Accepted Manuscript Analytical Methods A fast and environmental friendly analytical procedure for determination of melamine in milk exploiting fluorescence quenching Carina F. Nascimento, Diogo L. Rocha, Fábio R.P. Rocha PII: DOI: Reference:

S0308-8146(14)01192-3 http://dx.doi.org/10.1016/j.foodchem.2014.07.144 FOCH 16210

To appear in:

Food Chemistry

Received Date: Revised Date: Accepted Date:

4 January 2014 2 July 2014 30 July 2014

Please cite this article as: Nascimento, C.F., Rocha, D.L., Rocha, F.R.P., A fast and environmental friendly analytical procedure for determination of melamine in milk exploiting fluorescence quenching, Food Chemistry (2014), doi: http://dx.doi.org/10.1016/j.foodchem.2014.07.144

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

A fast and environmental friendly analytical procedure for determination of melamine in milk exploiting fluorescence quenching

Carina F. Nascimento, Diogo L. Rocha, Fábio R. P. Rocha* Centro de Energia Nuclear na Agricultura, Universidade de São Paulo P.O. Box 96 – 13400-970 – Piracicaba – SP– Brazil

*Corresponding author. FAX: +55 19 3429 4610 E-mail address: [email protected]

1

Abbreviated running title: A fast and green procedure for determination of melamine in milk

Abstract A fast and environmental friendly procedure was developed for melamine determination as an adulterant of protein content in milk. Triton X-114 was used for sample clean-up and as a fluorophore, whose fluorescence was quenched by the analyte. A linear response was observed from 1.0 to 6.0 mg L-1 melamine, described by the Stern-Volmer equation Io/I = (0.999±0.002) + (0.0165±0.004) CMEL (r = 0.999). The detection limit was estimated at 0.8 mg L-1 (95% confidence level), which allows detecting as low as 320 µg melamine in 100 g of milk. Coefficients of variation (n=8) were estimated at 0.4% and 1.4% for analytical and reference signals, respectively. Recoveries to melamine spiked to milk samples from 95 to 101% and similar slopes of calibration graphs obtained with and without milk indicated the absence of matrix effects. Results for different milk samples agreed with those obtained by high performance liquid chromatography at the 95% confidence level.

Keywords: Milk; Melamine; Cloud point extraction; Triton X-114; Fluorescence; Adulteration.

2

1. Introduction Melamine (1,3,5-triazine-2,4,6 triamine) is a synthetic compound widely used in plastics and resins production to increase thermal resistance (Guan, Yu, & Chi, 2013; Wang, Dong, Li, & Luo, 2013; Zeng, Yang, Wang, Li, & Qu, 2011). Trace amounts of melamine can be found in foods due to migration from packaging walls (Zhu, Gamez, Chen, Chingina, & Zenobi, 2009). On the other hand, the high nitrogen content (ca. 66% w/w) and low cost makes melamine a common adulterant of milk and dairy products aiming to increase the apparent protein content (Zhang, Wu, Donghua, & Song, 2011). This adulteration is not easily identified in routine analysis because the non-proteic nitrogen cannot be detected by usual procedures for protein determination, such as Kjeldahl (Sun, Lixin, Ai, Liang, & Wu, 2010) and Dumas (Saint-Denis & Goupy, 2004) methods. In a long term, consumption of products contaminated by melamine leads to formation of insoluble crystals in the bloodstream, which accumulate in kidneys and bladder. This can cause tissue damage and may lead to death (Wang et al., 2013; Sun et al., 2010; Tyan, Yang, Jong, Wang, & Shiea, 2009). As a consequence, WHO (World Health Organization) and FDA (U.S. Food and Drug Administration) established daily intake limits for melamine by adults and children as 2.5 and 1.0 mg kg-1 of body weight, respectively. The threshold limit for milk was established as 2.5 mg L-1 (World Health Organization; U.S. Food and Drug Administration). In extreme episodes, melamine concentrations reached 3300 mg kg-1 in adulterated milk (Rambla-Alegre, Peris-Vicente, Marco-Peiró, Beltrán-Martinavarro, & Esteve-Romero, 2002). Therefore, fast and reliable procedures for monitoring melamine in milk and its derivatives is of utmost importance due to high intake of these foods, especially by infants.

