Rapid Detection of Acetamiprid in Foods using Surface-Enhanced Raman Spectroscopy (SERS) Wisiani Wijaya, Shintaro Pang, Theodore P. Labuza, and Lili He
Acetamiprid is a neonicotinoid pesticide that is commonly used in modern farming. Acetamiprid residue in food commodities can be a potential harm to human and has been implicated in the honey bee hive die off crisis. In this study, we developed rapid, simple, and sensitive methods to detect acetamiprid in apple juice and on apple surfaces using surface-enhanced Raman spectroscopy (SERS). No pretreatment of apple juice sample was performed. A simple surface swab method was used to recover acetamiprid from the apple surface. Samples were incubated with silver dendrites for several minutes and SERS spectra were taken directly from the silver surface. Detection of a set of 5 apple juice samples can be done within 10 min. The swab-SERS method took 15 min for a set of 5 samples. Resulting spectral data were analyzed using principal component analysis. The highest acetamiprid peak at 634 cm−1 was used to detect and quantify the amount of acetamiprid spiked in 1:1 water–methanol solvent, apple juice, and on apple surface. The SERS method was able to successfully detect acetamiprid at 0.5 μg/mL (0.5 ppm) in solvent, 3 μg/mL (3 ppm) in apple juice, and 0.125 μg/cm2 on apple surfaces. The SERS methods provide simple, rapid, and sensitive ways to detect acetamiprid in beverages and on the surfaces of thick skinned fruits and vegetables.
Abstract:
Keywords: acetamiprid, detection, SERS
We developed rapid, simple, and sensitive methods using surface-enhanced Raman spectroscopy to detect acetamiprid in apple juice and on apple surfaces. These methods require no pretreatment to detect acetamiprid in apple juice and combined a simple swab method to detect acetamiprid on apple surfaces. The methods can be modified to detect acetamiprid and other pesticides or toxins in beverages and on the surfaces of thick-skinned fruits and vegetables.
Introduction Acetamiprid is in the class of neonicotinoid pesticides that are commonly used to protect fruits and vegetables from insects. Neonicotinoids are now one of the most commonly used pesticides due to their binding selectivity toward the nicotinic acetylcholine receptor of insects (Tomizawa and Casida 2003). However, acetamiprid residue in foods might be potentially harmful to humans (Kocaman and Topaktas 2007; Brunet and others 2008). The US Environmental Protection Agency (EPA) sets tolerance levels between 0.01 and 50 ppm of acetamiprid residue, depending on the type of food (Electronic Code of Federal Regulations 2013). For apples, as an example, the tolerance level is 1.0 parts per million (1000 μg/kg) using the whole apple slurry after grinding (Electronic Code of Federal Regulations 2013). Acetamiprid residue detection is usually conducted using highpressure liquid chromatography (HPLC). The limit of detection (LOD) by using HPLC can vary depending on the type of HPLC and the detectors (UV or mass spectroscopy) (Martinez and others 2002). However, the drawbacks of using HPLC are that it requires tedious sample preparation and trained personnel. An alternative for rapid detection of pesticide residues is the enzyme-linked immunosorbent assay (ELISA) (Watanabe and others 2006). However, this method is less commonly used than HPLC because it
MS 20131236 Submitted 9/2/2013, Accepted 1/9/2014. Authors Wijaya, Pang, and He are with Dept. of Food Science, Univ. of Massachusetts Amherst, 100 Holdsworth Way, Amherst, MA, 01003, USA. Author Labuza is with Dept. of Food Science and Nutrition, Univ. of Minnesota, 1334 Eckles Avenue, St. Paul, MN, 55108, USA. Direct inquiries to author He (E-mail: [email protected]).
is more susceptible to interference by organic solvents or other components found in foods. Recently, much interest has been given to surface-enhanced Raman spectroscopy (SERS), especially for its capacity for rapid, simple, and sensitive detection, with increased applications in agricultural security and food safety areas (He and others 2011a, 2011b; Craig and others 2013). SERS is a combination technique of Raman spectroscopy and nanotechnology. Raman spectroscopy, a vibrational spectroscopic technique, has become a more widely used experimental method for molecule identification and structural characterization of various compounds. It provides high structural information content, which is a so-called molecular “fingerprint” (McCreery 2000). The time for collecting the spectra is usually a few seconds. However, the major drawback is its low sensitivity. Placement of the analyte on noble metal nanoscale-roughened surfaces (typically silver or gold) enhances the inherently weak Raman molecular signatures tremendously. Parts per billion (ppb) limits of detection have been achieved, and in some cases, a single molecule can be detected using this method (Kneipp and others 1997; Haynes and others 2005). A portable Raman instrument has potential for on-site detection. The first report of using SERS to detect organophosphorus pesticides was in 1987 (Alak and Vo-Dinh 1987). This study demonstrated the potential of real-life detection of accidentally spilled pesticide at a relative high concentration in soil samples (1.25% by weigh). Recently there have been several SERS studies on pesticides. However, most of them have mainly focused on substrate development by testing simple pesticide solutions (Shende and others 2004a, 2004b; Wong-ek and others 2011; Liu and others 2012a, 2012b), and spectral characterization (Guerrini
R
C 2014 Institute of Food Technologists
doi: 10.1111/1750-3841.12391 Further reproduction without permission is prohibited
Vol. 79, Nr. 4, 2014 r Journal of Food Science T743
T: Toxicology & Chemical Food Safety
Practical Application:
SERS detection of acetamiprid in foods. . . and others 2011). For the applications in food matrices, many of the SERS studies still used a complex extraction procedure, which increased the analytical time (Liu and others 2012a, 2012b; Xie and others 2012), or achieved unsatisfactory sensitivity (Shende and others 2004a, 2004b). A recent study reported a rapid SERS detection of 50 ppb chlorpyrifos-methyl artificially added to orange juice in 12 min, including the time for solvent extraction and solid-phase extraction (Shende and others 2010). In this study, we developed simple, rapid, and sensitive SERSbased methods to detect acetamiprid in apple juice with no pretreatment and on apple surface using a simple swab method. To the best of our knowledge, this is the first report of detecting acetamiprid in apple juice and on apple surface using SERS methods.
