Using Standing Gold Nanorod Arrays as Surface-Enhanced Raman

Jan 12, 2017 - ... and Sensitive Analysis of Allura Red and Sunset Yellow in Beverages ... by AuNPs-Analyte/AuNPs Double-Decker Structure on Wax-Coate...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/JAFC

Using Standing Gold Nanorod Arrays as Surface-Enhanced Raman Spectroscopy (SERS) Substrates for Detection of Carbaryl Residues in Fruit Juice and Milk Fouad K. Alsammarraie and Mengshi Lin* Food Science Program, Division of Food System & Bioengineering, University of Missouri, Columbia, Missouri 65211-5160, United States S Supporting Information *

ABSTRACT: In recent years, there have been increasing concerns about pesticide residues in various foods. On the other hand, there is growing attention in utilizing novel nanomaterials as highly sensitive, low-cost, and reproducible substrates for surfaceenhanced Raman spectroscopy (SERS) applications. The objective of this study was to develop a SERS method for the rapid detection of pesticides that were extracted from different types of food samples (fruit juice and milk). A new SERS substrate was prepared by assembling gold nanorods into standing arrays on a gold-coated silicon slide. The standing nanorod arrays were neatly arranged and were able to generate a strong electromagnetic field in SERS measurement. The as-prepared SERS substrate was utilized to detect carbaryl in acetonitrile/water solution, fruit juices (orange and grapefruit), and milk. The results show that the concentrations of carbaryl spiked in fruit juice and milk were linearly correlated with the concentrations predicted by the partial least-squares (PLS) models with r values of 0.91, 0.88, and 0.95 for orange juice, grapefruit juice, and milk, respectively. The SERS method was able to detect carbaryl that was extracted from fruit juice and milk samples at a 50 ppb level. The detection limits of carbaryl were 509, 617, and 391 ppb in orange juice, grapefruit juice, and milk, respectively. All detection limits are below the maximum residue limits that were set by the U.S. EPA. Moreover, satisfactory recoveries (82−97.5%) were accomplished for food samples using this method. These results demonstrate that SERS coupled with the standing gold nanorod array substrates is a rapid, reliable, sensitive, and reproducible method for the detection of pesticide residues in foods. KEYWORDS: gold nanorod, SERS, pesticide, fruit juice, milk



INTRODUCTION The use of pesticides is important in agriculture because pesticides can increase the crop yield and enhance the quality of foods.1 However, pesticide residues are potential hazards to consumers’ health, and they are particular concerns especially in minimally processed foods such as fruits and vegetables and in other food products such as milk. There is growing epidemiologic evidence about the relationship between pesticides and various human diseases.2,3 Consequently, the use of pesticides can lead to great jeopardy for the environment and public health. Therefore, there is a crucial need for suitable analytical methods that could quickly detect pesticide residues in various food matrices. There are many traditional methods that have been used to measure pesticide residues in foods, such as gas chromatography (GC), high-performance liquid chromatography (HPLC), electrochemiluminescence, mass spectroscopy, and voltammetric methods.4−6 However, there are many disadvantages of using these methods, because they are labor-intensive and time-consuming and mostly they need complicated sample pretreatment. In addition, the scarceness of a quick, easy, and economic detection method to detect pesticides in a complex food matrix such as milk is a major hindrance for the food industry. In recent years, there is increasing interest in using novel nanotechnology and nanosized materials for detection of food contaminants. Particularly, surface-enhanced Raman spectroscopy (SERS) has received much attention in the area © XXXX American Chemical Society

of analytical chemistry and food safety. For example, SERS has been used to detect melamine in milk,7 illegal substances in fish,8 and biological contaminants such as bacteria in food samples.9 SERS is a technique that enhances Raman scattering from analyte molecules that are adsorbed on the roughened metal surface, particularly on silver and gold nanostructures.10−12 The enhancement factor could reach 106−1010 or higher.13−15 SERS has become one of the most utilized spectroscopic methods to detect and identify biological and chemical species due to its high sensitivity and specificity.7,16,17 SERS has many advantages compared to other conventional methods. For instance, the sample preparation for SERS is simple. Liquid samples could be measured directly on a SERS substrate. However, chromatographic methods need more time for sample preparation in which food contaminants must be extracted or separated from food samples using organic solvents for HPLC procedure and additionally refined for contaminants in GC method. Also, the time for collecting a SERS spectrum is very short in comparison to chromatographic methods. All of these features make SERS an ideal alternative method to detect chemical contaminants in food samples.18 Received: October 25, 2016 Revised: December 22, 2016 Accepted: December 27, 2016