3

The most serious cases of intoxication were reported in China in 2008 and thus a reference HPLC procedure for raw milk and dairy products was established (Sun et al., 2010; Rambla-Alegre et al., 2002; Wen et al., 2010). Other procedures exploited ionexchange chromatography (Ono et al., 1998), gas chromatography coupled to mass spectrometry (Pan et al., 2013), MALDI-TOF (Su et al., 2013), chemiluminescence (Zeng et al., 2011; Zhang et al., 2011), immunoassays (Fodey et al., 2011), and Raman spectrometry (Ma et al., 2013). These procedures generally involve expensive equipment and time-consuming sample preparation, including extraction and preconcentration steps. Fluorimetric procedures were previously proposed for melamine determination in milk (Feng et al., 2012; Attia, Bakir, Abdel-aziz, & Abdel-mottaleb, 2011; Zhang et al., 2012; Liu et al., 2012; Wang et al., 2011; Tang et al., 2013). Feng et al. exploited the inhibition of the oxidation of acridine red by potassium permanganate in the presence of melamine (Feng et al., 2012) and quenching of fluorescence of the Ru(II) carbonyl complexes was exploited for melamine determination at the nanomolar range (Attia et al., 2011). Other works have exploited the quenching of the fluorescence of quantum dots by melamine, i.e. the inner filter effect caused by Au nanoparticles on the fluorescence of CdTe (Zhang et al., 2012) or the oxidation of CuInS2 by H2O2 (Liu et al., 2012). In spite of achieving high sensitivity, sample preparation in these fluorimetric procedures was time-consuming (up to 100 min) and large amounts of toxic reagents were consumed (e.g. lead acetate and DMSO (Attia et al., 2011). In addition, the fluorescent compounds are expensive (Feng et al., 2012) or and some of them are not commercially available and their synthesis are also time-consuming (Attia et al., 2011; Zhang et al., 2012; Liu et al., 2012). Immunoassays also show these drawbacks and

4

involve time-consuming production of antibodies (Wang et al., 2011). These characteristics make the procedures unsuitable for routine analysis. Cloud point extraction (CPE) is a green alternative for liquid-liquid extraction aiming at analyte preconcentration and sample clean-up (Rocha et al., 2013; Lopes, et al., 2007; Coelho & Arruda, 2005). In milk analysis, CPE was used for separation and preconcentration of casein (Lopes et al., 2007), penicillin (Kukusamude et al., 2010), adrenaline (Qin, Gong, Miao, & Yan, 2012), manganese (Rod, Borhani, & Shemirani, 2006 ) and sulphonamides (Zhang et al., 2011). The non-ionic surfactant Triton X-114 has been employed in most of the procedures (Zhang, Duan, & Wang, 2011) in view of the cloud point between 23 and 25 °C, as well as low cost and high biodegradability (Lopes et al., 2007; Kukusamude et al., 2010; Qin et al., 2012; Rod et al., 2006; Zhang et al., 2011; Pytlakowska et al., 2013; Stalikas, 2002; Szymanowski, 2000). This surfactant absorbs at UV and shows fluorescence (Tabrizi, 2007), which can be a hindrance for some spectroanalytical procedures. The need for a simple, fast and environmental friendly procedure became evident by taking into account the approaches for melamine determination in milk. The objective of this work was then to develop a novel analytical procedure to achieve this goal, exploiting CPE for sample preparation. For the first time, quenching of Triton X114 fluorescence was exploited for melamine determination.

2. Experimental 2.1. Reagents and solutions All solutions were prepared in ultrapure water (resistivity higher than 18 MΩ cm). Melamine stock solution (Merck, Germany) was prepared by dissolving 100.0 mg 5

of the reagent in ca. 40 mL of water at 40 °C and diluting up to 100 mL. Reference working solutions (1.0 to 6.0 mg L-1) were daily prepared from dilutions of the stock in water. Stock solutions of 5.0% (m/v) Triton X-114 (Sigma, Germany) and 0.3 mol L-1 trichloroacetic acid (Merck, Germany) were prepared by dissolving the reagents in water.