Materials and Methods Silver dendrite nanosubstrate preparation Silver (Ag) dendrites were prepared by a simple displacement reaction that involves zinc (Fisher Scientific, Pittsburgh, Pa., U.S.A.) and silver nitrate (Fisher Scientific; He and others 2010). Briefly, a zinc plate previously rinsed with HCl (to remove the zinc oxide protective layer) was submerged into a 100 mM silver nitrate solution for exactly 1 min. The dark gray film layer (i.e., silver dendrites) formed on the surface of the zinc plate is then peeled off and rinsed with water several times. Prepared Ag dendrites were stored in deionized water at room temperature prior to use. The prior work showed these were stable for at least 6 months before use. Ag dendrites have been demonstrated to be reliable SERS
substrate for providing satisfactory enhancement and consistent signals due to the unique crystallization of 50 nm Ag nanoparticles that can provide consistent hot spots for SERS analysis (He and others 2010).
SERS sample preparations Detection in water–methanol (1:1). Acetamiprid (Fisher Scientific) was dissolved in deionized water–methanol (1:1) and diluted into 0.1, 0.5, 1, 3, 5, 8, and 10 μg/mL. In a microcentrifuge tube, 10 μL of Ag dendrites were added to 1 mL of each concentration of the spiked samples. The tubes were then vortexed for 30 s, and constantly agitated for 1 min by a Nutating mixer (Fisher Scientific) at 24 rpm. A quick centrifugation, that is 10 s at 2200 × g in the 2 mL tubes, was done to settle the silver dendrites, and then 5 μL of the Ag dendrites from the bottom of the sample tube was transferred onto a glass microscope slide and allowed to dry (3 min at room temperature). SERS spectra were taken directly from the silver surface. Pure solvent (no acetamiprid) was used as a negative control. Detection in apple juice. Apple juice (100% commercial Apple Juice Unsweetened) purchased from a local grocery store was used. Apple juice samples were spiked with acetamiprid to concentrations of 3, 10, and 50 μg/mL. No sample pretreatment was done on the apple juice. Ag dendrites were directly added into apple juice samples, that is same procedure as for water–methanol (1:1), except that constant agitation was done for 5 min. Apple juice without acetamiprid was used as a negative control, and
Figure 1–(A) A raw SERS spectrum of acetamiprid in water–methanol (1:1) solvent and (B) second-derivative SERS spectra of 634 cm−1 peaks at different concentrations (0, 0.1, 0.5, 1, 3, 5, 8, and 10 μg/mL) adsorbed on silver dendrites.
T: Toxicology & Chemical Food Safety
Figure 2–PCA and PLS plots of SERS spectra of different concentrations of acetamiprid in water–methanol (1:1) solvent. For the PCA plot, data points of 0 μg/mL (◦), 0.1 μg/mL (+), 0.5 μg/mL (), 1 μg/mL (•), and 3 μg/mL () were used to determine the limit of detection. For PLS plot, data points of 0, 1, 3, 5, 8, and 10 μg/mL were used to study the linear trend.
T744 Journal of Food Science r Vol. 79, Nr. 4, 2014
SERS detection of acetamiprid in foods. . . nm laser and a 10× confocal microscope objective was used in this study. Each spectrum was scanned from 2000 to 400 cm−1 with 2 mW laser power and a 25-μm slit width for 5 seconds integration time. Eight scans were done automatically from different spots on the surface and then averaged by the instrument before analysis.
Data analysis SERS spectral data were analyzed using TQ Analyst Software (Thermo Fisher Scientific). The spectra were processed using second-derivative transformation to remove baseline and separate overlapped peaks. Outlier peaks were removed to justify for a more accurate quantification of peak to concentration ratio. Detection of acetamiprid was determined using principal component analysis (PCA). The PCA procedure reduces a multidimensional data set to its most dominant features, removes random variation, and retains the principal components (PCs) that capture the variation between sample treatments. The information provided by the PCA shows the variance within a class and between different classes. The PC score reveals the percentage of data variance. Spectra of different concentrations of acetamiprid were analyzed by partial least squares (PLS). PLS was constructed by calibrating
Figure 3–(A) Second-derivative SERS spectra of the control water:methanol solution at 50 μg/mL (light blue line) versus the second-derivative SERS spectra of acetamiprid at 0, 3, 10, and 50 μg/mL in apple juice (average of all data points). Raw spectra is shown in Figure S1. (B) Second-derivative SERS spectra of acetamiprid in apple juice zoomed in the range of 610 to 640 cm−1 at full scale. (C) PCA plot of different acetamiprid concentrations in apple juice analyzed. Vol. 79, Nr. 4, 2014 r Journal of Food Science T745
T: Toxicology & Chemical Food Safety
50 μg/mL acetamiprid in deionized water–methanol (1:1) was used as a positive control. The testing was done in duplicate. Detection on apple surface. The apples (Purity Organic Fuji) were purchased from a local grocery store. Swab sticks (ITW R , Kernersville, N.C., U.S.A. diameter 0.25 in.) were Texwipe from Fisher Scientific. Acetamiprid was diluted in water–methanol (1:1) to final concentrations of 10, 50, and 100 μg/mL. Fifty microliters of each concentration of acetamiprid were transferred onto the apple surface, and allowed to dry for 3 min at room conditions. Swab sticks were soaked in water–methanol (1:1) for 10 s prior to using. The acetamiprid-exposed apple surface (2 × 2 cm) was swabbed for 30 s. The swab was then soaked and vortexed in 1 mL of water–methanol (1:1) for 3 min. The same procedure for Raman analysis as performed in water–methanol (1:1) was then applied. The final amount of acetamiprid exposed to the apple surface was calculated to be 0.5, 2.5, and 5 μg (0.125, 0.625, and 1.25 μg/cm2 ). Apple surface swab without acetamiprid was used as a negative control, and 50 μg/mL acetamiprid in deionized water–methanol (1:1) was used as a positive control. The tests were done in triplicates. Raman instrumentation. A DXR Raman microscope (Thermo Fisher Scientific, Madison, Wis., U.S.A.) with a 780-
SERS detection of acetamiprid in foods. . . sample concentrations based on the spectral information and their actual (spiked) values. The higher the correlation coefficient and the lower the root mean square error of calibration (RMSEC), the better quality of the model. The model was also self-validated using leave-one-out validation method, which uses all but one sample to build a calibration model and repeats for each sample in the data set. The closer the range between the root mean square error of validation (RMSECV) and RMSEV values, the better robustness of the model. Prior to PCA and PLS, standard normal variance, smoothing, and second-derivative transformation were done to provide better discrimination (Wu and others 1995).
