of Fe3+ was quantified by the changes of longitudinal plasmon wavelength. A series of different concentration of Ag+ solution (1 × 10−3, 10−4, 10−5, 10−6, 10−7, 10−8 M) were detected based on the resulting GNR dispersion after determination of Fe3+. A 0.1 mL portion of Ag+ was then mixed with 10 mL of the resulting GNR dispersion after determination of Fe3+. To the mixture, 50 μL of 0.1 M Na2S2O3 was added and allowed to react for 30 min at room temperature before the vis−NIR absorption measurement. Similarly, various concentrations of Cu2+ solution were measured based on the resulting GNR dispersion after determination of Ag+. A 0.1 mL portion of Cu2+ was then mixed with 10 mL of the resulting GNR dispersion after determination of Ag+. To the mixture, another 50 μL of 0.1 M Na2S2O3 was added and allowed to react for 30 min at room temperature before the vis−NIR absorption measurement. Lastly, the detection of Hg was performed based on previous nanorods after determination of Cu2+. Various concentrations of HgCl2 solution were used to test. A 0.1 mL portion of standard Hg2+ solution was added to 10 mL of nanorod dispersion, and then, 50 μL of 0.01 M NaBH4 was added to the mixture. Because the plasmon band of the resulting nanoparticles red-shifted with increasing reaction time, various stages were monitored at different reaction time intervals, and the final spectrum was recorded as the signal to detect the Hg2+ concentration. In this work for determination of the metal ions including Fe3+, Ag+, Cu2+, and Hg2+ ions, three replicated samples of the same metal ion were used for each test, and the average value was used to show the metal ion. Characterization. Transmission electron microscopy (TEM) was performed on a JEM-2010 transmission electron microscope at 80 kV. Absorption spectra of the GNRs and core−shell GNR dispersions were measured using Lambda 35 (PerkinElmer, USA).

otherwise, were of analytical reagent grade and used as received. The aqueous solutions were prepared in doubly distilled water. Preparation of GNRs without AgNO3. The GNRs were prepared according to the seeded growth method without the addition of AgNO3 as reported previously.31 A 20 mL aqueous solution containing 2.5 × 10−4 M HAuCl4, and 2.5 × 10−4 M trisodium citrate was prepared in a conical flask. Next, 0.6 mL of ice cold 0.1 M NaBH4 solution was added to the solution all at once while stirring. The solution turned pink immediately after adding NaBH4, indicating particle formation. The particles in this solution were used as seeds within 2−5 h after preparation. To prepare GNRs, 1.5 mL of 0.02 M HAuCl4 was added to 30 mL of 0.1 M CTAB, followed by the addition of 0.8 mL 0.08 M ascorbic acid. Ascorbic acid here worked as a mild reducing agent and changed the color of the solution from dark yellow to colorless. Finally, 70 μL of the seed solution was added to the solution and the color changed gradually in 30 min. Preparation of GNRs with AgNO3. The seed solution of the GNRs was prepared according to the technique reported previously.32,33 The CTAB solution (1.5 mL, 0.1 M) was mixed with 100 μL of 0.02 M HAuCl4, and 100 μL of ice cold 0.01 M NaBH4 was added to the solution to form a brownish yellow solution. Vigorous stirring was continued for 2 min, and then, the seed solution was kept at room temperature (25 °C) and used at least 2 h after preparation. To synthesize the GNRs, 1.5 mL of 0.02 M HAuCl4 and 1.0 mL of 0.01 M AgNO3 were added to 30 mL of 0.1 M CTAB, followed by addition of 0.8 mL of 0.08 M ascorbic acid. Ascorbic acid served as a mild reducing agent and changed the solution from dark yellow to colorless. Afterward, 70 μL of the seed solution was added and the color of the solution changed gradually within 15 min. Modification of Preparation of GNRs. The approach of preparation of GNRs was adopted based on the early report.34,35 Briefly, to synthesize the GNRs, 1.5 mL of 0.02 M HAuCl4 and 1.0 mL was added to 30 mL of 0.1 M CTAB, followed by addition of 0.8 mL of 0.08 M ascorbic acid and changing the solution from dark yellow to colorless. Afterward, 200 μL of the GNRs prepared with AgNO3 as seed solution was added and the color of the solution changed gradually within 20 min. Determination of Fe3+, Ag+, Cu2+, and Hg2+. A 0.1 mL portion of Fe3+ standard solution or sample solution was added to 10 mL of GNR dispersion prepared with modification. The