A

DOI: 10.1021/acs.jafc.6b04774 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry

mL of distilled water. The AuNR solution was kept at room temperature as a stock solution. Preparation of Standing AuNR Arrays. A volume of 0.1 mL of CTAB solution (2.5 mM) was redispersed with 1.0 mL of the AuNR stock solution that was already recentrifuged. The solution then underwent 60 s of sonication for full dispersion of the AuNR solution. Five drops of as-prepared AuNR solution (10 μL) were distributed on the gold-coated silicon slides (1 cm × 5 cm). The gold-coated silicon slides were then transmitted into a polystyrene Petri dish. A volume of 50 μL of distilled water was then put in the polystyrene Petri dish. The Petri dish was wrapped by the parafilm and put in an incubator at 37 °C. After 5 days, the drops of as-prepared AuNRs were fully dried, and standing AuNR layers were created on the gold-coated silicon slides. The gold slide was then immersed in pure ethanol for 5 min to eliminate the CTAB layer from the top of the standing AuNR arrays. Finally, the gold-coated silicon slides were left for air-drying at room temperature, and then the slides were directly used as SERS substrates.32 Detection of Carbaryl in Foods. Briefly, each sample of orange juice, grapefruit juice, and milk was spiked with various concentrations of carbaryl (0−1000 ppb). Then, the sample was centrifuged at 3913g force for 10 min. After that, a volume (10 μL) of the supernatant was directly put on the standing AuNR arrays and dried at room temperature for SERS measurement. SERS Measurement. A Renishaw RM1000 Raman spectrometer system (Gloucestershire, UK) equipped with a Leica DMLB microscope (Wetzlar, Germany) was used in this study. This system is equipped with a 785 nm near-infrared diode laser source. During the measurement, light from the high-power diode laser was directed and focused onto the sample at a microscope stage through a 50× objective. Raman scattering signals were detected by a 578 × 385 pixel CCD array detector. Spectral data were collected by WiRE 3.4 software (Gloucestershire, UK). In this study, spectra of samples were collected with 1 min of exposure time, 0% focus, and ∼20 mW laser power in the extended mode. The detection range for carbaryl was 1000−1700 cm−1. The detection range was determined in a way that the range was as narrow as possible but no obvious signals were missed. Data Analysis. The software Delight version 3.2.1 (D-Squared Development Inc., LaGrande, OR, USA) was used in data analysis. SERS spectral data were analyzed following the method that was used in our previously published papers.37 Moreover, Gaussian smoothing with 6 cm−1 was employed to remove high-frequency noises from the instrument. In addition, a second-derivative transformation with a 12 cm−1 gap was used before running PLS to eliminate overlapping of spectra. The PLS model, a multivariate statistical regression model, was established to predict analyte concentrations in tested food samples. The number of PLS latent variables was optimized on the basis of the lowest root-mean-square error of prediction (RMSEP) values to avert overfitting of spectral data.

There are various methods that have been utilized to synthesize high-quality SERS substrates, such as microsphere lithography, nanolithography, and self-assembling methods.19−22 Among these methods, self-assembly is considered simpler and more cost-effective compared to lithographic methods. Gold nanomaterials, such as nanorods, nanocubes, and nanospheres, are widely utilized to manufacture SERS substrates by an evaporation-induced self-assembling procedure.23−26 Particularly, gold nanorods (AuNRs) have achieved very good results in SERS applications because of their unique and anisotropic structures that can generate a strong electromagnetic field.24,25,27,28 Some previous studies have shown that laying a specific amount of cetyltrimethylammonium bromide (CTAB) on the surface of AuNRs will help to form a monolayer of vertically aligned AuNRs on a silicon substrate, which can be used as an ideal SERS substrate.29,30 The aim of this study was to develop a SERS method coupled with standing AuNR arrays to detect carbaryl in fruit juice (i.e., orange and grapefruit) and milk. AuNRs were assembled on gold-coated silicon slides for the following two benefits: first, conjugation between the AuNRs and the gold layer could produce a vigorous electromagnetic field; second, because of the wet feature of gold film, the AuNR droplets would not dwindle after evaporation. Therefore, it will form a thin layer of CTAB on the top of AuNRs, and this layer of CTAB could be readily washed away by ethanol. A food sample can be immediately put on the AuNR arrays for detection. The vertically aligned AuNRs will be used in SERS after elimination of the CTAB layer from the top of AuNRs. The solution of food samples will be uniformly deposited on the aligned AuNR substrates to collect SERS spectra. To our knowledge, this is the first study using SERS coupled with standing AuNR arrays to detect carbaryl in fruit juice and milk.



