Generalized Ratiometric Indicator Based Surface-Enhanced Raman

Article pubs.acs.org/ac

Generalized Ratiometric Indicator Based Surface-Enhanced Raman Spectroscopy for the Detection of Cd2+ in Environmental Water Samples Yao Chen, Zeng-Ping Chen,* Si-Yu Long, and Ru-Qin Yu State Key Laboratory of Chemo/Biosensing and Chemometrics, College of Chemistry and Chemical Engineering, Hunan University, Changsha, Hunan 410082, People’s Republic of China S Supporting Information *

ABSTRACT: The concept of generalized ratiometric indicator based surface-enhanced Raman spectroscopy was first introduced and successfully implemented in the detection of Cd2+ in environmental water samples using Au nanoparticles (AuNPs) modified by trithiocyanuric acid (TMT). Without the use of any internal standard, the proposed method achieved accurate concentration predictions for Cd2+ in environmental water samples with recoveries in the ranges of 91.8−108.1%, comparable to the corresponding values obtained by atomic absorption spectroscopy. The limit of detection and limit of quantification were estimated to be 2.9 and 8.7 nM, respectively. More importantly, other species present in water samples which cannot react with TMT and have weaker binding ability to AuNPs than TMT do not interfere with the quantification of Cd2+. Therefore, it is expected that the combination of the generalized ratiometric indicator based surface-enhanced Raman spectroscopy with the proposed AuNP−TMT probing system can be a competitive alternative for the primary screening of Cd2+ pollution.

C

solution to quantification of trace components in complex matrixes. It has been applied to a wide variety of biomedical and environmental analytical applications.18−20 However, the SERS effect relies heavily on enhancing substrates such as gold and silver nanoparticles. The SERS signals of analyte depend on not only the analyte concentration but also the physical properties of enhancing substrates (e.g., the particle size and shape of colloids and the degree of aggregation). The difficulty in obtaining highly stable and reproducible SERS signals renders SERS a qualitative or semiquantitative detection technique at the present stage. Internal standard methods have been used to correct SERS intensity variations induced by the variations in the physical properties of the enhancing substrate, as well as the laser power and alignment/focusing.21−24 A prerequisite for the application of conventional internal standard methods based on peak area/height ratios is that the internal standard must have one or more SERS peaks in spectrally silent regions of the analyte of interest, other coexisting SERS-active compounds, and possible background fluorescence interference. Therefore, different internal standards are generally needed for the detection of different analytes of interest. For some complex cases such as quantitative determination of proteins in living

admium is recognized as a highly toxic heavy metal which may have several chronic effects on humans such as cancer, renal damage, hypercalciuria, and enhanced tumor growth.1−6 Cadmium contamination of water is prevalent around the world.7,8 The Agency for Toxic Substances and Disease Registry (ATSDR) and U.S. Environmental Protection Agency (EPA) classify Cd as seventh on the Top 20 Hazardous Substances Priority List.9 The maximum limits of U.S. EPA and World Health Organization (WHO) standards9,10 for bottled water are about 4 and 40 nM, respectively. Over the years, various analytical techniques for detection of Cd2+ have been developed, including electrothermal atomic absorption spectrometry (ETAAS),11 atomic fluorescence spectrometry (AFS),12 ratiometric fluorescent probe,13 and inductively coupled plasma mass spectrometry (ICPMS).14 Although these methods can realize accurate quantification of Cd2+ in environmental samples, separation and preconcentration procedures for the determination of Cd2+ are generally needed for most applications because of the extremely low concentration of Cd2+ and the complicated matrixes in environmental and biological samples.15,16 Simple, sensitive, and cost-effective methods for the detection of Cd2+ are therefore still highly desirable. Due to its excellent molecular specificity, reduced photobleaching, and exquisite sensitivity, surface-enhanced Raman spectroscopy (SERS)17 has emerged as a potentially promising © XXXX American Chemical Society