2.2. Apparatus All excitation and emission spectra were obtained on a spectrofluorometer (Eclipse, Varian, Mulgrave, VIC, Australia) using a quartz cuvette with 1 cm optical path. The emission and excitation slits were adjusted to yield a 5.0-nm resolution (except in the evaluation of the effect of trichloroacetic acid concentration, in which the slit was set to 2.5 nm). A mechanic orbital shaker (Tecnal, Brazil), a temperaturecontrolled water bath (Marconi, M184, Brazil) and a centrifuge (Quimis, Brazil) were used for solutions homogenization and to promote phase separation.

2.3. Proposed procedure Liquid (whole, skimmed, semi-skimmed and pasteurized) and whole powdered, milk samples were obtained at the local market and preferably analyzed in the same day – pasteurized milk was stored under refrigeration until analysis. The powdered milk was prepared according to package instructions. For protein precipitation, 1.0 mL of 0.3 mol L-1 trichloroacetic acid solution was added to 0.5 mL of milk. The mixture was stirred and centrifuged for 5 min. Afterwards, 12 mL of water and 2.0 mL of 5.0% (m/v) Triton X-114 were added to 1.0 mL of the supernatant for the cloud point extraction. The mixture was manually stirred, heated in a 50 ºC water bath for 10 min to induce the cloud point and 6

centrifuged

for

5 min at 5000 rpm for phase separation. Sample preparation and fluorimetric measurements in the supernatant were performed in triplicate. All fluorescence measurements were carried out at (25±1) oC with excitation and emission wavelengths set at 235 and 302 nm, respectively. The effect of pH on the fluorescence quenching was evaluated from a 12 mmol L-1 Triton X-114 aqueous solution in the presence or absence of 10 mg L-1 melamine. The responses obtained for a whole milk sample and for solutions prepared in water were used for evaluation of the effects of TCA and Triton X-114 concentrations on sample clean-up. Calibration graphs were obtained from melamine solutions (n=5) prepared in water (proposed procedure) or whole milk (evaluation of matrix effects). Coefficients of variation were estimated from fluorescence measurements taken for a whole milk sample in the absence and presence of 10 mg L-1 melamine (eighth portions of each, treated according to the procedure described above). Melamine recoveries were estimated by spiking 2.0 or 4.0 mg L-1 of the analyte in four different types of milk (whole, skimmed and whole powder). A different set with five samples was used for accuracy assessment after spiking with 2.50 mg L-1 melamine. The detection limit was estimated as the lowest melamine concentration that yielded a fluorescence response significantly different from the reference signal at the 95% confidence level.

2.4. Reference procedure The reference procedure for accuracy assessment was based on micellar chromatography, validated by the authors according to EU Regulation 2002/654/EC (Rambla-Alegre et al., 2002), except for a short C18 chromatographic column (0.46 x

7

10 cm, 3.5 µm). Calibration was performed by matrix matching. Reference solutions (from 1 to 9 mg L-1 melamine) were prepared from 1.0 mL of a whole milk sample spiked with the suitable amounts of melamine. The final volume was adjusted to 10.0 mL with 0.05 mol L-1 SDS. The matrix matching was necessary because low recoveries (as low as 46%) were observed for some milk samples by the HPLC procedure, thus indicating matrix effects.

3. Results and discussion 3.1. General aspects In the present work, Triton X-114 acted as both CPE surfactant and fluorophore. Figure 1 shows emission spectra of Triton X-114 solutions in the absence and presence of different melamine concentrations (from 1.0 to 6.0 mg L-1). These results indicate that melamine quantification can be based on quenching of the surfactant fluorescence, which was proportional to the analyte concentration, according to the Stern-Volmer equation (see the inset graph on Figure 1), Lakowicz, 2006.