Results and Discussion Detection in water–methanol (1:1) The SERS fingerprint pattern of acetamiprid on silver dendrites is shown in Figure 1A. The highest peak was observed at 634 cm−1 , which is associated with the ring structure of the molecule. Eight different concentrations of acetamiprid in deionized water–methanol (1:1) were analyzed to determine the acetamiprid limit of detection (LOD) and linear range using the SERS method. The result (Figure 1B) showed that the concentration of acetamiprid is positively correlated to the acetamiprid peak intensity at 634 cm−1 in the SERS spectra. PCA was used to estimate LOD of acetamiprid in the solvent. According to Figure 2A, results at 0 and 0.1 μg/mL acetamiprid were clustered together; separation from 0 was not observed until a concentration of 0.5 μg/mL acetamiprid, which indicates that the LOD was about 0.5 μg/mL. The higher the concentration of acetamiprid, the bigger the PCA score separation distance from the negative control. The PLS plot (Figure 2B) shows a good linear relationship between 0 and 10 μg/mL with a RMSEC of 0.683 and correlation coefficient at 0.982. The RMSECV obtained using the leave-one-out validation method was 1.14 μg/mL (Figure S2). This result demonstrated the potential quantitative capacity of the SERS method for detection of acetamripid. Although the LOD of the SERS method is lower than many chromatographic methods for acetamiprid (0.001 to 0.01 mg/kg) (Obana and others 2002; Hirotaka and others 2003; Mateu-Sanchez and others 2003) and ELISA (0.053 ng/g) (Watanabe and others 2006), the SERS method was much simpler and faster.
juice, significant apple juice peaks were observed, for example at 732 cm−1 . With increasing concentration of acetamiprid, the acetamiprid “fingerprint” peak intensity increased, noticeably at high concentration (50 μg/mL) while there were no intensity differences at low concentrations (3 and 10 μg/mL). However, the acetamiprid concentration was found to correlate with the shifting of the acetamiprid peak. By zooming into around 610 to 640 cm−1 at full scale (Figure 3B), it is clear that the apple juice samples that contain different concentrations of acetamiprid were “sandwiched” in between the pure apple juice peak (negative control) and 50 μg/mL acetamiprid in solvent peak (positive control). The higher the concentration of acetamiprid in the apple juice sample, the more the shift to the left is observed. This could be due to the competition between the added acetamiprid and the apple juice components for the surface of Ag dendrites. More studies will be needed to elucidate the detailed mechanism. The secondderivative PCA of apple juice samples (Figure 3C) showed a clear separation of data points between the Ag dendrites control and 3 μg/mL (ppm) or higher acetamiprid concentration. This suggests that at least 3 μg/mL (ppm) of acetamiprid can be detected in apple juice using the SERS method without sample preparation. Because no sample pretreatment was done and the apple juice components will compete with acetamiprid, the recovery can be approximately estimated by Raman shift(50 μg/mL in apple juice−0 μg/mL) = 69.2%. Detection of a set Raman shift(50 μg/mL in solvent−0 μg/mL) of 5 apple juice samples can be done within 10 min.
Detection on apple surface The apple peel surface was spiked with 3 different amounts of acetamiprid: 0.5, 2.5, and 5.0 μg (0.125, 0.625, and 1.25 μg/cm2 ). Acetamiprid was recovered using a simple swab method. No significant background interference was observed in the apple peel surface samples (data not shown). The acetamiprid “fingerprint” peak at 634 cm−1 was observed, and its peak intensity was positively correlated to the spiked amount (Figure 4A). In the PCA plot (Figure 4B), the minimum amount of concentration that can be differentiated from the negative control (0 μg) was the sample that had 0.5 μg (0.125 μg/cm2 ) acetamiprid. There was no clear separation between the 0.5, 2.5, and 5.0 μg data, although a clear trend was observed; that is, the higher the concentration of acetamiprid, the bigger the PCA score separation Detection in apple juice distance from the negative control. This was probably due to the For the second-derivative spectra of the apple juice samples apple surface variation and manual swabbing variation in the trip(Figure 3A), because no pretreatment was done to the apple licate experiment. The approximate recovery can be estimated
Figure 4–(A) Second-derivative SERS spectra and (B) PCA plot of different amounts of acetamiprid swabbed from apple surface.
T: Toxicology & Chemical Food Safety
T746 Journal of Food Science r Vol. 79, Nr. 4, 2014
SERS detection of acetamiprid in foods. . . hight(5 μg swabbed from surface−0 μg ) by peakpeak × 100% = 90.0%. Considhight(5 μg/mL in solvent−0 μg) ering the apple weight was about 200 g, the LOD transferring into EPA units is 0.0025 ppm or 2.5 μg/kg, much lower than the tolerance level of 1000 μg/kg. The swab-SERS method took 15 min to test a set of 5 samples.
Conclusions In conclusion, detection of acetamiprid using a SERS method was successfully done in water–methanol (1:1) solvent, apple juice, and from swabs of the apple surface. No pretreatment was done to the apple juice samples. Thus, some background interference peaks in the second-derivative spectra were observed. Less background interference was observed with the apple peel surface using a swabbing technique for sample preparation. Detection of acetamiprid was based on its “fingerprint” peak at around 634 cm−1 and the LODs were 0.5 μg/mL in water–methanol (1:1) solvent, 3 μg/mL in apple juice, and 0.125 μg/cm2 on apple peel surface swab. The SERS method provides rapid, simple, and sensitive ways to detect pesticides in foods; they can also be modified to detect other pesticides. A portable Raman instrument can potentially be used for on-site pesticides detection, which will be evaluated in future studies.
Acknowledgment This research was supported by the U.S. Dept. of Agriculture (USDA)–National Inst. of Food and Agriculture (NIFA) #201267017-30194. The authors declare that there are no conflicts of interest.