Received: September 8, 2014 Accepted: November 10, 2014

A

dx.doi.org/10.1021/ac5033808 | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

being stirred, and then 2 mL of 1% (m/v) trisodium citrate solution was added quickly, resulting in a change in the color of the solution from pale yellow to purple. Subsequently, the mixture was refluxed for an additional 30 min and allowed to cool to room temperature. For the detection of Cd2+, the surfaces of Au nanoparticles (AuNPs) were modified by TMT through the addition of 40 μL of TMT methanol solution (340 μM) to 2 mL of AuNP solution followed by continuous stirring for 2 h. Sample Preparation and SERS Measurement. A total of 54 samples with different concentration levels of Cd2+ were prepared by diluting a standard solution of cadmium ion with ultrapure water (Table 1). To mimic real-world water samples

cells by the SERS technique, the conventional internal standard based methods are hardly applicable due to the difficulty in selecting appropriate internal standards. Zamarion et al. investigated the possibility of detecting hazardous metal ions by surface-enhanced Raman spectroscopy based on trithiocyanuric acid-modified gold nanoparticles without the use of internal standards.25 They claimed that linear relationships were observed between the concentrations of hazardous metal ions and the intensity ratios of certain pairs of SERS peaks of trithiocyanuric acid, which theoretically should be nonlinear. However, there were no detailed explanations of the above observation. Furthermore, the performance of the claimed linear models had not been verified on independent test samples, so there is no reason to believe that their approach is generically applicable. Recently, some of the present authors have proposed a multiplicative effects model for surface-enhanced Raman spectroscopy (MEMSERS) based on both the internal standard addition detection mode and the internal standard tagging detection mode,26,27 which can effectively eliminate the detrimental effects caused by variations in SERS enhancing substrates. MEMSERS solves the above-mentioned problems associated with the conventional internal standard based methods and can help realize accurate quantitative SERS assays. Unlike conventional internal standard methods, MEMSERS does not require the internal standards used to have SERS peaks in spectrally silent regions of the analyte of interest, other coexisting SERS-active compounds, and possible background fluorescence interference. Therefore, SERS enhancing substrates labeled with one specific internal standard are applicable to quantitative SERS analysis of many different analytes. Nevertheless, the MEMSERS method based on either the internal standard addition detection mode or the internal standard tagging detection mode still involves the use of internal standards, which complicates the implementation of quantitative SERS assays. In this study, MEMSERS was further extended to situations of generalized ratiometric indicator (whose SERS spectrum will deform when it is bound with the analyte of interest) based quantitative SERS assays without the use of any internal standard. The multiplicative effects model for generalized ratiometric indicator based surface-enhanced Raman spectroscopy (MEMGRI) was then applied to the quantitative SERS analysis of Cd2+ in water samples using a generalized ratiometric SERS indicator, trithiocyanuric acid.

Table 1. Experiment Design of Cd2+ Samples Prepared with Ultrapure Water sample no.a

c (μM)

sample no.a

c (μM)

sample no.a

c (μM)

1−6 7−12 13−18

0.000 0.010 0.062

19−24 25−30 31−36

0.083 0.100 0.120

37−42 43−48 49−54

0.170 0.190 0.210

a Samples 1−12, 19−24, 31−36, and 43−54 are calibration samples; the rest are test samples.

contaminated by Cd2+, another 54 samples with three different levels of Cd2+ concentrations (0.062, 0.100, and 0.170 μM) were also prepared by spiking the appropriate amount of Cd2+ into three different water matrixes, i.e., tap water, water collected from the Xiangjiang River (Changsha, China), and soil water (the supernatant of a mixture of 4.8 g of soil collected from cropland and 50 mL of ultrapure water after 30 min of ultrasonic treatment and centrifugation). For each level of Cd2+ concentration and each water matrix, there were six replicate samples. A 10 μL volume of each Cd2+ sample was mixed with 50 μL of TMT-modified AuNP colloids and 2.5 μL of 2 M KCl. After thorough mixing, about 10 μL of the above mixture was sampled by a capillary tube, and its SERS spectrum was measured by a Renishaw InVia micro-Raman spectrometer (Renishaw, Wotton under Edge, U.K.) equipped with a Peltier charge-coupled device (CCD) detector, a Leica microscope, and a diode laser (785 nm, 300 mW). All SERS spectra were collected within the range of 317−1477 cm−1 in static mode using a 50× objective. An exposure time of 7 s was employed, and four scans were accumulated for each spectrum with a resolution of 1 cm−1. Generalized Ratiometric Indicator Based SurfaceEnhanced Raman Spectroscopy for the Detection of Cd2+. Nitrogen atoms and sulfur atoms have great affinity for gold atoms. TMT can bind to AuNPs through its thiol sulfur and heterocyclic nitrogen atoms.31,32 The most probable mode of interaction of TMT with AuNPs might involve a tridentate coordination owing to the symmetry and close proximity of the thiol and nitrogen binding sites (Figure 1). The remaining thiol sulfur atom and two heterocyclic nitrogen atoms are available for chelating with Cd2+. According to Zamarion et al.,25 Cd2+ interacts primarily with the ring nitrogen atoms, and a bidentate coordination may be formed through a weaker Cd2+−S bond. The interaction between TMT and Cd2+ can result in deformations of the SERS spectrum of TMT. As shown in Figure 2, the addition of Cd2+ solution (0.890 μM) to the AuNP−TMT mixture significant increased the SERS intensities of TMT at 434 cm−1 (C−S stretching mode) and 972 cm−1