MATERIALS AND METHODS

Materials and Chemicals. Silver nitrate, ascorbic acid, hydrogen tetrachloroaurate solution (HAuCl4, 30 wt % in dilute HCl), and CTAB were purchased from Sigma-Aldrich (St. Louis, MO, USA). Carbaryl was obtained from Fisher Scientific (Pittsburgh, PA, USA). All of these materials were utilized without any purification. Silicon slides were purchased from Ted Pella (Redding, CA, USA). Moreover, polystyrene Petri dishes (W, 94 mm; H, 15 mm; VWR, PA, USA) and parafilm (PM-992, American National Can Co., Menasha, WI, USA) were used in this study. Organic orange juice, grapefruit juice, and milk were purchased from a local grocery store. A 100 ppm (w/v) of carbaryl stock solution was prepared by mixing with organic solvent system (acetonitrile/H2O = 1:1, v/v). Solutions of 0, 50, 100, 200, 400, 600, 800, and 1000 ppb of carbaryl were prepared by serial dilutions from the 100 ppm solution. The solvent without pesticides was utilized in this study as a control. Synthesis of AuNRs. AuNRs were synthesized using a seedmediated method. Briefly, 0.10 mL of HAuCl4 (5 mM) and 1.9 mL of CTAB (0.2 M) were mixed in a glass vial. The seeds were produced by adding 0.12 mL of freshly prepared ice-cold NaBH4 (10 mM) into the glass vial that was already submitted to intensive shaking. The seed solution was aged for 30 min at room temperature, and the solution was then ready to be used. To prepare the growth solution, 44 mL of 0.5 mM HAuCl4 was mixed with 40 mL of CTAB solution (0.2 M) in a volumetric flask. Then, 800 μL of AgNO3 (10 mM), 1.6 mL of HCl (1.0 M), and 1.1 mL of fresh ascorbic acid (0.1 M) were injected into the mixture. After the solution had been stirred for 30 s, the yellow color of the growth solution vanished. Then, the growth solution was injected with 10 μL of seed solution under gentle stirring. Eventually, the solution was kept overnight in a water bath (28 °C) for the growth of AuNRs.31 The AuNR solution was subjected to centrifugation at 1409g force for 7 min for purification purpose and redispersed in 20

n

RMSEP =

∑i (cî − ci)2 n

(1)

In this equation, n is the number of samples, ĉi is the predicted pesticide concentration (ppb), and ci is the actual pesticide concentration (ppb). The correlation coefficient (r) and RMSEP were utilized to estimate the model. In general, the higher the r value or the lower the RMSEP value is, the better predictability the model has. The detection limit (DL) with 99.86% confidence interval was calculated using the PLS calibration curve based on characteristic peaks of SERS spectra using the formula33

DL = 3σ /m

(2)

in which σ is the standard error of predicted concentration and m is the slope of the calibration curve. In a PLS model, σ equals RMSEP. B

DOI: 10.1021/acs.jafc.6b04774 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry



RESULTS AND DISCUSSION Characterization of Gold Nanorods. The synthesized AuNRs display two separate surface plasmon resonance (SPR) bands that are known as transverse and longitudinal plasmon bands as shown in Figure 1A. A strong band at 725 nm is due