■

RESULTS AND DISCUSSION GNR Preparation. GNRs were first prepared using the silver ion-assisted seed-mediated method in the presence of CTAB based on the previous method.32,33 In this protocol, the addition of AgNO3 provides some advantages by allowing better control of the shape of the synthesized GNRs. In this study, the GNR was designed to determine Fe3+, Ag+, Cu2+, and Hg2+ ions. Obviously, the Ag+ ion can not be directly detected by this type of GNR because the Ag has been uniformly distributed in the synthesized GNRs, which can not distinguish the elemental Ag to be measured originated from the selfcontaining of GNRs or exotic target Ag+ in the medium. Therefore, another approach was adopted to prepare GNRs without AgNO3 based on previous report.31 However, the poor qualified GNRs contained plenty of spherical nanoparticles, and polydispersion nanorods could not meet the requirement of experiments in this study. A modification of preparing GNRs was thereby employed.34,35 First, the GNRs were synthesized with AgNO3. Then the as-prepared GNRs as seed to synthesize GNR@Au by depositing Au on the as-prepared nanorods in the absence of Ag+, in which AuCl4− ions would be complexed to CTAB micelles and get reduced by ascorbic acid on the surface of GNRs, along with a blue-shift of LPW from 887 to 790 nm illustrated in Figure 1, and the TEM images are shown in 2313

dx.doi.org/10.1021/ac303305j | Anal. Chem. 2013, 85, 2312−2319

Analytical Chemistry

Article

Figure 1. Absorption spectrum (b) obtained by a uniform layer of Au depositing on the surface of the original GNRs (a).

Figure 3. Absorption spectra acquired from the GNRs reacting with FeCl3 at time durations (10 min).

Figure 2a and b. The prepared nanorods get thicker due to a uniform layer of Au depositing on the surface of the original

longitudinal band may completely disappear. This is also characterized by TEM images shown in Supporting Information Figure S1, illustrating the shortened GNRs obtained by FeCl3 reacting with GNRs. The LPW decreases with the increase of Fe3+ concentration, implying that this method may quantitatively determine Fe 3+ ions. A series of FeCl 3 concentrations were measured under proper experimental conditions. Quantitative calculation gives a linear equation between the Fe3+ concentration within the range of 0.1−40 × 10−5 M and the degree of blue-shift of LPW: Δλ = −21.96 + 47.91 × 104c and the linear regression coefficient is 0.9960. Where, Δλ is the degree of blue-shift of LPW of nanorods and c represents the concentration of Fe3+ ion. The calibration curve is shown in Supporting Information Figure S2, and the results show that Fe3+ concentrations as small as 10−6 M can be detected by this method. During the process of shortening GNRs, FeCl3 serves as an oxidation agent which oxidizes the GNRs. Solutions of other compounds containing Fe3+ such as Fe(NO3)3 can also shorten the GNRs in the experiments. Reaction 1 presents the chemical reaction involved in the observed phenomenon. Fe3 + + Au 0 + 2Cl− = AuCl 2− + Fe 2 +

(1)

which can take place in the presence of CTAB because of a change in the reduction potential of Au(0) upon complexation,36 leading to gradual oxidation of GNRs. A higher concentration of CTAB is required for GNR oxidation. GNRs separated from CTAB by centrifugation redisperse in water which exhibit no blue-shift of the LPW. However, these GNRs dispersed in 0.1 M CTAB again show the same oxidation behavior as as-synthesized GNRs. The sensing performance for metal ions is strongly influenced by the assay condition such as GNR concentration. The number of nanorods would influence the sensitivity for each species. The particle concentration being too high or too low may lead to a serious imbalance between the nanorods and metal ions, thus decreasing the sensitivity of this assay. In this study, repeated experiments require larger amounts of GNRs, which are obtained by different batches of the GNRs prepared under the same condition, showing slight change in size and number of GNRs. The absorption intensity seems more convenient and accurate compared with the concentration of GNRs in these experiments. Therefore, the absorption intensity of the GNRs, especially in the quantitative analysis, is selected to be in the range of 0.3−0.5 in the subsequent experiments.