■

EXPERIMENTAL SECTION Reagents and Chemicals. All solvents and chemicals were of analytical grade and used without further purification. Trithiocyanuric acid (TMT; 95%), methanol, and standard solutions of metal ions used in this study were purchased from Aladdin Chemistry Co., Ltd. (Shanghai, China). Sodium citrate dihydrate (C6H5Na3O7·2H2O) and chlorauric acid tetrahydrate (HAuCl4·4H2O) were obtained from Sigma-Aldrich Chemical Reagent Co., Ltd. (Shanghai, China). Ultrapure water (18.2 MΩ cm−1) was produced by an Aquapro water system (Aquapro, Chongqing, China). Preparation of Au Nanoparticles. Au particles with average diameters of approximately 13 nm (Figure S-1, Supporting Information) were prepared by citrate reduction of HAuCl4.28−30 All glassware was cleaned in aqua regia (HCl:HNO3 = 3:1, v/v), rinsed with ultrapure water, and then dried in an oven prior to use. An aqueous solution of HAuCl4 (0.01%, m/v; 50 mL) was brought to reflux temperature while B



Article

Figure 1. Schematic diagram for the SERS detection of Cd2+ using the generalized ratiometric indicator trithiocyanuric acid.

Figure 2. SERS spectra of AuNP−TMT and AuNP−TMT−Cd2+.

(ring vibration peak) while simultaneously decreasing the Raman intensities at 488 cm−1 (C−S stretching mode) and 875 cm−1 (in-plane bending vibration of S−H).25,33 In general, besides the concentration of the target analyte, the physical properties of the enhancing substrates also have significant effects on the intensity of the SERS signals. For analytical purposes, the following MEMGRI was adopted for the subsequent quantitative SERS detection of Cd2+ using the generalized ratiometric indicator trithiocyanuric acid:

(cTMT), which is constant across samples and in large excess with respect to the total concentration of Cd2+ (ci,Cd2+) in the ith sample, [TMT−Cd2+]i is therefore approximately equal to ci,Cd2+, and eq 1 can be rewritten as x i = bi(c TMT·rTMT + ci ,Cd2+·Δr) + d i Δr = rTMT−Cd2+ − rTMT

In eq 2, model parameter bi explicitly accounts for the multiplicative confounding effects on SERS intensities caused by changes in variables other than the concentrations of the analytes in the ith calibration sample, such as the physical properties of the enhancing substrates and the intensity and alignment/focusing of the laser excitation source. The multiplicative parameters bi (i = 1, 2, ..., N) for N calibration samples in the above MEMGRI are different for different calibration samples and can be estimated from their SERS spectra by the modified optical length estimation and correction (OPLECm) method.34 After the estimation of bi, two calibration models can then be built by multivariate linear calibration methods such as partial least-squares (PLS) regression. One is between xi and bi (bi = α1 + xiβ1, i = 1, 2, ..., N), and the other is between xi and bici,Cd2+ (bici,Cd2+ = α2 + xiβ2, i = 1, 2, ..., N). After the establishment of the above two calibration models, the total concentration (ctest,Cd2+) of Cd2+ in any test sample can then be estimated from its measured SERS spectrum xtest according to ctest,Cd2+ = (α2 + xtestβ2)/(α1 + xtestβ1). The confounding multiplicative effects of the physical properties of the enhancing substrate and the intensity and alignment/focusing of the laser

x i = bi([TMT]i ·rTMT + [TMT−Cd2 +]i ·rTMT−Cd2+) + d i (i = 1, 2, ..., N )

(2)