For the preparation of standing AuNR arrays, our previous studies have shown that AuNRs can be vertically aligned on a silicon substrate by either a two-step or a one-step method without the need to modify the AuNR top surface.35 Because of the shrinkage of nanorod droplets, a thick CTAB layer would be formed on the upper part of the standing nanorod arrays. To utilize AuNRs as a SERS material, it is very important to eliminate the CTAB layer and, at the same time, not to damage the morphology of the standing arrays, which is considered a challenging mission. To do that, the droplets of CTAB/AuNR solution evaporated slowly under a high-humidity environment that provided enough time for AuNRs to arrange in a side-byside pattern. The assembled AuNRs eventually precipitated on a gold-coated silicon slide and formed vertically aligned AuNR arrays. A drop of pesticide solution was put directly on the AuNR substrate surface and tested by SERS after the CTAB layer was washed and removed using ethanol. AuNRs were closely packed with their tips facing up on the silicon slide, which will be very useful for inducing a powerful electromagnetic field. Detection of Carbaryl in Acetonitrile Solution. The standing AuNR arrays were used to detect carbaryl in acetonitrile/water solution. Figure 2A shows the averaged SERS spectra (n = 8) of different concentrations of carbaryl. Distinctive peaks at 1382, 1442, and 1578 cm−1 could be obviously observed in the Raman spectra. The intensity of these peaks increased as the concentration of carbaryl increased from 0 to 1000 ppb. Some studies have been reported to detect carbaryl in food samples using different SERS substrates, such as Q-SERS and silver nanoparticle-coated Si nanowire SiNW arrays.36,37 In most of those studies, ppm levels of the carbaryl solutions were utilized to spike food samples. However, in our study, ppb levels of the carbaryl solutions were used, which demonstrate better performance and higher sensitivity of the standing AuNR arrays compared to other types of SERS substrates. A strong peak at 1382 cm−1 is due to the symmetric vibration of the naphthalene ring. A peak at 1442 cm−1 arises from unspecified vibrations of the monosubstituted naphthalene ring that exists in the carbaryl structure. A strong peak at 1578 cm−1 could refer to the stretching of CC double bonds in the naphthalene ring.37 In general, our results are in good agreement with previous papers.32,37 As shown in Figure 2B, the r value and RMSEP are 0.92 and 1.364 × 10−7, respectively. The results demonstrate that the as-prepared SERS substrates can be used for the quantification of carbaryl in acetonitrile/ water solution. Detection of Carbaryl in Orange Juice, Grapefruit Juice, and Milk. The supernatants of spiked orange juice, grapefruit juice, and milk were directly measured by SERS after they were centrifuged for 10 min. In all three types of food samples, SERS was able to detect carbaryl as low as 50 ppb. Figure 3A shows the average SERS spectra of carbaryl in grapefruit juice. The most significant and distinctive peaks can be observed at 1382, 1442, and 1578 cm−1. The intensity of those peaks increased with increasing concentration of carbaryl in grapefruit juice from 0 to 1000 ppb. Figure 3B shows that the PLS model of the SERS spectra in grapefruit juice displays a good linear correlation between the actual concentration of carbaryl and predicted the concentration of carbaryl in grapefruit juice (r = 0.88; RMSEP = 1.665 × 10−7). These results indicate that carbaryl in grapefruit juice can be quantified by SERS coupled with the standing AuNR arrays.

Figure 1. UV−vis spectrum of synthesized AuNRs (A); TEM image of AuNRs (B); SEM image of the vertical AuNR arrays on gold-coated silicon slide (C).

to the resonant propagation of surface plasmon along the longitudinal axis. A weak band at 530 nm is probably caused by the impurities or due to the presence of a small amount of spherical and cubic gold nanoparticles in the AuNRs solution (Figure 1B). The UV−vis absorption spectrum exhibits a narrow absorption band, which refers to the monodispersed morphology of AuNRs. Moreover, the longitudinal or rod shape of the AuNRs is the dominant one in the purified solution, which makes them suitable for use as SERS substrates with a 785 nm near-infrared diode laser source. The SEM image demonstrates that gold nanorod arrays are flat and vertically aligned on a gold−silicon slide (Figure 1C). The gold nanorods were obviously packed side-by-side on the slide. The TEM image shows that AuNRs have rod shape with a length of 87.8 ± 8.9 nm and a width of 27 ± 4.6 nm. The aspect ratio (length/width) of AuNRs is ∼3.2 (Figure 1B). AuNRs have a more complicated anisotropic structure compared to spherical gold nanoparticles. AuNRs have more complexity of scattering properties and absorbance in which AuNRs produce multiple plasmon bands including the transverse plasmon (short axis) and the longitudinal plasmon (long axis). As a result, AuNRs have shown higher performance in optical imaging applications. Excitation of SPR induces a vigorous electromagnetic field at the surface of AuNRs that could expand over a distance compared to the spherical AuNP’s diameter. Therefore, AuNRs can be utilized to enhance the scattering spectra of chemical species that are adsorbed onto AuNRs, especially in SERS applications.34 C

DOI: 10.1021/acs.jafc.6b04774 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry

Figure 2. Average SERS spectra (n = 8) of carbaryl in acetonitrile/water solution with concentrations from 50 to 1000 ppb (A); predicted carbaryl concentration versus actual carbaryl concentration using the PLS models (B): smoothing 6 cm−1, baseline adjustment by subtracting a second-order polynomial function; spectral region of 1000−1700 cm−1; spectral number n = 64.

prediction values for carbaryl in orange juice were obtained with an r value of 0.91 and an RMSEP of 1.425 × 10−7. The PLS model shows that there is a good linear correlation between the actual concentrations and the predicted concentrations of carbaryl in orange juice, which indicates that the PLS model is valid for use in this study. Detection of chemical contaminants in milk is challenging because milk contains complex food matrices such as proteins. Figure S1 shows the average SERS spectra of the milk samples containing carbaryl. The most distinctive peaks of carbaryl at 1382 and 1442 cm−1 could be clearly observed, and the intensity is correlated with the concentration of carbaryl. In Figure S1A, a peak at 1578 cm−1 was affected to some degree by other milk components. This result is expected in complex food systems such as milk. Milk proteins have chemical reactive