Figure 2. TEM images of GNRs and corresponding core−shell nanorods. (a) As-synthesized GNRs. (b) Decreased aspect ratio of GNRs obtained by a uniform layer of Au depositing on the surface of the as-synthesized GNRs. (c) Shortened GNRs acquired from b reacting with FeCl3. (d) GNR@Au2S/AuAgS/Ag3AuS2 obtained by c reacting with Na2S2O3. (e) Core−shell GNRs acquired by GNR@ Au2S/AuAgS/Ag3AuS2 reacting with Cu(NO3)2 and Na2S2O3. (f) Result of Hg2+ reacting with CuAgAu−composite GNRs.

GNRs, leading to the decrease of aspect ratio of the resultant nanorods. Clearly, LSPR spectra and the TEM images illustrate that GNR@Au have the similar characteristics as GNRs, confirming the validity of the protocol. The GNR@Au is still termed as GNR in the following discussion. Determination of Fe3+ Based on the Decrease of Aspect Ratio of GNRs. When FeCl3 was added to the GNR dispersion, real-time vis−NIR absorption spectroscopy was monitored as displayed in Figure 3, which reveals that the LPW exhibits a gradual blue-shift while the transverse plasmon wavelength (TPW) stays at around 520 nm and a decrease in intensity of the absorption spectra. Finally, the absorbance of 2314

dx.doi.org/10.1021/ac303305j | Anal. Chem. 2013, 85, 2312−2319

Analytical Chemistry

Article

Determination of Ag+ Ion Based on the Formation of Core−Shell Nanorods. After accomplishment of determination of Fe3+, the detection of Ag+ based on GNRs was investigated. With the addition of Ag+ and Na2S2O3 to the shortened GNR dispersion, a gradual red-shift of LPW occurred with the reaction time, as shown in Figure 4. The

In the case of detection of Ag+ concentration, our results show that Ag+ concentration as small as 10−8 M can be detected by this method. Measurement of Cu2+ Ion in the Solution. With the completion of detecting Ag+, the addition of Cu(NO3)2 solution to the produced GNR@Au 2 S/AuAgS/Ag 3 AuS 2 dispersion in the presence of Na2S2O3 led to a distinct blueshift instead of red-shift of LPW, as also shown in Figure 4. Apparently, the contrary tendency of LPW implies the nanostructure of nanorod is varied with Cu2+ reacting with GNR@Au2S/AuAgS/Ag3AuS2 and Na2S2O3. The TEM image shown in Figure 2e indicates that the new generated nanorods are still core−shell nanorods. For convenience, the produced core−shell nanorod is termed as CuAgAu−composite GNR in the following discussion. Further investigation demonstrates that the Cu2+ ion can induce the change of nanostructure of the resultant nanorods in the presence of Na2S2O3. With the progress of the reaction, the thickness of the shell of CuAgAu−composite GNR is increased; however, the nanorod core is shortened, demonstrated by a TEM image shown in the Supporting Information (Figure S4). Obviously, the Cu2+ can etch the core of core−shell GNR in the presence of Na2S2O3, leading to a decrease of aspect ratio of the nanorod and, at the same time, a significant blue-shift of LPW. The degree of blue-shift of LPW is directly related to the amounts of Cu2+ ions in the aqueous solution. Various concentrations of Cu(NO3)2 were investigated, and the corresponding LSPR spectra are shown in Figure 6a. Meanwhile, Figure 6b shows a relationship between the degree of blue-shift of LPW and logarithm of Cu2+ concentrations with the regression equation Δλ = 140.6 + 13.57log c (regression coefficient R = 0.9879), which indicates that the Cu2+ ion can be quantitatively detected within the range of 1.0 × 10−10 to 1.0 × 10−5 M. Trace concentrations as low as 1 × 10−10 M of Cu2+ can be detected in the optimized experimental conditions. Detection of Hg2+ Ion after the Measurement of Cu2+ Ion. With the addition of Hg(NO3)2 and NaBH4 to CuAgAu− composite GNR dispersion, another red-shift of LPW will occur, as shown in Figure 7. All the spectra were recorded after 5 min of Hg(II) addition, i.e., the minimum time at which a plateau was reached, and no further spectral shifts were observed. A control experiment was carried out by adding the same amount of NaBH4 to same volume of CuAgAu− composite GNR dispersion, and no detectable change of

Figure 4. Absorption spectra acquired from the GNRs reacting with AgNO3 at first (solid lines) and then reacting with Cu(NO3)2 in the presence of Na2S2O3 for different time durations (dash lines). The black curve represents as-synthesized GNRs.