Kneipp K, Wang Y, Kneipp H, Perelman LT, Itzkan I, Dasari RR, Feld MS. 1997. Single molecule detection using surface-enhanced Raman scattering (SERS). Phys Rev Lett 78:1667–70. Kocaman AY, Topaktas M. 2007. In vitro evaluation of the genotoxicity of acetamiprid. Environ Mol Mutagen 490:483–90. Liu B, Han G, Zhang Z, Liu R, Jiang C, Wang S, Han MY. 2012a. Shell thickness-dependent Raman enhancement for rapid identification and detection of pesticide residues at fruit peels. Anal Chem 84:255–61. Liu B, Zhou P, Liu X, Sun X, Li H, Lin M. 2012b. Detection of pesticides in fruits by surface-enhanced raman spectroscopy coupled with gold nanostructures. Food Bioprocess Tech 6:710–8. Martinez Vidal JL, Gill Garcia MD, Martinez Galera M, Lopez Lopez T. 2002. Determination of acetamiprid by HPLC-fluorescence with post-column photoderivatization and HPLC-MASS selection determination. J Liq Chromatogr Relat Technol 25:2695–07. Mateu-Sanchez M, Moreno M, Arebola FJ, Mart´ınez Vidal, LJ. 2003. Analysis of acetamiprid in vegetables using gas chromatography–tandem mass spectrometry. Anal Sci 19: 701–4. McCreery RL. 2000. Raman spectroscopy for chemical analysis. Canada: Wiley-Interscience. 448 p. Obana H, Okihashi M, Akutsu K, Kitagawa Y, Hori S. 2002. Determination of acetamiprid, imidacloprid, and nitenpyram residues in vegetables and fruits by high-performance liquid chromatography with diode-array detection. J Agric Food Chem 50:4464–67. Shende C, Gift A, Inscore F, Maksymiuk P, Farquharson S. 2004a. Inspection of pesticide residues on food by surface-enhanced Raman spectroscopy. Proc SPIE 5271:28–34. Shende CS, Inscore F, Gift A, Maksymiuk, P, Farquharson S. 2004b. Analysis of pesticides on or in fruit by surface-enhanced Raman spectroscopy. Proc SPIE 5587:170–6. Shende S, Inscore F, Sengupta A, Stuart J, Farquharson S. 2010. Rapid extraction and detection of trace chlorpyrifos-methyl in orange juice by surface-enhanced Raman spectroscopy. Sens Instrum Food Qual 4:101–107. Tomizawa M, Casida JE. 2003. Selective toxicity of neonicotinoids attributable to specificity of insect and mammalian nicotinic receptors. Annu Rev Entomol 48:339–64. Watanabe E, Miyake S, Baba K, Eun H, Endo S. 2006. Immunoassay for acetamiprid detection: application to residue analysis and comparison with liquid chromatography. Anal Bioanal Chem 386:1441–8. Wong-ek K, Horprathum M, Eiamchai P, Limnonthakul P, Patthanasettakul V, Chindaudom P, Nuntawong N. 2011. Portable surface-enhanced Raman spectroscopy for insecticide detection using silver nanorod film fabricated by magnetron sputtering. Proc SPIE 7911:791108–1–11. Wu W, Walczak B, Massart DL, Prebble KA, Last IR. 1995. Spectral transformation and wavelength selection in near-infrared spectra classification. Anal Chim Acta 315: 243–55. Xie Y, Mukamurezi G, Sun Y, Wang H, Qian H, Yao W. 2012. Establishment of rapid detection method of methamidophos in vegetables by surface enhanced Raman spectroscopy. Eur Food Res Technol 234:1091–8.
References
Supporting Information Additional Supporting Information may be found in the online version of this article at the publisher’s website: Figure S1. The raw spectra of the control water:methanol solution at 50 μg/mL (dark blue line) vs the raw SERS spectra of acetamiprid at 0, 3, 10, and 50 μg/mL in apple juice (average of all data points). The second derivative SERS spectra is shown on the manuscript in Figure 3a; The second derivative spectra was used in the manuscript to help reduce noise and remove baseline, making it easier to see the differences between spectra. Figure S2. The PLS plot with the root mean square error coefficient of variation (RMSECV) of SERS spectra of different concentrations of acetamiprid in water-methanol (1:1) solvent where RMSECV = 1.14 μg/mL. Data points of 0, 1, 3, 5, 8, 10 μg/mL were used.
Vol. 79, Nr. 4, 2014 r Journal of Food Science T747
T: Toxicology & Chemical Food Safety
Alak AM, Vo-Dinh T. 1987. Surface-enhanced Raman spectroscopy of organophosphorous chemical agents. Anal Chem 59:2149–53. Brunet JL, Maresca M, Fantini J, Belzunces LP. 2008. Intestinal absorption of the acetamiprid neonicotinoid by Caco-2 cells: transepithelial transport, cellular uptake and efflux. J Environ Sci Health, Part B 43:261–70. Craig AP, Franca AS, Irudayaraj J. 2013. Surface-enhanced Raman spectroscopy applied to food safety. Annu Rev Food Sci Technol 4:369–80. Electronic Code of Federal Regulations. 2013. 40 CFR 180.578, Acetamiprid; tolerances for residues. Available from: http://www.ecfr.gov/cgi-bin/text-idx?c=ecfr&SID= 7d60c7b05ece6fdddaa11b861770fbc5&rgn=div8&view=text&node=40:25.0.1.1.27.3.19. 299&idno=40. Accessed 2013 August 30. Guerrini L, Sanchez-Cortes S, Cruz VL, Martinez S, Ristori S, Feis A. 2011. Surface-enhanced Raman spectra of dimethoate and omethoate. J Raman Spectrosc 42:980–5. Haynes CL, McFarland AD, Van Duyne RP. 2005. Surface-enhanced Raman spectroscopy. Anal Chem 77:338a–46a. He L, Lin M, Li H, Kim NJ. 2010. Surface-enhanced Raman spectroscopy coupled with dendritic silver nanosubstrate for detection of restricted antibiotics. J Raman Spectrosc 41:739– 44. He L, Lamont E, Veeregowda B, Sreevatsan S, Haynes CL, Diez-Gonzalez F, Labuza TP. 2011a. Aptamer-based surface-enhanced Raman scattering detection of ricin in liquid foods. Chem Sci 2:1579–82. He L, Rodda T, Haynes CL, Deschaines T, Strother T, Diez-Gonzalez F, Labuza TP. 2011b. Detection of a foreign protein in milk using surface-enhanced Raman spectroscopy coupled with antibody-modified silver dendrites. Anal Chem 83:1510–3. Hirotaka O, Okihashi M, Kazuhiko A, Kitagawa Y, Hori S. 2003. Determination of neonicotinoid pesticide residues in chromatography mass spectrometry. J Agric Food Chem 51: 2501–5.