Figure 4A shows the average SERS spectra acquired from orange juice samples that were contaminated by carbaryl. Prominent peaks of carbaryl can be observed at 1382 and 1442 cm−1. However, a peak at 1578 cm−1 was slightly affected because of the interference from other food components. A main challenge to use SERS for the detection of food chemical contaminants is that SERS collects signals from both target analytes and food components that are considered interferences.38 Orange juice contains a high amount of sugar that formed a transparent layer on the SERS substrate after the droplet was dried. The sugar has preferential interaction with AuNRs that would decrease the enhancement ability of SERS substrate.39 However, sugar and other carbohydrates in orange juice did not generate considerable interfering peaks in the SERS spectra in this study. As illustrated in Figure 4B, PLS D

DOI: 10.1021/acs.jafc.6b04774 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry

Figure 3. Average SERS spectra (n = 8) of carbaryl with concentration from 50 to 1000 ppb spiked in grapefruit juice (A); predicted carbaryl concentration versus actual carbaryl concentration using the PLS models (B): smoothing 6 cm−1, baseline adjustment by subtracting a second-order polynomial function; spectral region of 1000−1700 cm−1; spectral number n = 64.

issue did not cause any problem for carbaryl detection because carbaryl has two prominent peaks at 1382 and 1442 cm−1. Figure S1B shows the PLS model indicating a good linear correlation between the actual concentrations and the predicted concentrations of carbaryl in milk. The r value and RMSEP are 0.95 and 1.146 × 10−7, respectively. The r and RMSEP values for all food types are summarized in Table 1. The results demonstrate that the SERS method coupled with standing AuNR arrays is a reliable and accurate method to quantify carbaryl in different types of food samples containing complex matrices. A second-derivative transformation is a useful tool in spectral data analysis. It can separate overlapping peaks, accomplish baseline correction, and improve spectral resolution. Raman spectra of carbaryl in all three types of food samples were

functional groups, such as sulfur linkages and primary amines that have an affinity to bond with the surface of AuNRs. As a result, this attribute will prevent target molecules from binding to the SERS substrate and eventually affects Raman spectra. In addition, the electromagnetic enhancement of SERS substrate decreases remarkably as the distance between the target molecules and the substrate surface increases.39 Therefore, sample pretreatment is important to eliminate or minimize the influence of those large molecules, such as milk proteins or fat. A common method to remove these large molecules from milk samples is centrifugation, which has been used in this study. There were some random and small Raman peaks in the spectra that are probably due to random noises from the Raman spectrometer during the measurement or interactions between food components and target molecules. However, this E

DOI: 10.1021/acs.jafc.6b04774 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry

Figure 4. Average SERS spectra (n = 8) of carbaryl with concentrations from 50 to 1000 ppb spiked in orange juice (A); predicted carbaryl concentration versus actual carbaryl concentration using the PLS models (B): smoothing 6 cm−1, baseline adjustment by subtracting a second-order polynomial function; spectral region of 1000−1700 cm−1; spectral number n = 64.

Detection Limits and Recoveries of Carbaryl in Different Food Samples. The detection limit of SERS for carbaryl extracted from real food samples was calculated on the basis of eq 2. As shown in Figure 7, a calibration curve of the PLS method shows a good linear relationship between the actual and predicted concentrations of carbaryl that was extracted from orange juice samples. The detection limits and maximum residue limits (MRLs) of carbaryl for the three types of foods and acetonitrile/water solution are shown in Table 1. The detection limits of carbaryl in orange juice, grapefruit juice,

preprocessed by the second-derivative transformation. Figure 5 shows the most notable and characteristic peak of carbaryl at 1382 cm−1. This figure indicates that SERS is able to detect and distinguish spectral patterns between different concentrations of carbaryl that were mixed with food samples. PLS analysis was done on spectra that were collected from carbaryl in acetonitrile/water solution and food samples to obtain the RMSEP values with different latent variables. Figure 6 illustrates that an optimum number of latent variables of carbaryl in milk sample is three with the lowest RMSEP value. F

DOI: 10.1021/acs.jafc.6b04774 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry

Figure 5. Second-derivative transformation of a peak at 1382 cm−1 of an average SERS spectrum (n = 8) acquired from different concentrations (50−1000 ppb) of carbaryl solutions.