TEM image present in Figure 2d illustrates that core−shell nanorods, GNR@Au2S/AuAgS/Ag3AuS2, are produced after the GNRs reacting with Ag+ and Na2S2O3, inducing the formation of a layer of shell, Au2S/AuAgS/Ag3AuS2, around the GNR.37,38 The metal chalcogenide layer is a narrow band gap semiconductor with a large refractive index at typical optical frequencies. As a result, red-shifts in the plasmon peak occur from the larger refractive index chalcogenide layer covering the GNR. To verify the red-shift of LPW is a result of GNRs reacting with Ag+ in the presence of Na2S2O3, a control experiment without addition of Ag+ was performed as displayed in Figure 5. The experimental results indicate that the addition of Na2S2O3 to the GNR dispersion leads to a decrease of absorption intensity of the plasmon band and no clear red-shift of LPW was observed, demonstrating that the significant redshift of LPW resulted from Ag+ reacting with GNRs in the presence of Na2S2O3. In addition, a series of concentrations of standard Ag+ solution was analyzed using this method shown in Supporting Information Figure S3, suggesting that the Ag+ ions might be quantitatively analyzed utilizing this GNR-based assay.

Figure 5. Comparison of GNRs reacting with Na2S2O3 in the absence (a) and presence (b) of AgNO3. The black curves in parts a and b represent the as-synthesized GNRs. 2315

dx.doi.org/10.1021/ac303305j | Anal. Chem. 2013, 85, 2312−2319

Analytical Chemistry

Article

Figure 6. Absorption spectra show the spectral shift at several Cu2+ concentrations (a) and the arrow indicates the increase of Cu2+ concentrations. The concentration range between 1.00 × 10−10 and 1.00 × 10−6 M shows the spectra within the linear dynamic range of the calibration curve (b).

Figure 7. Absorption spectra show the spectral shift at several Hg2+ concentrations reacting with CuAgAu−composite GNR, the black line represents the as-prepared GNRs, and the dashed line represents Au2S/AuAgS/Ag3AuS2 (a). The concentration range between 1.00 × 10−8 and 1.56 × 10−4 shows the spectra within the linear dynamic range of the calibration curve (b).

shell nanorod GNR@Au2S/AuAgS/Ag3AuS2, subsequently, the Cu2+ may be detected due to GNR@Au2S/AuAgS/Ag3AuS2 nanorods transforming to another core−shell nanorod of CuAgAu−composite. At last, the water-dissolved Hg(II) can be further analyzed by Hg(II) reacting with CuAgAu-composite GNR in the presence of NaBH4. The evolution of the changes of GNR nanostructure is well characterized by Figure 2. It suggests that the four kinds of metal ions may be successively determined. The selectivity of this assay was investigated. For the determination of Fe3+ ion, the accomplishment of analysis comes from the oxidation of GNRs. Accordingly, the nonoxidizing metal ions do not interfere with the detection of Fe3+ with the exception of Cr(VI) in the aqueous solution. Ag+ is evaluated by comparing the responses to other metal ions such as Cu2+, Co2+, Ni2+, Pb2+, Hg2+ Cd2+, Zn2+, Fe3+, and Al3+ at the concentration of 10−3 M. No noticeable plasmon peak wavelength change is found at this concentration. The other ions investigated here hardly show any effects on the plasmon peak wavelength implying that most of these metal ions cannot form complex with Au in the shell layer around the GNRs.38 Thereafter, Cu2+ is evaluated by comparing the responses to other metal ions such as Ag+, Co2+, Ni2+, Pb2+, Hg2+, Cd2+, Zn2+, Fe3+, Ca2+, Mn2+, and Al3+ at a concentration of 10−3 M, as shown in Supporting Information Figure S5.