Please, insert Figure 1

The effect of the temperature on the fluorescence quenching was evaluated within 20 and 35 ºC, and the slopes of the Stern-Volmer graphs increased progressively up to 3.1 %. In addition, the fluorescence lifetime of Triton X-114 was reduced from 10.8 to 5.5 ns in the presence of melamine. These experimental observations indicate that the quenching process is dynamic (collisional) (Lakowicz, 2006), thus involving energy transference from the excited fluorophore (Triton X-114) to the quencher (melamine). 8

Sample clean-up was required prior to fluorescence measurements and removal of proteins was carried out by denaturation with trichloroacetic acid (TCA). The resultant whey suspension can contain interfering hydrophobic species (e.g. fats) and thus an additional clean-up step was exploited using CPE instead of commonly employed organic solvents, such as acetonitrile and methanol (Zeng et al., 2011; Wang et al., 2011). The hydrophobic species, such as clusters of fats, were extracted to the surfactant-rich phase. It was evaluated from measurements in both phases that melamine remained quantitatively at the supernatant, in which the Triton X-114 amount is close to its critical micelle concentration.

3.2. Procedure optimization Due to TCA addition for sample clean-up, the influence of pH on Triton X-114 fluorescence was evaluated in the presence and absence of melamine (Table 1). As previously observed (Tabrizi, 2007), the fluorescence intensity of Triton X-114 increased with pH, but this was also observed in the presence of melamine. The Io/I ratio did not show significant differences from 2.0 to 4.0 (ca. 3%) and the pH close to 3.0 (provided by addition of TCA in the previous step) was suitable for melamine quantification. Please, insert Table 1 The effects of TCA (from 0.10 to 0.30 mol L-1) and Triton X-114 (from 6 to 18 mmol L-1) concentrations in the sample clean-up were evaluated aiming at the closest fluorescence response for milk and reference solutions (Figure 2). At lower TCA concentrations (0.10 mol L-1) turbid suspensions were obtained from milk, which indicate incomplete protein precipitation. The particles in suspension caused light scattering and affected the fluorescence measurements. For higher concentrations, 9

similar fluorescence intensities (Io) were obtained in the presence and absence of sample. TCA at 0.30 mol L-1 yielded the closest I values and then was selected for protein precipitation. Please, insert Figure 2 Triton X-114 should lead a suitable reference signal (Io) and quantitatively remove fats to avoid their interference of fluorescence measurements. Therefore, the effect of the surfactant concentration was evaluated using a milk sample and a reference solution; Figure 2B shows the results expressed as the final Triton X-114 concentrations. As expected, the Triton X-114 concentration in the surfactant-poor phase remained constant (only the amount of the surfactant-rich phase increased). Moreover, the I values were not affected, which indicates that the removal of melamine during sample clean-up was not significant. This was expected due to the pH 3.0 of the extract before CPE, in which > 99% melamine is positively charged (pKa = 5) (Feng et al., 2012). The closest Io/I ratios were reached with 12 mmol L-1 Triton X-114 (1.69±0.05 and 1.68±0.10 with and without milk, respectively) and this concentration was chosen for further studies.

3.3. Analytical features and application A linear response was observed from 1.0 to 6.0 mg L-1 (1.0 to 6.2 mg kg-1 ) melamine in milk, as described by eq. 1, in which I° and I are the fluorescence intensities in the absence and presence of melamine, respectively, and CMEL is melamine concentration (mg L-1). The slope value (0.0165 L mg-1) is the product of the bimolecular quenching constant and the fluorescence lifetime in the absence of the quencher. The intercept close to 1.00 and the r value (0.999) confirm the agreement to the Stern-Volmer model and thus the viability to exploit the fluorescence quenching for 10

analytical purposes. Differently of the inset of Fig. 1, equation 1 took into account the sample dilution in the clean-up step.