(1)

where xi is the SERS spectrum of the ith calibration sample, [TMT]i and [TMT−Cd2+]i are the concentrations of free TMT and Cd2+-bound TMT in the ith sample, and rTMT and rTMT−Cd2+ represent the molecular scattering properties of free TMT and Cd2+-bound TMT. The multiplicative parameter bi explicitly accounts for the multiplicative confounding effects on SERS intensities caused by changes in variables other than the analytes’ concentrations in the ith calibration sample, such as the physical properties of the enhancing substrates and the intensity and alignment/focusing of the laser excitation source. The more significant the multiplicative confounding effects, the greater the variation of bi. di is a composite term that represents background interference(s) and the nonmultiplicative effects caused by variations in the physical properties of the enhancing substrates on the ith sample. Since the summation of [TMT]i and [TMT−Cd2+]i equals the total concentration of TMT used for the detection of Cd2+ C



Article

Figure 3. SERS spectra of a sample composed of 6.8 μM TMT measured at six different laser focusing positions.

Figure 4. I972/I875 vs ci,Cd2+ for the calibration samples prepared with ultrapure water.

■

excitation source on the quantitative results have been readily corrected. Data Analysis. SERS signals in the range of 380−1252 cm−1 were selected for subsequent quantitative analysis to ensure that the SERS peaks of both free TMT and Cd2+-bound TMT can be readily observed in the selected regions. The quantitative performance of the MEMGRI model was evaluated in terms of the root-mean-square error of prediction (RMSEP = (∑i N= 1(ci,Cd2+ − ĉi,Cd2+)2/N)1/2, where ci,Cd2+ and ĉi,Cd2+ are the actual and predicted total concentrations of Cd2+ in the ith test sample, respectively, and N is the number of test samples) and average relative prediction error (ARPE = (1/N)∑i N= 1|ci,Cd2+ − ĉi,Cd2+|/ci,Cd2+ × 100) and compared with that of the popular multivariate linear calibration methodPLSand an atomic absorption spectrometer equipped with a graphite furnace atomizer (PerkinElmer, United Kingdom). Leave-one-out cross-validation was employed to determine the optimal MEMGRI and PLS calibration models.

RESULTS AND DISCUSSION

The Raman reporter TMT plays an important role in the quantification of Cd2+ in water samples using surface-enhanced Raman spectroscopy based on a generalized ratiometric indicator. To acquire a satisfactory detection sensitivity and concentration range for Cd2+, the concentration of TMT used should be optimized. Concentration levels of TMT ranging from 2.8 to 11.3 μM were scrutinized (Figure S-2, Supporting Information). The intensity ratio between the peaks at 972 and 875 cm−1 remained at a stable and relatively low level when the concentration of TMT was increased from about 5.7 to 10.2 μM. In terms of observing more obvious changes in the intensity ratio between the peaks at 972 and 875 cm−1 during the addition of Cd2+, the concentration level of TMT could be set at a value between 5.7 and 10.2 μM. Nevertheless, on the one hand, the higher the concentration level of TMT, the lower the detection sensitivity for Cd2+. On the other hand, a lower concentration level of TMT means a narrower detection range for Cd2+. Therefore, for the concentration range of Cd2+ (0.010−0.210 μM) studied in this work, the concentration of D



Article

Figure 5. Model parameters bi estimated by OPLECm for calibration samples prepared with ultrapure water.

Figure 6. Concentration predictions for Cd2+ in the ultrapure water calibration (red circle) and test (blue triangle) samples obtained by MEMGRI.

TMT was chosen to be 6.8 μM to get a good balance between the detection sensitivity and concentration range for Cd2+. As mentioned above, besides the concentrations of the analytes, the SERS signals of a sample also depend on the physical properties of the enhancing substrate (e.g., the particle size and shape of the colloids and the degree of aggregation) and the intensity and alignment/focusing of the laser excitation source. As shown in Figure 3, due to the possible heterogeneity of AuNPs, the SERS spectra of the same sample composed of 6.8 μM TMT at different laser focusing positions differed significantly. The effects of the heterogeneity of the AuNPs on the SERS signals caused the relationship between the concentration of Cd2+ and the intensity ratio (I972/I875) between the two SERS peaks at 972 and 875 cm−1 to deviate significantly from a linear model (Figure 4). Therefore, it is difficult to predict the concentrations of Cd2+ in water samples by a univariate calibration model based on the intensity ratio between the two SERS peaks at 972 and 875 cm−1.