LPW occurred. This suggests that the red-shift of LPW results from the Hg2+ reacting with CuAgAu−composite GNR in the presence of the NaBH4. This strong reducing agent is also capable of reducing any oxidized form of Hg into Hg(0). Therefore, this approach may detect water-dissolved Hg including inorganic and organic mercury. It is possible that the Hg(0) was first reduced from Hg(II) by the NaBH4 and then another composite is formed between Hg(0) and CuAgAu−composite GNR. A new layer of shell with higher refractive index is thus produced around the nanorod and the LPW moves to longer wavelength. The difference between maximum and the starting wavelength of the longitudinal band shows a direct correlation with Hg(II) concentration. Figure 7b shows the relationship between the degree of red-shift of LPWs and Hg(II) concentrations. There is a linear relationship within the range of 1.00 × 10−8 to 1.56 × 10−4 M of Hg2+ solution, and the regress equation is Δλ = 34.58 + 4.397log c with the related coefficient R = 0.9968. Where Δλ is the difference between redshift of LPW and the starting LPW of nanorod and c represents the Hg2+ concentration. A travel level of HgCl2 as low as 1.00 × 10−8 M can be detected in this assay. As the above discussion, Fe3+ ion can be first measured accompanied by the decrease of aspect ratio of GNRs and then the Ag+ ion can be determined based on the formation of core− 2316

dx.doi.org/10.1021/ac303305j | Anal. Chem. 2013, 85, 2312−2319

Analytical Chemistry

Article

Au2S/AuAgS/Ag3AuS2 nanorods to produce another core− shell nanorod of CuAgAu−composite, resulting in a significant blue-shift. In this way, the separated reactions can be carried out readily and reliable LSPR spectra can be obtained in the process of determination of Ag+, Cu2+, and Hg2+. After GNRs reacting with Fe3+, Ag+, Cu2+, and Hg2+, a heteronanostructure of core−shell nanorod was produced. The scanning transmission electron microscope (STEM) is a very powerful and highly versatile instrument capable of atomic resolution imaging and nanoscale analysis. Further validation of elemental analysis is demonstrated by STEM-EDAX mapping, and the Hg, Cu, Ag, and S signals are shown in Figure 9,

Similarly, no interference metal ions are observed. Lastly, the investigation of Hg2+ illustrates that there are no interference ions in the ordinarily metal ions such as Ag+, Cu2+, Co2+, Ni2+, Pb2+, Hg2+, Cd2+, Zn2+, Fe3+, Mn2+, Mg2+, and Al3+, as illustrated in Supporting Information Figure S6. In fact, the particular nanostructure and composition of the formed nanorods are the results of the GNRs reacting with these metal ions, respectively, which allow this sensor only responding to these special metal ions in the analytical procedure. A highly selective assay is therefore established to successively detect the Fe3+, Ag+, Cu2+, and Hg2+. A sample containing Fe3+, Ag+, Cu2+, and Hg2+ was tested to validate this assay. The sample was first added to the GNR dispersion and a gradual blue-shift of LPW occurred until it reached the minimal wavelength. Afterward, a proper amount of Na2S2O3 was introduced into the system with the final concentration maintained 10−4 M. The wavelength of longitudinal plasmon band increased with the reaction time and approached the max wavelength. Shortly, a contrary change, blue-shift of LPW, appeared again resulting from the nanorods reacting with Cu2+. When no blue-shift further occurred, NaBH4 solution was added to the mixture and kept the final concentration at 10−5 M. A significant red-shift of LPW occurred, and the degree of red-shift was determined by the amounts of Hg2+. In this way, Fe3+, Ag+, Cu2+, and Hg2+ may be successively detected with high selectivity, as shown in Figure 8.

Figure 9. TEM and STEM images as well as EDAX elemental mapping of core−shell nanorods: Cu−K map, Hg−L map, Ag−L map, and S−K map images taken simultaneously with an EDX map.