Table 1. Prediction Results of PLS Models, Detection Limits (DL), and MRLs of SERS Method for Detection of Carbaryl in Food Samples sample carbaryl in acetonitrile (pure pesticide) carbaryl in orange juice carbaryl in grapefruit juice carbaryl in milk

r value

standard error

slope

DL (ppb)

MRL (ppm)

0.92

1.364

0.86

476

N/A

0.91 0.88 0.95

1.425 1.665 1.146

0.84 0.81 0.88

509 617 391

10 10 1

and milk were 509, 617, and 391 ppb, respectively. The MRLs of carbaryl are 10 ppm in orange and grapefruit juice and 1 ppm in milk.40 Our results demonstrate that the SERS method coupled with AuNR substrate was able to meet the MRL requirement of carbaryl that was set by the U.S. EPA in all three types of foods that have been used in this study. The recovery

Figure 7. Calibration curve of carbaryl with concentrations from 50 to 1000 ppb extracted from orange juice using the PLS models.

Figure 6. Root mean square error of prediction (RMSEP) values of carbaryl spectra with different latent variables obtained from the PLS models. G

DOI: 10.1021/acs.jafc.6b04774 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Journal of Agricultural and Food Chemistry



values of the method were calculated on the basis of the PLS model of carbaryl that was extracted from food samples.41 As shown in Table 2, satisfactory recovery values were obtained, ranging from 82 to 99.8% for the samples that were spiked with 600 and 1000 ppb of carbaryl.

carbaryl in acetonitrile (pure pesticide) carbaryl in orange juice carbaryl in grapefruit juice carbaryl in milk

spiked (ppb) 600 1000 600 1000 600 1000 600 1000

quantified (ppb) 515 998 571.1 907.46 561 819 545 975

± ± ± ± ± ± ± ±

4.6 1.9 2.5 5.3 3.2 9.6 3.8 2.1

recovery (%) 86 99.8 95.2 91 93.5 82 91 97.5

In summary, a simple and reproducible method was developed to synthesize the standing AuNR arrays that were used as SERS substrate. It took only 10 min to prepare food samples for measurement by SERS method. The new standing AuNR arrays were able to induce potent electromagnetic field. The AuNRs were utilized in SERS to detect and quantify carbaryl in orange juice, grapefruit juice, and milk. The results show that the SERS method could detect and quantify carbaryl as low as 50 ppb in three types of foods. The detection limits for all three types of foods meet the MRLs of carbaryl. Satisfactory recovery values of carbaryl that was extracted from real food samples were obtained. The results demonstrate that the standing AuNR arrays can be used in the SERS method to detect chemical contaminants in different food products. Future studies are needed to optimize this method using different substrate materials, fabrication methods, organic solvents, and extraction protocols.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jafc.6b04774. Average SERS spectra of carbaryl with concentrations from 50 to 1000 ppb spiked in milk (PDF)