The model parameters bi (i = 1, 2, ..., N) estimated by OPLECm (Figure 5) visually reflect the extent of influence of the physical properties of the enhancing substrate and the intensity and alignment/focusing of the laser excitation source on the SERS signals of the calibration samples. It can be seen from Figure 5 that bi varies from 1.0 to 2.7, indicating the need to use MEMGRI to eliminate the significant detrimental effects of the physical properties of the enhancing substrate and the intensity and alignment/focusing of the laser excitation source when carrying out quantitative assays of Cd2+ in water samples using surface-enhanced Raman spectroscopy based on TMT as the generalized ratiometric indicator. The concentration predictions of MEMGRI for the Cd2+ samples prepared with ultrapure water are shown in Figure 6. Clearly, MEMGRI provided rather accurate concentration predictions for Cd2+ in both the calibration and test samples. The variation of its prediction values for samples with the same Cd2+ concentration level was considerably small, demonstrating the successful elimination of the detrimental effects of the physical properties of the enhancing substrate and the intensity E



Article

metal ions considered in this study result in no obvious changes, which illustrates that the probing system possesses high selectivity for Cd2+. To test the practical applicability of the proposed AuNP− TMT probing system for the detection of Cd2+ in real-world environmental water samples, the MEMGRI calibration of the Cd2+ samples prepared with ultrapure water was applied to the Cd2+ samples prepared with tap water, Xiangjiang River water, and soil water as well. Table 3 lists the concentration predictions for Cd2+ in environmental water samples obtained by MEMGRI. The recovery rates achieved by MEMGRI were in the range of 91.8−108.1%, which were comparable to (if not better than) the corresponding values obtained by atomic absorption spectroscopy (AAS). Such excellent results suggested that other chemical species present in environmental water samples did not interfere with the quantification of Cd2+, and the MEMGRI model established using calibration samples prepared with ultrapure water in the laboratory can therefore be applied to real-world samples. A reasonable explanation for this phenomenon is that the Au nanoparticles were covered with TMT, leaving no binding sites for other chemical species. Such a favorable feature makes the SERS detection of Cd2+ using the generalized ratiometric indicator trithiocyanuric acid stand out from other methods for Cd2+ quantification.

and alignment/focusing of the laser excitation source on the quantitative SERS detection of Cd2+. Table 2 lists the detailed Table 2. Concentrations of Cd2+ in the Test Samples Prepared with Ultrapure Water Determined by MEMGRI and PLS method MEMGRI

PLS

a

number of samples

true concn (μM)

mean predicted concn (μM)

6 6 6 6 6 6

0.062 0.100 0.170 0.062 0.100 0.170

0.063 0.103 0.172 0.081 0.096 0.153

RMSEP (μM)

ARPE (%)

0.006

5.0

0.017

15.6

(±0.004)a (±0.008) (±0.007) (±0.008) (±0.006) (±0.012)

The numbers in parentheses are standard deviations.

results of MEMGRI and those of the PLS model for comparison. Obviously, MEMGRI far outperformed the PLS model in terms of both RMSEP and ARPE values. The RMSEP and ARPE values of MEMGRI were 0.006 μM and 5.0%, respectively, which are about one-third of the corresponding values of the PLS model. Additionally, the limit of detection (LOD) and limit of quantification (LOQ)35,36 were estimated to be 2.9 and 8.7 nM, respectively. These results fully demonstrated that the combination of MEMGRI with the generalized ratiometric indicator trithiocyanuric acid could realize accurate quantitative detection of Cd2+ in ultrapure water samples using surfaceenhanced Raman spectroscopy. To investigate the selectivity of generalized ratiometric indicator TMT based surface-enhanced Raman spectroscopy for the detection of Cd2+ in water samples, metal ions commonly found in environmental water samples (Cd2+, As3+, Ca2+, Cr3+, Fe3+, Hg2+, Mg2+, Mn2+, Ni2+, and Zn2+ at the same concentration of 0.830 μM) were separately added to the AuNP−TMT (6.8 μM) mixture, and the SERS intensity ratio between the SERS peaks at 972 and 875 cm−1 was calculated (Figure 7). As shown in Figure 7, only the addition of Cd2+ could induce a significant increase in the SERS intensity ratio between the SERS peaks at 972 and 875 cm−1. Other