respectively. The elemental Hg, Cu, and Ag in core−shell nanorods are the results of these metal ions reacting with nanorods to form heteronanostructure, and the elemental S is introduced by the addition of Na2S2O3 during the determination of the elemental Ag and Cu. There is no Fe in the core− shell nanorods because the Fe3+ was reduced to Fe2+ dissolved in solution during the chemical reaction process. Determination of Any One or Several Kinds of Metal Ions. As discussed aforementioned, the four kinds of metal ions can be successively detected using this assay when all these metal ions exist in the test sample. In fact, any one or several kinds of these metal ions contained in the sample can also be determined using this assay. The experimental results show the absorption spectra acquired by one kind of metal ion are similar to the spectra of the successive determination with the exception of the Hg2+ shown in Supporting Information Figure S7. In the determination of single component Hg, blue-shift rather than red-shift of LPW occurred, induced by Hg2+. The degree of blue-shift of LPW is proportional to the amount of Hg2+, and quantitative analysis can also be performed. The blue-shift of LPW is attributed to the decrease of aspect ratio of the resulting GNR.15 On the other hand, although the single component Cu2+ can also lead to the occurrence of blue-shift of LPW, the related mechanism is different from the successive detection, which will be explored in the future study. If there are two kinds of metal ions in the sample, these metal ions can also be determined respectively. Figure 10 displays the spectra obtained by two kinds of metal ions. When there are three kinds of metal ions in the sample, the corresponding LSPR spectra are similar to Figure 8. Interestingly, it can be seen from these figures that the determination of Hg2+ in different samples which may contain one, two, or three kinds of metal ions exhibit respective different change trends of their LSPR spectra. If the sample

Figure 8. LSPR spectra obtained by the GNRs reacting with Fe3+, Ag+, Cu2+, and Hg2+, respectively.

In the above procedure, blue-shift of LPW induced by Cu2+ will immediately appear after a red-shift resulted from Ag+ in the presence of Na2S2O3. The rapid switch of LSPR spectra from red-shift to blue-shift brings inconvenience to determine the concentration of Ag+. To obtain reliable spectra for reference purposes, we slightly change the experimental procedure. After the completion of determination of Fe3+, a chelating agent ethylene diamine tetraacetic acid (EDTA) was added to the sample, leading to the formation of stable complexes of EDTA−Cu(II) and EDTA−Hg(II). With the addition of Na2S2O3, a type of core−shell nanorod, GNR@ Au2S/AuAgS/Ag3AuS2, will be formed in the presence of Ag+ ion, leading to the longitudinal plasmon peak gradually moving to the longer wavelength to reach at maximum and stable. However, upon the addition of Fe3+ ion, the wavelength of longitudinal plasmon band shifts back until the minimal wavelength appears. As is well-known, Fe3+ can react with EDTA to form more stable complex; thus, the completely replaced Cu2+ by Fe3+ will continually react with the GNR@ 2317

dx.doi.org/10.1021/ac303305j | Anal. Chem. 2013, 85, 2312−2319

Analytical Chemistry

Article

Figure 10. Absorption spectra obtained from two kinds of metal ions in the sample determined by a single type of GNR.

contains only one component Hg2+ ion, a blue-shift of LPW will occur along with the decrease of absorption intensity. The blue-shift, however, will be replaced by red-shift of LPW if the Cu2+ exists in the sample. When the Hg2+, Cu2+, and Ag+ ions coexist in the sample, the determination of Hg2+ will be accomplished by the appearance of blue-shift of plasmonic band, accompanied by the increase of absorption intensity of transverse and longitudinal bands. These suggest that the change of plasmonic properties results from the formation of different nanostructure and composition of the generated nanorods.

GNR in the following experiments. When FeCl3 was added to the GNR dispersion in the presence of CTAB, the aspect ratio of the nanorods was gradually decreased with the progress of reaction between GNRs and FeCl3. TEM images shown in Figure S1 illustrate that starting GNRs exhibit a narrow size distribution and GNRs decrease in length in the process of reaction and then become nanospheres. With the addition of Ag+ to the GNRs in the presence of Na2S2O3, GNRs transformed to core−shell nanorod of GNR@Au2S/AuAgS/ Ag3AuS2. Interestingly, the core−shell nanorods can react with Cu2+ and Na2S2O3, resulting in a clear LPW blue-shift, suggesting a change occurred in the nanostructure. Compared to the precursor GNR@Au2S/AuAgS/Ag3AuS2, the shell of the resulting nanorods is clearly changed, and the nanorod core is obviously shortened as shown in Figure S4. With the progress of the reaction, a considerable part of the core disappeared, as shown in Figure 4c. Despite the increase of the shell being a result of the previous core−shell nanorods reacting with Cu2+ and Na2S2O3, the formed shell seems still homostructure. The change of nanostructure of these nanorods demonstrate that the reagents can not only react with shell but also further pass through this compact shell and then subsequently react with the inner GNR core. This material is available free of charge via the Internet at http://pubs.acs.org.