‫ܫ‬° = 0.999 + 0.0165C‫( ܮܧܯ‬1) ‫ܫ‬

The detection limit was estimated as 0.8 mg L-1 (95% confidence level), which allows the detection of adulteration by as low as 320 µg melamine in 100 g of milk. The coefficients of variation were estimated as 0.4% and 1.4% (n = 8) for a whole milk sample in the absence and presence of melamine, respectively. The procedure showed suitable sensitivity for the determination of melamine in milk, even after the 45-fold dilution during sample preparation. It consumes only 49 mg of trichloroacetic acid and 100 mg Triton X-114 per determination, which is equivalent to only US$ 0.10. The slopes of the calibration graphs (r = 0.999) obtained in the absence [Io/I = (0.999±0.002) + (0.0165±0.004) CMEL] and the presence [Io/I = (1.000±0.002) + (0.0161±0.004) CMEL] of the sample agreed at the 95% confidence level. In addition, recoveries for melamine spiked to the samples (Table 2) were within 95 and 101%. These results indicate the absence of matrix effects and confirm the efficiency of the sample clean-up prior to fluorimetric measurements. From a paired t-test, it can be concluded that results obtained for spiked milk samples (Table 3) agreed with those achieved by the HPLC reference procedure (Rambla-Alegre et al., 2002) at the 95% confidence level. Please, insert Tables 2 and 3 Table 4 shows a comparison of the analytical features achieved in different approaches for determination of melamine in milk. It highlights that the proposed 11

procedure is more environmental friendly and requires one of the simplest sample pretreatments, thus improving precision and requiring one of the lowest analysis time. In comparison to the fluorimetric procedures previously described (Feng et al., 2012; Attia et al., 2011; Zhang et al., 2012; Liu et al., 2012; Wang et al., 2011; Tang et al., 2013), the proposed procedure is simpler (avoid chemical derivatization and complex sample clean-up), inexpensive, faster and requires only commonly used reagents. In spite of exploiting fast detection, some procedures require time-consuming sample pretreatment. For example, the FI-CL procedure (Zheng et al., 2011) takes only 25 s for measurement, but sample treatment involves TCA and acetonitrile, sonication, centrifugation, clean-up by SPE (sample loading, washing and elution) followed by solvent evaporation to dryness and dissolution in water. Such time-consuming sample treatment was also exploited in other works (Sun et al., 2010; Wen et al., 2010; Feng et al., 2012; Su et al., 2013). On the other hand, most of the sample pretreatments require toxic reagents and solvents, such as methanol (Wen et al., 2010; Su et al., 2013; Feng et al., 2012; Wang et al., 2011), chloroform (Zhang et al., 2012; Su et al., 2013; Tang et al., 2013), methylene chloride (Feng et al., 2012) and lead acetate (Sun et al., 2010; Attia et al., 2011). In addition, several procedures require the synthesis of reagents and materials which are not commercially available, such as fluorescein-labeled tracers (Wang et al., 2011), gold nanoparticles (Guan et al., 2013; Zhang et al., 2012), modified silver nanoparticles (Ma et al., 2013), modified sorbents (Su et al., 2013), Ru(II) carbonyl complex (Attia et al., 2011), and quantum dots (Liu et al., 2012; Tang et al., 2013; Zhang et al., 2012). Previously reported procedures achieved lower detection limits (as low as 60 ng L-1) and wide linear response ranges. However, the analytical features achieved in the proposed procedure allow direct determination of melamine in milk by taking into

12

account the threshold limit established by WHO (World Health Organization) and FDA (U.S. Food and Drug Administration). Some previous proposals also showed poor precision (Su et al., 2013) and recoveries from the melamine spiked in milk samples (Wang et al., 2011). Please, insert Table 4

4. Conclusions A simple, reliable, fast and environmental friendly procedure was developed for melamine determination in adulterated milk by exploiting, for the first time, the fluorescence quenching of Triton X-114. A simple clean-up step with trichloroacetic acid and cloud point extraction was successfully exploited to avoid matrix effects. The use of organic solvents and chromatographic separation were then avoided. These analytical features make the proposed procedure attractive for detection of adulteration in milk, thus contributing to avoid the catastrophic consequences of this illegal action as recently reported in several countries.