Acetamiprid is a neonicotinoid pesticide that is commonly used in modern farming. Acetamiprid residue in food commodities can be a potential harm to human and has been implicated in the honey bee hive die off crisis. In this study, we developed rapid, simple, and sensitive methods to detect acetamiprid in apple juice and on apple surfaces using surface-enhanced Raman spectroscopy (SERS). No pretreatment of apple juice sample was performed. A simple surface swab method was used to recover acetamiprid from the apple surface. Samples were incubated with silver dendrites for several minutes and SERS spectra were taken directly from the silver surface. Detection of a set of 5 apple juice samples can be done within 10 min. The swab-SERS method took 15 min for a set of 5 samples. Resulting spectral data were analyzed using principal component analysis. The highest acetamiprid peak at 634 cm−1 was used to detect and quantify the amount of acetamiprid spiked in 1:1 water–methanol solvent, apple juice, and on apple surface. The SERS method was able to successfully detect acetamiprid at 0.5 μg/mL (0.5 ppm) in solvent, 3 μg/mL (3 ppm) in apple juice, and 0.125 μg/cm2 on apple surfaces. The SERS methods provide simple, rapid, and sensitive ways to detect acetamiprid in beverages and on the surfaces of thick skinned fruits and vegetables.
Abstract:
Keywords: acetamiprid, detection, SERS
We developed rapid, simple, and sensitive methods using surface-enhanced Raman spectroscopy to detect acetamiprid in apple juice and on apple surfaces. These methods require no pretreatment to detect acetamiprid in apple juice and combined a simple swab method to detect acetamiprid on apple surfaces. The methods can be modified to detect acetamiprid and other pesticides or toxins in beverages and on the surfaces of thick-skinned fruits and vegetables.
Introduction Acetamiprid is in the class of neonicotinoid pesticides that are commonly used to protect fruits and vegetables from insects. Neonicotinoids are now one of the most commonly used pesticides due to their binding selectivity toward the nicotinic acetylcholine receptor of insects (Tomizawa and Casida 2003). However, acetamiprid residue in foods might be potentially harmful to humans (Kocaman and Topaktas 2007; Brunet and others 2008). The US Environmental Protection Agency (EPA) sets tolerance levels between 0.01 and 50 ppm of acetamiprid residue, depending on the type of food (Electronic Code of Federal Regulations 2013). For apples, as an example, the tolerance level is 1.0 parts per million (1000 μg/kg) using the whole apple slurry after grinding (Electronic Code of Federal Regulations 2013). Acetamiprid residue detection is usually conducted using highpressure liquid chromatography (HPLC). The limit of detection (LOD) by using HPLC can vary depending on the type of HPLC and the detectors (UV or mass spectroscopy) (Martinez and others 2002). However, the drawbacks of using HPLC are that it requires tedious sample preparation and trained personnel. An alternative for rapid detection of pesticide residues is the enzyme-linked immunosorbent assay (ELISA) (Watanabe and others 2006). However, this method is less commonly used than HPLC because it
MS 20131236 Submitted 9/2/2013, Accepted 1/9/2014. Authors Wijaya, Pang, and He are with Dept. of Food Science, Univ. of Massachusetts Amherst, 100 Holdsworth Way, Amherst, MA, 01003, USA. Author Labuza is with Dept. of Food Science and Nutrition, Univ. of Minnesota, 1334 Eckles Avenue, St. Paul, MN, 55108, USA. Direct inquiries to author He (E-mail: [email protected]).
is more susceptible to interference by organic solvents or other components found in foods. Recently, much interest has been given to surface-enhanced Raman spectroscopy (SERS), especially for its capacity for rapid, simple, and sensitive detection, with increased applications in agricultural security and food safety areas (He and others 2011a, 2011b; Craig and others 2013). SERS is a combination technique of Raman spectroscopy and nanotechnology. Raman spectroscopy, a vibrational spectroscopic technique, has become a more widely used experimental method for molecule identification and structural characterization of various compounds. It provides high structural information content, which is a so-called molecular “fingerprint” (McCreery 2000). The time for collecting the spectra is usually a few seconds. However, the major drawback is its low sensitivity. Placement of the analyte on noble metal nanoscale-roughened surfaces (typically silver or gold) enhances the inherently weak Raman molecular signatures tremendously. Parts per billion (ppb) limits of detection have been achieved, and in some cases, a single molecule can be detected using this method (Kneipp and others 1997; Haynes and others 2005). A portable Raman instrument has potential for on-site detection. The first report of using SERS to detect organophosphorus pesticides was in 1987 (Alak and Vo-Dinh 1987). This study demonstrated the potential of real-life detection of accidentally spilled pesticide at a relative high concentration in soil samples (1.25% by weigh). Recently there have been several SERS studies on pesticides. However, most of them have mainly focused on substrate development by testing simple pesticide solutions (Shende and others 2004a, 2004b; Wong-ek and others 2011; Liu and others 2012a, 2012b), and spectral characterization (Guerrini
R
C 2014 Institute of Food Technologists
doi: 10.1111/1750-3841.12391 Further reproduction without permission is prohibited
Vol. 79, Nr. 4, 2014 r Journal of Food Science T743
T: Toxicology & Chemical Food Safety
Practical Application:
SERS detection of acetamiprid in foods. . . and others 2011). For the applications in food matrices, many of the SERS studies still used a complex extraction procedure, which increased the analytical time (Liu and others 2012a, 2012b; Xie and others 2012), or achieved unsatisfactory sensitivity (Shende and others 2004a, 2004b). A recent study reported a rapid SERS detection of 50 ppb chlorpyrifos-methyl artificially added to orange juice in 12 min, including the time for solvent extraction and solid-phase extraction (Shende and others 2010). In this study, we developed simple, rapid, and sensitive SERSbased methods to detect acetamiprid in apple juice with no pretreatment and on apple surface using a simple swab method. To the best of our knowledge, this is the first report of detecting acetamiprid in apple juice and on apple surface using SERS methods.