■

CONCLUSIONS In summary, we have successfully developed a multiplex assay to determine different metal ions based on single type of nanoparticle. Under proper conditions, Fe3+, Hg2+, Cu2+, and Ag+ can react with GNRs, resulting in the changes of nanostructure and composition. The four kinds of metal ions are therefore determined with cyclic changes of LPW of nanorods. Moreover, the GNR-based assay can not only determine all four kinds of metal ions successively but also can detect which of any one or several kinds of metal ions. This GRN-based assay is sensitive to detect Fe3+, Hg2+, Cu2+, and Ag+ as low as 10−6, 10−8, 10−10, and 10−8 M, respectively. In addition, this nanosensor maintains high selectivity to determine these metal ions. This assay abrogates the need for complicated chemosensors or sophisticated equipment, providing a simple and highly selective detection platform.

■

■

AUTHOR INFORMATION

Corresponding Author

*Tel.: 86-731-58290045. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

ASSOCIATED CONTENT

■

* Supporting Information S

ACKNOWLEDGMENTS This work was supported by Natural Science Foundation of China (21075035) and Hunan Provincial Natural Science Foundation of China (10JJ5004)

Modification of preparing gold nanorods (GNRs) was performed to determine Fe3+, Ag+, Cu2+, and Hg2+ ions. GNRs were first prepared using silver ion-assisted seedmediated method, which serve as seed to prepare GNR@Au by depositing Au on the as-synthesized nanorods in the absence of Ag+. Apparently, a uniform layer of Au was deposited on the surface of the original GNRs, leading to the decrease of aspect ratio of the resultant nanorods. The GNR@Au is still termed as

■

REFERENCES

(1) Jiang, G.; Hore, M. J. A.; Gam, S.; Composto, R. J. ACS Nano 2012, 6, 1578−1588. 2318