13

Acknowledgements The authors acknowledge the fellowships and financial support from the Brazilian agencies Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP). Prof. V.L. Tornisielo and R.F. Pimpinato are thanked for their assistance with the reference procedure and A. D. Batista for the critical comments. This is a contribution of the National Institute of Advanced Analytical Science and Technology (INCTAA).

References Attia, M. S., Bakir, E., Abdel-aziz, A. A., & Abdel-mottaleb, M. S. A. (2011) Determination of melamine in different milk batches using a novel chemosensor based on the luminescence quenching of Ru (II) carbonyl complex. Talanta, 84, 27–33. Coelho, L. M., & Arruda, M. A. Z. (2005). Preconcentration procedure using cloud point extraction in the presence of electrolyte for cadmium determination by flame atomic absorption spectrometry. Spectrochimica Acta Part B, 60, 743–748. Feng, W., Lv, C., Liping, Y., Cheng, J., & Yan, C. (2012). Determination of melamine concentrations in dairy samples. Food Science and Technology, 47, 147–153. Fodey, T. L., Thompson, C. S., Traynor, I. M., Haughey, S. A., Kennedy, D. G., & Crooks, S. R. H. (2011). Development of an Optical Biosensor Based Immunoassay to Screen Infant Formula Milk Samples for Adulteration with Melamine. Analytical Chemistry, 83, 5012–5016.

14

Guan, H., Yu, J., & Chi, D. (2013). Label-free colorimetric sensing of melamine based on chitosan-stabilized gold nanoparticles probes. Food Control, 32, 35–41. Kukusamude, C., Santalad, A., Boonchiangma, S., Burakham, R., Srijaranai, S., & Chailapakul, O. (2010). Mixed micelle-cloud point extraction for the analysis of penicillin residues in bovine milk by high performance liquid chromatography. Talanta, 81, 486–492. Lakowicz, J. R. (2006). Principles of fluorescence spectroscopy, (3th. ed.) Baltimore: Springer, p. 954. Liu, S., Hu, J., Zhang, H., & Sun, X. (2012). CuInS2 quantum dots-based fluorescence turn off/on probe for detection of melamine. Talanta, 101, 368–373. Lopes, A. S., Garcia, J. S., Catharino, R. R., Santos, L. S., Eberlin, M. N., & Arruda, M. A. Z. (2007). Cloud point extraction applied to casein proteins of cow milk and their identification by mass spectrometry. Analytica Chimica Acta, 590, 166–172. Ma, P., Liang, F., Sun, Y., Jin, Y., Chen, Y., Wang, X., Zhang, H., Gao, D., & Song, D. (2013). Rapid determination of melamine in milk and milk powder by surfaceenhanced

Raman

spectroscopy and

using

cyclodextrin-decorated

silver

nanoparticles. Microchimica Acta, 180, 1173–1180. Ono, S., Funato, T., Inoue, Y., Munechika, T., Yoshimura, T., Morita, H., Rengakuji, S., & Shimasaki, C. (1998). Determination of melamine derivatives, melame, meleme, ammeline and ammelide by high-performance cation-exchange chromatography. Journal of Chromatography A, 815, 197–204. Pan, X., Wu, P., Yang, D., Wang, L., Shen, X., & Zhu, C. (2013). Simultaneous determination of melamine and cyanuric acid in dairy products by mixed-mode solid phase extraction and GC–MS. Food Control, 30, 545–548