Materials and Methods Silver dendrite nanosubstrate preparation Silver (Ag) dendrites were prepared by a simple displacement reaction that involves zinc (Fisher Scientific, Pittsburgh, Pa., U.S.A.) and silver nitrate (Fisher Scientific; He and others 2010). Briefly, a zinc plate previously rinsed with HCl (to remove the zinc oxide protective layer) was submerged into a 100 mM silver nitrate solution for exactly 1 min. The dark gray film layer (i.e., silver dendrites) formed on the surface of the zinc plate is then peeled off and rinsed with water several times. Prepared Ag dendrites were stored in deionized water at room temperature prior to use. The prior work showed these were stable for at least 6 months before use. Ag dendrites have been demonstrated to be reliable SERS
substrate for providing satisfactory enhancement and consistent signals due to the unique crystallization of 50 nm Ag nanoparticles that can provide consistent hot spots for SERS analysis (He and others 2010).
SERS sample preparations Detection in water–methanol (1:1). Acetamiprid (Fisher Scientific) was dissolved in deionized water–methanol (1:1) and diluted into 0.1, 0.5, 1, 3, 5, 8, and 10 μg/mL. In a microcentrifuge tube, 10 μL of Ag dendrites were added to 1 mL of each concentration of the spiked samples. The tubes were then vortexed for 30 s, and constantly agitated for 1 min by a Nutating mixer (Fisher Scientific) at 24 rpm. A quick centrifugation, that is 10 s at 2200 × g in the 2 mL tubes, was done to settle the silver dendrites, and then 5 μL of the Ag dendrites from the bottom of the sample tube was transferred onto a glass microscope slide and allowed to dry (3 min at room temperature). SERS spectra were taken directly from the silver surface. Pure solvent (no acetamiprid) was used as a negative control. Detection in apple juice. Apple juice (100% commercial Apple Juice Unsweetened) purchased from a local grocery store was used. Apple juice samples were spiked with acetamiprid to concentrations of 3, 10, and 50 μg/mL. No sample pretreatment was done on the apple juice. Ag dendrites were directly added into apple juice samples, that is same procedure as for water–methanol (1:1), except that constant agitation was done for 5 min. Apple juice without acetamiprid was used as a negative control, and
Figure 1–(A) A raw SERS spectrum of acetamiprid in water–methanol (1:1) solvent and (B) second-derivative SERS spectra of 634 cm−1 peaks at different concentrations (0, 0.1, 0.5, 1, 3, 5, 8, and 10 μg/mL) adsorbed on silver dendrites.
T: Toxicology & Chemical Food Safety
Figure 2–PCA and PLS plots of SERS spectra of different concentrations of acetamiprid in water–methanol (1:1) solvent. For the PCA plot, data points of 0 μg/mL (◦), 0.1 μg/mL (+), 0.5 μg/mL (), 1 μg/mL (•), and 3 μg/mL () were used to determine the limit of detection. For PLS plot, data points of 0, 1, 3, 5, 8, and 10 μg/mL were used to study the linear trend.
T744 Journal of Food Science r Vol. 79, Nr. 4, 2014
SERS detection of acetamiprid in foods. . . nm laser and a 10× confocal microscope objective was used in this study. Each spectrum was scanned from 2000 to 400 cm−1 with 2 mW laser power and a 25-μm slit width for 5 seconds integration time. Eight scans were done automatically from different spots on the surface and then averaged by the instrument before analysis.
Data analysis SERS spectral data were analyzed using TQ Analyst Software (Thermo Fisher Scientific). The spectra were processed using second-derivative transformation to remove baseline and separate overlapped peaks. Outlier peaks were removed to justify for a more accurate quantification of peak to concentration ratio. Detection of acetamiprid was determined using principal component analysis (PCA). The PCA procedure reduces a multidimensional data set to its most dominant features, removes random variation, and retains the principal components (PCs) that capture the variation between sample treatments. The information provided by the PCA shows the variance within a class and between different classes. The PC score reveals the percentage of data variance. Spectra of different concentrations of acetamiprid were analyzed by partial least squares (PLS). PLS was constructed by calibrating
Figure 3–(A) Second-derivative SERS spectra of the control water:methanol solution at 50 μg/mL (light blue line) versus the second-derivative SERS spectra of acetamiprid at 0, 3, 10, and 50 μg/mL in apple juice (average of all data points). Raw spectra is shown in Figure S1. (B) Second-derivative SERS spectra of acetamiprid in apple juice zoomed in the range of 610 to 640 cm−1 at full scale. (C) PCA plot of different acetamiprid concentrations in apple juice analyzed. Vol. 79, Nr. 4, 2014 r Journal of Food Science T745
T: Toxicology & Chemical Food Safety
50 μg/mL acetamiprid in deionized water–methanol (1:1) was used as a positive control. The testing was done in duplicate. Detection on apple surface. The apples (Purity Organic Fuji) were purchased from a local grocery store. Swab sticks (ITW R , Kernersville, N.C., U.S.A. diameter 0.25 in.) were Texwipe from Fisher Scientific. Acetamiprid was diluted in water–methanol (1:1) to final concentrations of 10, 50, and 100 μg/mL. Fifty microliters of each concentration of acetamiprid were transferred onto the apple surface, and allowed to dry for 3 min at room conditions. Swab sticks were soaked in water–methanol (1:1) for 10 s prior to using. The acetamiprid-exposed apple surface (2 × 2 cm) was swabbed for 30 s. The swab was then soaked and vortexed in 1 mL of water–methanol (1:1) for 3 min. The same procedure for Raman analysis as performed in water–methanol (1:1) was then applied. The final amount of acetamiprid exposed to the apple surface was calculated to be 0.5, 2.5, and 5 μg (0.125, 0.625, and 1.25 μg/cm2 ). Apple surface swab without acetamiprid was used as a negative control, and 50 μg/mL acetamiprid in deionized water–methanol (1:1) was used as a positive control. The tests were done in triplicates. Raman instrumentation. A DXR Raman microscope (Thermo Fisher Scientific, Madison, Wis., U.S.A.) with a 780-
SERS detection of acetamiprid in foods. . . sample concentrations based on the spectral information and their actual (spiked) values. The higher the correlation coefficient and the lower the root mean square error of calibration (RMSEC), the better quality of the model. The model was also self-validated using leave-one-out validation method, which uses all but one sample to build a calibration model and repeats for each sample in the data set. The closer the range between the root mean square error of validation (RMSECV) and RMSEV values, the better robustness of the model. Prior to PCA and PLS, standard normal variance, smoothing, and second-derivative transformation were done to provide better discrimination (Wu and others 1995).