dx.doi.org/10.1021/ac303305j | Anal. Chem. 2013, 85, 2312−2319

Analytical Chemistry

Article

(2) Katz-Boon, H.; Rossouw, C. J.; Weyland, M.; Funston, A. M.; Mulvaney, P.; Etheridge, J. Nano Lett. 2011, 11, 273−278. (3) Kim, S.; Kim, S. K.; Park, S. J. Am. Chem. Soc. 2009, 131, 8380− 8381. (4) Castellana, E. T.; Gamez, R. C.; Russell, D. H. J. Am. Chem. Soc. 2011, 133, 4182−4185. (5) Wijaya, A.; Schaffer, S. B.; Pallares, I. G.; Hamad-Schifferli, K. ACS Nano 2009, 3, 80−86. (6) Yasun, E.; Gulbakan, B.; Ocsoy, I.; Yuan, Q.; Shukoor, M. I.; Li, C.; Tan, W. Anal. Chem. 2012, 84, 6008−6015. (7) Wang, L.; Liu, Y.; Li, W.; Jiang, X.; Ji, Y.; Wu, X.; Xu, L.; Qiu, Y.; Zhao, K.; Wei, T.; Li, Y.; Zhao, Y.; Chen, C. Nano Lett. 2011, 11, 772− 780. (8) Huang, H.; He, C.; Zeng, Y.; Xia, X.; Yu, X.; Yi, P.; Chen, Z. Biosensor Bioelectron 2009, 24, 2255−2259. (9) Yang, D. P.; Cui, D. X. Chem. Asian J. 2008, 3, 2010−2022. (10) Kim, D. K.; Kerman, K.; Saito, M.; Sathuluri, R. R.; Endo, T.; Yamamura, S.; Kwon, Y. S.; Tamiya, E. Anal. Chem. 2007, 79, 1855− 1864. (11) Long, P.; Cao, T. C.; Park, S.; Kim, M.; Sim, S. J. Lab Chip 2011, 11, 2591−2597. (12) Huang, H.; Qu, C.; Liu, X.; Huang, S.; Xu, Z.; Zhu, Y.; Chu, P. K. Chem. Commun. 2011, 47, 6897−6899. (13) Chen, G.; Jin, Y.; Wang, L.; Deng, J.; Zhang, C. Chem. Commun. 2011, 47, 12500−12502. (14) Oyelere, A. K.; Chen, P. C.; Huang, X.; El-Sayed, I. H.; El-Sayed, M. A. Bioconjugate Chem. 2007, 18, 1490−1497. (15) Rex, M.; Hernandez, F. E.; Campiglia, A. D. Anal. Chem. 2006, 78, 445−451. (16) Liu, J. M.; Wang, H. F.; Yan, X. P. Analyst 2011, 136, 3904− 3910. (17) Wang, G.; Chen, Z.; Wang, W.; Yan, B.; Chen, L. Analyst 2011, 136, 174−178. (18) Tian, H.; Wang, Y.; Xue, Z.; Qu, Y.; Chai, F.; Hao, J. Sci. Total Environ. 2011, 16, 3078−3081. (19) Liu, W.; Zhou, Q.; An, J.; Sun, Y.; Liu, R. J. Hazard. Mater. 2010, 173, 737−743. (20) Sun, J.; Mao, J. D.; Gong, H.; Lan, Y. J. Hazard. Mater. 2009, 168, 1569−1574. (21) Chen, J. H.; Lin, H.; Luo, Z. H.; He, Y. S.; Li, G. P. Desalination 2011, 27, 7265−273. (22) Shen, C.; Shen, Y.; Wen, Y.; Wang, H.; Liu, W. Water Res. 2011, 45, 5200−5210. (23) Cabon, J. Y.; Giamarchi, P.; Le Bihan, A. Anal. Chim. Acta 2010, 664, 114−120. (24) Viyannalage, L. T.; Liu, Y.; Dimitrov, N. Langmuir 2008, 24, 8332−8337. (25) Li, G.; Wu, D.; Wang, Y.; Xie, W.; Zhang, X.; Liu, B.; Geostand. Geoanal. Res. 2012, 35, 334−340. (26) Zheng, F.; Hu, Bi. Talanta 2007, 73, 372−379. (27) Li, G.; Wu, D.; Wang, Y.; Xie, W.; Zhang, X.; Liu, B. J. Sep. Sci. 2012, 35, 334−340. (28) Wiederhold, J. G.; Cramer, C. J.; Daniel, K.; Infante, I.; Bourdon, B.; Kretzschmar, R. Environ. Sci. Technol. 2010, 44, 4191− 4197. (29) Tran, E.; Cohen, A. E.; Murray, R. W.; Rampi, M. A.; Whitesides, G. M. J. Am. Chem. Soc. 2009, 131, 2141−2150. (30) Mauzeroll, J.; Hueske, E. A.; Bard, A. J. Anal. Chem. 2003, 75, 3880−3889. (31) Pérez-Juste, J.; Liz-Marzán, L. M.; Carnie, S.; Chan, D. Y. C.; Mulvaney, P. Adv. Funct. Mater. 2004, 14, 571−579. (32) Jana, N. R.; Gearheart, L.; Murphy, C. J. J. Phys. Chem. B 2001, 105, 4065−4067. (33) Busbee, B. D.; Obare, S. O.; Murphy, C. J. Adv. Mater. 2003, 15, 414−416. (34) Jana, N. R.; Gearheart, L.; Murphy, C. J. J. Phys. Chem. B 2001, 105, 4065−4067. (35) Wu, H. Y.; Huang, W. L.; Huang, M. H. Cryst. Growth Des. 2007, 7, 831−835.

(36) Rodríguez-Fernández, J.; Pérez-Juste, J.; Mulvaney, P.; LizMarzán, L. M. J. Phys. Chem. B 2005, 109, 14257−14261. (37) Huang, H.; Liu, X.; Zeng, Y.; Yu, X.; Liao, B.; Yi, P.; Chu, P. K. Biomaterials 2009, 30, 5622−5630. (38) Huang, H.; Qu, C.; Liu, X.; Huang, S.; Xu, Z.; Liao, B.; Zeng, Y.; Chu, P. K. ACS Appl. Mater. Interfaces 2011, 3, 183−190.

2319

dx.doi.org/10.1021/ac303305j | Anal. Chem. 2013, 85, 2312−2319