15

Pytlakowska, K., Kozik, V., & Dabioch, M. (2013). Complex-forming organic ligands in cloud-point extraction of metal ions: A review. Talanta, 110, 202–228. Qin, H., Gong, Y., Miao, C., & Yan, Z. (2012). Ultrasonic-thermostatic-assisted cloud point extraction coupled to high-performance liquid chromatography for the analysis of adrenalines residues in milk. European Food Research and Technology, 234, 543–550. Rambla-Alegre, M., Peris-Vicente, J., Marco-Peiró, S., Beltrán-Martinavarro, B., & Esteve-Romero, J. (2002). Development of an analytical methodology to quantify melamine in milk using micellar liquid chromatography and validation according to EU Regulation 2002/654/EC. Talanta, 81, 894–900. Rocha, D. L., Batista, A. D., Rocha, F. R. P., Donati, G. L., & Nóbrega, J. A. (2013). Greening sample preparation in inorganic analysis. Trends in Analytical Chemistry, 45, 79–92. Rod, A. R., Borhani, S., & Shemirani, F. (2006). Cloud point preconcentration and flame atomic absorption spectrometry: application to the determination of manganese in milk and water samples European Food Research and Technology, 223, 649–653. Saint-Denis, T., & Goupy, J. (2004). Optimization of a nitrogen analyser based on the Dumas method. Analytica Chimica Acta, 515, 191–198. Stalikas, C. D. (2002). Micelle-mediated extraction as a tool for separation and preconcentration in metal analysis Trends in Analytical Chemistry, 21, 343–355. Su, X., Zhou, H., Chen, F., Gao, B., Liu, Z., Zhang, Y., Liu, F., Liu, F., Li, Z., & Gao, Z. (2013). Modified SBA-15 matrices for high-throughput screening of melamine in milk samples by MALDI-TOF MS. International Journal of Mass Spectrometry, 338, 39–44.

16

Sun, F., Ma, W., Xu, L., Zhu, Y., Liu, L., Peng, C., Wang, L., Kuang, H., & Xu, C. (2010). Analytical methods and recent developments in the detection of melamine. Trends in Analytical Chemistry, 29, 1239–1249. Sun, H., Lixin, W., Ai, L., Liang, S., & Wu, H. (2010). A sensitive and validated method for determination of melamine residue in liquid milk by reversed phase high-performance liquid chromatography with solid-phase extraction. Food Control, 21, 686–691. Szymanowski, J. (2000). Surfactant enhanced non-classical extraction. Journal of Radioanalytical and Nuclear Chemistry, 246, 635–642. Tabrizi, A. B. (2007). Cloud point extraction of salicylic acid from human urine and its spectrofluorimetric determination. Chemia Analityczna, 52, 823–833. Tang, G., Du, L., & Su, X. (2013). Detection of melamine based on the fluorescence resonance energy transfer between CdTe QDs and Rhodamine B. Food Chemistry, 141, 4060–4065. Tyan, Y., Yang, M., Jong, S. B., Wang, C., & Shiea, J. (2009). Melamine contamination. Analytical and Bioanalytical Chemistry, 395, 729–733. U.S. Food and Drug Administration. URL http://www.fda.gov. Acessed 19.05.13. Wang, P., Dong, J. X., Li, N. B., & Luo, H. Q. (2013). Highly sensitive detection of melamine based on gemini surfactant using enhanced resonance Rayleigh scattering signals. Spectrochimica Acta Part A, 103, 38–44. Wang, Q., Haughey, S. A., Sun. Y., Eremin, S. A., Li, Z., Liu, H., Xu, Z., Shen, Y., & Lei, H. (2011). Development of a fluorescence polarization immunoassay or the detection of melamine in milk and milk powder. Analytical and Bioanalytical Chemistry, 399, 2275–2284.

17

Wen, Y., Liu, H., Han, P., Gao, Y., Luan, F., & Li, X. (2010). Determination of melamine in milk powder, milk and fish feed by capillary electrophoresis: a good alternative to HPLC. Journal of the Science of Food and Agriculture, 90, 2178– 2182. World Health Organization.URL http://www.who.int. Acessed 16.05.13. Zeng, H., Yang, R., Wang, Q., Li, J., & Qu, L. (2011). Determination of melamine by flow injection analysis based on chemiluminescence system. Food Chemistry, 127, 842–846. Zhang, J., Wu, M., Donghua, C., & Song, Z. (2011). Ultrasensitive determination of melamine in milk products and biological fluids by luminol-hydrogen peroxide chemiluminescence. Journal of Food Composition and Analysis, 24, 1038–1042. Zhang, M., Cao, X., Li, H., Guan, F., Guo, J., Shen, F., Luo, Y., Sun, C., & Zhang, L. (2012). Sensitive fluorescent detection of melamine in raw milk based on the inner filter effect of Au nanoparticles on the fluorescence of CdTe quantum dots. Food Chemistry, 135, 1894–1900. Zhang, W., Duan, C., & Wang, M. (2011). Analysis of seven sulphonamides in milk by cloud point extraction and high performance liquid chromatography. Food Chem, 126, 779–785. Zhu, L., Gamez, G., Chen, H., Chingina, K., & Zenobi, R. (2009). Rapid detection of melamine in untreated milk and wheat gluten by ultrasound-assisted extractive electrospray

ionization

mass

spectrometry

(EESI-MS).