Results and Discussion Detection in water–methanol (1:1) The SERS fingerprint pattern of acetamiprid on silver dendrites is shown in Figure 1A. The highest peak was observed at 634 cm−1 , which is associated with the ring structure of the molecule. Eight different concentrations of acetamiprid in deionized water–methanol (1:1) were analyzed to determine the acetamiprid limit of detection (LOD) and linear range using the SERS method. The result (Figure 1B) showed that the concentration of acetamiprid is positively correlated to the acetamiprid peak intensity at 634 cm−1 in the SERS spectra. PCA was used to estimate LOD of acetamiprid in the solvent. According to Figure 2A, results at 0 and 0.1 μg/mL acetamiprid were clustered together; separation from 0 was not observed until a concentration of 0.5 μg/mL acetamiprid, which indicates that the LOD was about 0.5 μg/mL. The higher the concentration of acetamiprid, the bigger the PCA score separation distance from the negative control. The PLS plot (Figure 2B) shows a good linear relationship between 0 and 10 μg/mL with a RMSEC of 0.683 and correlation coefficient at 0.982. The RMSECV obtained using the leave-one-out validation method was 1.14 μg/mL (Figure S2). This result demonstrated the potential quantitative capacity of the SERS method for detection of acetamripid. Although the LOD of the SERS method is lower than many chromatographic methods for acetamiprid (0.001 to 0.01 mg/kg) (Obana and others 2002; Hirotaka and others 2003; Mateu-Sanchez and others 2003) and ELISA (0.053 ng/g) (Watanabe and others 2006), the SERS method was much simpler and faster.
juice, significant apple juice peaks were observed, for example at 732 cm−1 . With increasing concentration of acetamiprid, the acetamiprid “fingerprint” peak intensity increased, noticeably at high concentration (50 μg/mL) while there were no intensity differences at low concentrations (3 and 10 μg/mL). However, the acetamiprid concentration was found to correlate with the shifting of the acetamiprid peak. By zooming into around 610 to 640 cm−1 at full scale (Figure 3B), it is clear that the apple juice samples that contain different concentrations of acetamiprid were “sandwiched” in between the pure apple juice peak (negative control) and 50 μg/mL acetamiprid in solvent peak (positive control). The higher the concentration of acetamiprid in the apple juice sample, the more the shift to the left is observed. This could be due to the competition between the added acetamiprid and the apple juice components for the surface of Ag dendrites. More studies will be needed to elucidate the detailed mechanism. The secondderivative PCA of apple juice samples (Figure 3C) showed a clear separation of data points between the Ag dendrites control and 3 μg/mL (ppm) or higher acetamiprid concentration. This suggests that at least 3 μg/mL (ppm) of acetamiprid can be detected in apple juice using the SERS method without sample preparation. Because no sample pretreatment was done and the apple juice components will compete with acetamiprid, the recovery can be approximately estimated by Raman shift(50 μg/mL in apple juice−0 μg/mL) = 69.2%. Detection of a set Raman shift(50 μg/mL in solvent−0 μg/mL) of 5 apple juice samples can be done within 10 min.
Detection on apple surface The apple peel surface was spiked with 3 different amounts of acetamiprid: 0.5, 2.5, and 5.0 μg (0.125, 0.625, and 1.25 μg/cm2 ). Acetamiprid was recovered using a simple swab method. No significant background interference was observed in the apple peel surface samples (data not shown). The acetamiprid “fingerprint” peak at 634 cm−1 was observed, and its peak intensity was positively correlated to the spiked amount (Figure 4A). In the PCA plot (Figure 4B), the minimum amount of concentration that can be differentiated from the negative control (0 μg) was the sample that had 0.5 μg (0.125 μg/cm2 ) acetamiprid. There was no clear separation between the 0.5, 2.5, and 5.0 μg data, although a clear trend was observed; that is, the higher the concentration of acetamiprid, the bigger the PCA score separation Detection in apple juice distance from the negative control. This was probably due to the For the second-derivative spectra of the apple juice samples apple surface variation and manual swabbing variation in the trip(Figure 3A), because no pretreatment was done to the apple licate experiment. The approximate recovery can be estimated
Figure 4–(A) Second-derivative SERS spectra and (B) PCA plot of different amounts of acetamiprid swabbed from apple surface.
T: Toxicology & Chemical Food Safety
T746 Journal of Food Science r Vol. 79, Nr. 4, 2014
SERS detection of acetamiprid in foods. . . hight(5 μg swabbed from surface−0 μg ) by peakpeak × 100% = 90.0%. Considhight(5 μg/mL in solvent−0 μg) ering the apple weight was about 200 g, the LOD transferring into EPA units is 0.0025 ppm or 2.5 μg/kg, much lower than the tolerance level of 1000 μg/kg. The swab-SERS method took 15 min to test a set of 5 samples.
Conclusions In conclusion, detection of acetamiprid using a SERS method was successfully done in water–methanol (1:1) solvent, apple juice, and from swabs of the apple surface. No pretreatment was done to the apple juice samples. Thus, some background interference peaks in the second-derivative spectra were observed. Less background interference was observed with the apple peel surface using a swabbing technique for sample preparation. Detection of acetamiprid was based on its “fingerprint” peak at around 634 cm−1 and the LODs were 0.5 μg/mL in water–methanol (1:1) solvent, 3 μg/mL in apple juice, and 0.125 μg/cm2 on apple peel surface swab. The SERS method provides rapid, simple, and sensitive ways to detect pesticides in foods; they can also be modified to detect other pesticides. A portable Raman instrument can potentially be used for on-site pesticides detection, which will be evaluated in future studies.
Acknowledgment This research was supported by the U.S. Dept. of Agriculture (USDA)–National Inst. of Food and Agriculture (NIFA) #201267017-30194. The authors declare that there are no conflicts of interest.