■

CONCLUSIONS In this paper, we have introduced the concept of generalized ratiometric indicator based surface-enhanced Raman spectroscopy. Through the combination of the proposed new concept and Au nanoparticles modified with a generalized ratiometric indicator (TMT), the concentrations of Cd2+ in real-world environmental water samples were accurately quantitatively determined with recovery rates in the range of 91.8−108.1%. Other chemical species present in environmental water samples which cannot react with TMT and have weaker binding ability to Au nanoparticles than TMT do not interfere with the quantification of Cd2+. The concept of generalized ratiometric indicator based surface-enhanced Raman spectroscopy can be readily extended to the rapid, low-cost, highly sensitive and selective detection of other chemical species in systems with complex matrixes.

Figure 7. SERS intensity ratio between 972 and 875 cm−1 when metal ions (Cd2+, As3+, Ca2+, Cr3+, Fe3+, Hg2+, Mg2+, Mn2+, Ni2+, and Zn2+ at the same concentration of 0.830 μM) were separately added to the AuNP−TMT (6.8 μM) mixture. F



Article

Table 3. Concentrations of Cd2+ in Different Samples Determined by MEMGRI and AAS MEMGRI sample

concn added (μM)

tap water

Xiangjiang River water

soil water

a

0.062 0.100 0.170 0.062 0.100 0.170 0.062 0.100 0.170

concn found (μM) 0.063 0.103 0.156 0.067 0.098 0.169 0.058 0.104 0.171

(±0.006)a (±0.006) (±0.010) (±0.004) (±0.004) (±0.005) (±0.007) (±0.002) (±0.004)

AAS recovery (%) 101.6 103.0 91.8 108.1 98.0 99.4 93.5 104.0 100.6

concn found (μM) 0.068 0.094 0.173 0.065 0.095 0.180 0.053 0.088 0.174

(±0.003) (±0.015) (±0.018) (±0.004) (±0.007) (±0.003) (±0.001) (±0.010) (±0.006)

recovery (%) 109.7 94.0 101.8 104.8 95.0 105.9 85.5 88.0 102.4

The numbers in parentheses are standard deviations.

■

(15) Ferreira, S. L. C.; de Andrade, J. B.; Korn, M. D. A.; Pereira, M. D.; Lemos, V. A.; dos Santos, W. N. L.; Rodrigues, F. D.; Souza, A. S.; Ferreira, H. S.; da Silva, E. G. P. J. Hazard. Mater. 2007, 145, 358−367. (16) Pereira, M. G.; Arruda, M. A. Z. Microchim. Acta 2003, 141, 115−131. (17) Fleischmann, M.; Hendra, P. J.; McQuillan, A. J. J. Chem. Phys. Lett. 1974, 26, 163−166. (18) Baraldi, C.; Freguglia, G.; Tinti, A.; Sparta, M.; Alexandrova, A. N.; Gamberini, M. C. Spectrochim. Acta, Part A 2013, 103, 1−10. (19) Wu, Z. T.; Liu, Y. Z.; Zhou, X. D.; Shen, A. G.; Hu, J. M. Biosens. Bioelectron. 2013, 44, 10−15. (20) Smith, W. E. Chem. Soc. Rev. 2008, 37, 955−964. (21) Lee, S.; Choi, J.; Chen, L.; Park, B.; Kyong, J. B.; Seong, G. H.; Choo, J.; Lee, Y.; Shin, K. H.; Lee, E. K.; Joo, S. W.; Lee, K. H. Anal. Chim. Acta 2007, 590, 139−144. (22) Lorén, A.; Engelbrektsson, J.; Eliasson, C.; Josefson, M.; Abrahamsson, J.; Johansson, M.; Abrahamsson, K. Anal. Chem. 2004, 76, 7391−7395. (23) Zhang, D.; Xie, Y.; Deb, S. K.; Davison, V. J.; Ben-Amotz, D. Anal. Chem. 2005, 77, 3563−3569. (24) Deb, S. K.; Davis, B.; Knudsen, G. M.; Gudihal, R.; Ben-Amotz, D.; Davisson, V. J. J. Am. Chem. Soc. 2008, 130, 9624−9625. (25) Zamarion, V. M.; Timm, R. A.; Araki, K.; Toma, H. E. Inorg. Chem. 2008, 47, 2934−2936. (26) Song, J.; Chen, Z. P.; Jin, J. W.; Chen, Y.; Yu, R. Q. Chemom. Intell. Lab. Syst. 2014, 135, 31−36. (27) Xia, T. H.; Chen, Z. P.; Chen, Y.; Jin, J. W.; Yu, R. Q. Anal. Methods 2014, 6, 2363−2370. (28) Frens, G. Nat. Phys. Sci. 1973, 241, 20−22. (29) Grabar, K. C.; Freeman, R. G.; Hommer, M. B.; Natan, M. J. Anal. Chem. 1995, 67, 725−743. (30) Storhoff, J. J.; Elghanian, R.; Mucic, R. C.; Mirkin, C. A.; Letsinger, R. L. J. Am. Chem. Soc. 1998, 120, 1959−1964. (31) Wei, F.; Lam, R.; Cheng, S.; Lu, S.; Ho, D.; Li, N. Appl. Phys. Lett. 2010, 96, 133702−133704. (32) He, S.; Liu, D. B.; Wang, Z.; Cai, K. Y.; Jiang, X. Y. Sci. China: Phys., Mech. Astron. 2011, 54, 1757−1765. (33) Rostkowska, H.; Lapinski, L.; Khvorostov, A.; Nowak, M. J. J. Phys. Chem. A 2005, 109, 2160−2166. (34) Jin, J. W.; Chen, Z. P.; Li, L. M.; Steponavicius, R.; Thennadil, S. N.; Yang, J.; Yu, R. Q. Anal. Chem. 2012, 84, 320−326. (35) Boqué, R.; Ferré, J.; Faber, N. M.; Rius, F. X. Anal. Chim. Acta 2002, 451, 313−321. (36) Lorber, A. Anal. Chem. 1986, 58, 1167−1172.