REFERENCES

(1) Cooper, J.; Dobson, H. The benefits of pesticides to mankind and the environment. Crop Prot. 2007, 26, 1337−1348. (2) Elbaz, A.; Clavel, J.; Rathouz, P. J.; Moisan, F.; Galanaud, J.-P.; Delemotte, B.; Alpérovitch, A.; Tzourio, C. Professional exposure to pesticides and Parkinson disease. Ann. Neurol. 2009, 66, 494−504. (3) Baldi, I.; Cantagrel, A.; Lebailly, P.; Tison, F.; Dubroca, B.; Chrysostome, V.; Dartigues, J. F.; Brochard, P. Association between Parkinson’s disease and exposure to pesticides in southwestern France. Neuroepidemiology 2003, 22, 305−310. (4) Grimalt, S.; Pozo, Ó . J.; Sancho, J. V.; Hernández, F. Use of liquid chromatography coupled to quadrupole time-of-flight mass spectrometry to investigate pesticide residues in fruits. Anal. Chem. 2007, 79, 2833−2843. (5) Lin, M.; He, L.; Awika, J.; Yang, L.; Ledoux, D.; Li, H.; Mustapha, A. Detection of melamine in gluten, chicken feed, and processed foods using surface enhanced Raman spectroscopy and HPLC. J. Food Sci. 2008, 73, T129−T134. (6) Raghu, P.; Kumara Swamy, B. E.; Madhusudana Reddy, T.; Chandrashekar, B. N.; Reddaiah, K. Sol−gel immobilized biosensor for the detection of organophosphorous pesticides: a voltammetric method. Bioelectrochemistry 2012, 83, 19−24. (7) Liu, B.; Lin, M.; Li, H. Potential of SERS for rapid detection of melamine and cyanuric acid extracted from milk. Sens. Instrum. Food Qual. Saf. 2010, 4, 13−19. (8) Zhang, Y.; Huang, Y.; Zhai, F.; Du, R.; Liu, Y.; Lai, K. Analyses of enrofloxacin, furazolidone and malachite green in fish products with surface-enhanced Raman spectroscopy. Food Chem. 2012, 135, 845− 850. (9) Alexander, T. A.; Le, D. M. Characterization of a commercialized SERS-active substrate and its application to the identification of intact Bacillus endospores. Appl. Opt. 2007, 46, 3878−3890. (10) Albrecht, M. G.; Creighton, J. A. Anomalously intense Raman spectra of pyridine at a silver electrode. J. Am. Chem. Soc. 1977, 99, 5215−5217. (11) Campion, A.; Kambhampati, P. Surface-enhanced Raman scattering. Chem. Soc. Rev. 1998, 27, 241−250. (12) Fleischmann, M.; Hendra, P. J.; McQuillan, A. J. Raman spectra of pyridine adsorbed at a silver electrode. Chem. Phys. Lett. 1974, 26, 163−166. (13) Fang, Y.; Seong, N.-H.; Dlott, D. D. Measurement of the distribution of site enhancements in surface-enhanced Raman scattering. Science 2008, 321, 388−392. (14) Le Ru, E.; Blackie, E.; Meyer, M.; Etchegoin, P. Surface enhanced Raman scattering enhancement factors: a comprehensive study. J. Phys. Chem. C 2007, 111, 13794−13803. (15) Le Ru, E.; Etchegoin, P.; Meyer, M. Enhancement factor distribution around a single surface-enhanced Raman scattering hot spot and its relation to single molecule detection. J. Chem. Phys. 2006, 125, 204701. (16) Alexander, T. A.; Le, D. M. Characterization of a commercialized SERS-active substrate and its application to the identification of intact Bacillus endospores. Appl. Opt. 2007, 46, 3878− 3890. (17) Liu, B.; Han, G.; Zhang, Z.; Liu, R.; Jiang, C.; Wang, S.; Han, M. Y. Shell thickness-dependent Raman enhancement for rapid identification and detection of pesticide residues at fruit peels. Anal. Chem. 2012, 84, 255−261. (18) Zheng, J.; He, L. Surface-enhanced Raman spectroscopy for the chemical analysis of food. Compr. Rev. Food Sci. Food Saf. 2014, 13, 317−328. (19) Abu Hatab, N. A.; Oran, J. M.; Sepaniak, M. J. Surface-enhanced Raman spectroscopy substrates created via electron beam lithography and nanotransfer printing. ACS Nano 2008, 2, 377−385. (20) Alvarez-Puebla, R.; Cui, B.; Bravo-Vasquez, J.-P.; Veres, T.; Fenniri, H. Nanoimprinted SERS-active substrates with tunable surface plasmon resonances. J. Phys. Chem. C 2007, 111, 6720−6723. (21) Fang, C.; Bandaru, N. M.; Ellis, A. V.; Voelcker, N. H. Betacyclodextrin decorated nanostructured SERS substrates facilitate

Table 2. Recovery Values of Carbaryl Extracted from Food Samples sample

Article

AUTHOR INFORMATION

Corresponding Author

*(M.L.) Phone: (573) 884-6718. Fax: (573) 884-7964. E-mail: [email protected]. ORCID

Mengshi Lin: 0000-0002-6967-2257 Funding

This study was supported by a USDA NIFA Multi-State Project (NC1194). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We acknowledge assistance from Trang Nguyen in electron microscopy analysis. H