Kneipp K, Wang Y, Kneipp H, Perelman LT, Itzkan I, Dasari RR, Feld MS. 1997. Single molecule detection using surface-enhanced Raman scattering (SERS). Phys Rev Lett 78:1667–70. Kocaman AY, Topaktas M. 2007. In vitro evaluation of the genotoxicity of acetamiprid. Environ Mol Mutagen 490:483–90. Liu B, Han G, Zhang Z, Liu R, Jiang C, Wang S, Han MY. 2012a. Shell thickness-dependent Raman enhancement for rapid identification and detection of pesticide residues at fruit peels. Anal Chem 84:255–61. Liu B, Zhou P, Liu X, Sun X, Li H, Lin M. 2012b. Detection of pesticides in fruits by surface-enhanced raman spectroscopy coupled with gold nanostructures. Food Bioprocess Tech 6:710–8. Martinez Vidal JL, Gill Garcia MD, Martinez Galera M, Lopez Lopez T. 2002. Determination of acetamiprid by HPLC-fluorescence with post-column photoderivatization and HPLC-MASS selection determination. J Liq Chromatogr Relat Technol 25:2695–07. Mateu-Sanchez M, Moreno M, Arebola FJ, Mart´ınez Vidal, LJ. 2003. Analysis of acetamiprid in vegetables using gas chromatography–tandem mass spectrometry. Anal Sci 19: 701–4. McCreery RL. 2000. Raman spectroscopy for chemical analysis. Canada: Wiley-Interscience. 448 p. Obana H, Okihashi M, Akutsu K, Kitagawa Y, Hori S. 2002. Determination of acetamiprid, imidacloprid, and nitenpyram residues in vegetables and fruits by high-performance liquid chromatography with diode-array detection. J Agric Food Chem 50:4464–67. Shende C, Gift A, Inscore F, Maksymiuk P, Farquharson S. 2004a. Inspection of pesticide residues on food by surface-enhanced Raman spectroscopy. Proc SPIE 5271:28–34. Shende CS, Inscore F, Gift A, Maksymiuk, P, Farquharson S. 2004b. Analysis of pesticides on or in fruit by surface-enhanced Raman spectroscopy. Proc SPIE 5587:170–6. Shende S, Inscore F, Sengupta A, Stuart J, Farquharson S. 2010. Rapid extraction and detection of trace chlorpyrifos-methyl in orange juice by surface-enhanced Raman spectroscopy. Sens Instrum Food Qual 4:101–107. Tomizawa M, Casida JE. 2003. Selective toxicity of neonicotinoids attributable to specificity of insect and mammalian nicotinic receptors. Annu Rev Entomol 48:339–64. Watanabe E, Miyake S, Baba K, Eun H, Endo S. 2006. Immunoassay for acetamiprid detection: application to residue analysis and comparison with liquid chromatography. Anal Bioanal Chem 386:1441–8. Wong-ek K, Horprathum M, Eiamchai P, Limnonthakul P, Patthanasettakul V, Chindaudom P, Nuntawong N. 2011. Portable surface-enhanced Raman spectroscopy for insecticide detection using silver nanorod film fabricated by magnetron sputtering. Proc SPIE 7911:791108–1–11. Wu W, Walczak B, Massart DL, Prebble KA, Last IR. 1995. Spectral transformation and wavelength selection in near-infrared spectra classification. Anal Chim Acta 315: 243–55. Xie Y, Mukamurezi G, Sun Y, Wang H, Qian H, Yao W. 2012. Establishment of rapid detection method of methamidophos in vegetables by surface enhanced Raman spectroscopy. Eur Food Res Technol 234:1091–8.
References
Supporting Information Additional Supporting Information may be found in the online version of this article at the publisher’s website: Figure S1. The raw spectra of the control water:methanol solution at 50 μg/mL (dark blue line) vs the raw SERS spectra of acetamiprid at 0, 3, 10, and 50 μg/mL in apple juice (average of all data points). The second derivative SERS spectra is shown on the manuscript in Figure 3a; The second derivative spectra was used in the manuscript to help reduce noise and remove baseline, making it easier to see the differences between spectra. Figure S2. The PLS plot with the root mean square error coefficient of variation (RMSECV) of SERS spectra of different concentrations of acetamiprid in water-methanol (1:1) solvent where RMSECV = 1.14 μg/mL. Data points of 0, 1, 3, 5, 8, 10 μg/mL were used.
Vol. 79, Nr. 4, 2014 r Journal of Food Science T747
T: Toxicology & Chemical Food Safety
Alak AM, Vo-Dinh T. 1987. Surface-enhanced Raman spectroscopy of organophosphorous chemical agents. Anal Chem 59:2149–53. Brunet JL, Maresca M, Fantini J, Belzunces LP. 2008. Intestinal absorption of the acetamiprid neonicotinoid by Caco-2 cells: transepithelial transport, cellular uptake and efflux. J Environ Sci Health, Part B 43:261–70. Craig AP, Franca AS, Irudayaraj J. 2013. Surface-enhanced Raman spectroscopy applied to food safety. Annu Rev Food Sci Technol 4:369–80. Electronic Code of Federal Regulations. 2013. 40 CFR 180.578, Acetamiprid; tolerances for residues. Available from: http://www.ecfr.gov/cgi-bin/text-idx?c=ecfr&SID= 7d60c7b05ece6fdddaa11b861770fbc5&rgn=div8&view=text&node=40:25.0.1.1.27.3.19. 299&idno=40. Accessed 2013 August 30. Guerrini L, Sanchez-Cortes S, Cruz VL, Martinez S, Ristori S, Feis A. 2011. Surface-enhanced Raman spectra of dimethoate and omethoate. J Raman Spectrosc 42:980–5. Haynes CL, McFarland AD, Van Duyne RP. 2005. Surface-enhanced Raman spectroscopy. Anal Chem 77:338a–46a. He L, Lin M, Li H, Kim NJ. 2010. Surface-enhanced Raman spectroscopy coupled with dendritic silver nanosubstrate for detection of restricted antibiotics. J Raman Spectrosc 41:739– 44. He L, Lamont E, Veeregowda B, Sreevatsan S, Haynes CL, Diez-Gonzalez F, Labuza TP. 2011a. Aptamer-based surface-enhanced Raman scattering detection of ricin in liquid foods. Chem Sci 2:1579–82. He L, Rodda T, Haynes CL, Deschaines T, Strother T, Diez-Gonzalez F, Labuza TP. 2011b. Detection of a foreign protein in milk using surface-enhanced Raman spectroscopy coupled with antibody-modified silver dendrites. Anal Chem 83:1510–3. Hirotaka O, Okihashi M, Kazuhiko A, Kitagawa Y, Hori S. 2003. Determination of neonicotinoid pesticide residues in chromatography mass spectrometry. J Agric Food Chem 51: 2501–5.