ASSOCIATED CONTENT

* Supporting Information S

SEM image of Au nanoparticles and intensity ratio between the SERS peaks at 972 and 875 cm−1 versus the concentration of TMT. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +86 731 88821989. Fax: +86 731 88821989. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge the financial support of the National Natural Science Foundation of China (Grants 21275046 and 21475039), the National Instrumentation Program of China (Grant 2011YQ0301240102), the Specialized Research Fund for the Doctoral Program of Higher Education (Grant 20130161110027), and the Program for New Century Excellent Talents in University (Grant NCET-12-0161).

■

REFERENCES

(1) Stoeppler, M. Hazardous Metals in the Environment; Techniques and Instrumentation in Analytical Chemistry, Vol. 12; Elsevier: Amsterdam, 1992; p 177. (2) Daher, R. T. Anal. Chem. 1995, 67, 405−410. (3) Friberg, L; Kjellstrom, T; Nordberg, G. F. Handbook on the Toxicology of Metals; Elsevier: Amsterdam, 1986; Vol. 2, pp 130−143. (4) Robards, K.; Worsfold, P. Analyst 1991, 116 (6), 549−568. (5) Darwish, I. A.; Blake, D. A. Anal. Chem. 2001, 73, 1889−1895. (6) Davis, A. C.; Wu, P.; Zhang, X. F.; Hou, X. D.; Jones, B. T. Appl. Spectrosc. 2006, 41, 35−75. (7) Dobson, S. CadmiumEnvironmental Aspects; Environmental Health Criteria 135; World Health Organization: Geneva, 1992. (8) Yi, H. C.; Mei, P. J. Appl. Electrochem. 2008, 38, 1623−1627. (9) Agency for Toxic Substances and Disease Registry, 4770 Buford Hwy NE, Atlanta, GA 30341. (10) World Health Organization, Ave. Appia 20, 1211 Geneva 27, Switzerland. (11) Butcher, D. J. Appl. Spectrosc. 2006, 41, 15−34. (12) Reis, B. F.; Ródenas-Torralba, E.; Sancenón-Buleo, J.; MoralesRubio, A.; de la Guardia, M. J. Anal. At. Spectrom. 2002, 17, 537−540. (13) Liu, Z. P.; Zhang, C. L.; He, W. J.; Yang, Z. H.; Gao, X.; Guo, Z. J. Chem. Commun. 2010, 46, 6138−6140. (14) Palmer, C. D.; Lewis, M. E.; Geraghty, C. M.; Barbosa, F.; Parsons, P. J. Spectrochim. Acta, Part B 2006, 61, 980−990. G


Generalized Ratiometric Indicator Based Surface-Enhanced Raman

Recommend Documents