DOI: 10.1021/acs.jafc.6b04774 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry selective detection of endocrine disruptorchemicals. Biosens. Bioelectron. 2013, 42, 632−639. (22) Wang, H. H.; Liu, C. Y.; Wu, S. B.; Liu, N. W.; Peng, C. Y.; Chan, T. H.; Hsu, C. F.; Wang, J. K.; Wang, Y. L. Highly Ramanenhancing substrates based on silver nanoparticle arrays with tunable sub-10 nm gaps. Adv. Mater. 2006, 18, 491−495. (23) Lee, S. Y.; Hung, L.; Lang, G. S.; Cornett, J. E.; Mayergoyz, I. D.; Rabin, O. Dispersion in the SERS enhancement with silver nanocube dimers. ACS Nano 2010, 4, 5763−5772. (24) Thai, T.; Zheng, Y.; Ng, S. H.; Mudie, S.; Altissimo, M.; Bach, U. Self-assembly of vertically aligned gold nanorod arrays on patterned substrates. Angew. Chem., Int. Ed. 2012, 51, 8732−8735. (25) Wei, W.; Ge, G. Monolayer of nanorod vertical arrays selfassembled at the air/water interface. Part. Part. Syst. Charact. 2013, 837−841. (26) Wu, H. L.; Tsai, H. R.; Hung, Y. T.; Lao, K. U.; Liao, C. W.; Chung, P. J.; Huang, J. S.; Chen, I. C.; Huang, M. H. A comparative study of gold nanocubes, octahedra, and rhombic dodecahedra as highly sensitive SERS substrates. Inorg. Chem. 2011, 50, 8106−8111. (27) Alvarez-Puebla, R. A.; Agarwal, A.; Manna, P.; Khanal, B. P.; Aldeanueva-Potel, P.; Carbó-Argibay, E.; Pazos-Pérez, N.; Vigderman, L.; Zubarev, E. R.; Kotov, N. A. Gold nanorods 3D-supercrystals as surface enhanced Raman scattering spectroscopy substrates for the rapid detection of scrambled prions. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 8157−8161. (28) Peng, B.; Li, G.; Li, D.; Dodson, S.; Zhang, Q.; Zhang, J.; Lee, Y. H.; Demir, H. V.; Yi Ling, X.; Xiong, Q. Vertically aligned gold nanorod monolayer on arbitrary substrates: self-assembly and femtomolar detection of food contaminants. ACS Nano 2013, 7, 5993−6000. (29) Guerrero-Martínez, A.; Pérez-Juste, J.; Carbó-Argibay, E.; Tardajos, G.; Liz-Marzán, L. M. Gemini-surfactant-directed selfassembly of monodisperse gold nanorods into standing superlattices. Angew. Chem., Int. Ed. 2009, 48, 9484−9488. (30) Xie, Y.; Guo, S.; Guo, C.; He, M.; Chen, D.; Ji, Y.; Chen, Z.; Wu, X.; Liu, Q.; Xie, S. Controllable two-stage droplet evaporation method and its nanoparticles self-assembly mechanism. Langmuir 2013, 29, 6232−6241. (31) Ming, T.; Feng, W.; Tang, Q.; Wang, F.; Sun, L.; Wang, J.; Yan, C. Growth of tetrahexahedral gold nanocrystals with high-index facets. J. Am. Chem. Soc. 2009, 131, 16350−16351. (32) Zhang, Z.; Yu, Q.; Li, H.; Mustapha, A.; Lin, M. Standing gold nanorod arrays as reproducible SERS substrates for measurement of pesticides in apple juice and vegetables. J. Food Sci. 2015, 80, 450−458. (33) Strickland, A. D.; Batt, C. A. Detection of carbendazim by surface-enhanced Raman scattering using cyclodextrin inclusion complexes on gold nanorods. Anal. Chem. 2009, 81, 2895−2903. (34) Lohse, S. E.; Murphy, C. J. The quest for shape control: a history of gold nanorod synthesis. Chem. Mater. 2013, 25, 1250−1261. (35) Zhang, Z.; Lin, M. High-yield preparation of vertically aligned gold nanorod arrays via a controlled evaporation-induced self-assembly method. J. Mater. Chem. C 2014, 2, 4545−4551. (36) Wang, X.; Shi, W.; She, G.; Mu, L.; Lee, S. High-performance surface-enhanced Raman scattering sensors based on Ag nanoparticlescoated Si nanowire arrays for quantitative detection of pesticides. Appl. Phys. Lett. 2010, 96, 053104. (37) Liu, B.; Zhou, P.; Liu, X.; Sun, X.; Li, H.; Lin, M. Detection of pesticides in fruits by surface-enhanced Raman spectroscopy coupled with gold nanostructures. Food Bioprocess Technol. 2013, 6, 710−718. (38) Li, X.; Feng, S.; Hu, Y.; Sheng, W.; Zhang, Y.; Yuan, S.; Zeng, H.; Wang, S.; Lu, X. Rapid detection of melamine in milk using immunological separation and surface enhanced Raman spectroscopy. J. Food Sci. 2015, 80, 1196−1201. (39) Kim, A.; Barcelo, S. J.; Williams, R. S.; Li, Z. Melamine sensing in milk products by using surface enhanced Raman scattering. Anal. Chem. 2012, 84, 9303−9309. (40) EPA Part 180Tolerances and Exemptions for Pesticide Chemical Residues in Food; https://www.gpo.gov/fdsys/pkg/CFR-

2014-title40-vol24/xml/CFR-2014-title40-vol24-part180.xmlseqnum180.103 (accessed Aug 25, 2016). (41) Sentellas, S.; Saurina, J.; Hernández-Cassou, S.; Galceran, M. T.; Puignou, L. s. Multivariate calibration methods for quantification in strongly overlapping capillary electrophoretic peaks. J. Chromatogr. A 2001, 909, 259−269.

I

DOI: 10.1021/acs.jafc.6b04774 J. Agric. Food Chem. XXXX, XXX, XXX−XXX