Chemical

Communications, 5, 559–561.

18

Table 1 Influence of pH on Triton X-114 fluorescence in the absence (Iº) and presence (I) of melamine. Values correspond to triplicate measurements.

pH

Iº

I

Iº/I

2.0

308 ± 3

145 ± 4

2.13 ± 0.06

4.0

331 ± 9

160 ± 3

2.06 ± 0.07

6.0

366 ± 14

215 ± 1

1.70 ± 0.07

8.0

453 ± 30

261 ± 4

1.74 ± 0.12

10.0

570 ± 53

283 ± 2

2.01 ± 0.19

19

Table 2 Recoveries of melamine in different milk samples. Results correspond to measurements in triplicate.

Sample

Melamine (mg L-1)

Recovery (%)a

Added

Found

2.00

1.9±0.1

96±4

4.00

3.8±0.2

95±5

2.00

2.0±0.1

100±5

4.00

4.0±0.2

101±5

2.00

2.0±0.1

101±5

4.00

4.0±0.1

100±2

Pasteurized whole

2.00

2.0±0.2

100±10

milk

4.00

4.0±0.1

100±2

Whole milk powder

Integral milk UHT

Skimmed milk UHT

a. Values estimated before rounding-off to the correct number of significant figures.

20

Table 3 Mean values and standard deviations (n=3) for the determination of melamine in milk samples by the proposed procedure and HPLC. Melamine (mg kg-1)

Sample*

Fluorescence

HPLC

Whole powder

2.58 ± 0.14

2.52 ± 0.01

Whole UHT

2.68 ± 0.14

2.68 ± 0.01

Semi-skimmed UHT-1

2.81 ± 0.11

2.66 ± 0.01

Semi-skimmed UHT-2

2.79 ± 0.14

2.71 ± 0.01

2.51 ± 0.21

2.94 ± 0.07

Pasteurized skimmed -1

-1

*spiked with 2.50 mg L (2.58 mg kg )

21

Table 4. Analytical features of procedures for melamine determination in milk. Detection

Sample treatment

Linear range (mg L-1) 1.0–6.0 0.0002–1.6

LD (µg L-1) 800 0.06

Recovery (%) 95–101 98–111

Analysis time (min)a 20 100

RSD (%)

Reference

1.4 (n=8) 2.3 (n=6)

This work Feng et al., 2012

0.0003–0.1

42

99–102

70

1.9 (n=3)

methanol:water TCA and chloroform TCA and acetonitrile

0.02–0.9 0.01–0.1 0.001–1

9.3 20 0.6

79–119 103–104 95–99

35 40 90

19 (n=3) 1.4 (n=11) 4.4 (n=3)

TCA, acetonitrile and chloroform

0.0063– 0.504 0.1–100

1.2

99–104

20

3.5 (n=3)

6

85–107

25

4.2 (n=8)

0.2–80

120

86–102

90

3.3 (n=11)

Attia, Bakir, Abdel-aziz & Abdel-mottaleb, 2011 Wang et al., 2011 Zhang et al., 2012 Liu, Hu, Zhang & Sun, 2012 Tang, Du, Su, 2013 Guan, Yu &Chi, 2013 Zeng et al., 2011

0.1–50 0.2–100

18 5

85–99 85–109

90 25b

3.1 (n=11) < 7.6 (n=6)

0.8–80

80

96–110

90

1.22 (n=6)

Sun et al., 2010 Rambla-Alegre et al., 2002 Wen et al., 2010

0.1–10

100



90

≤ 14 (n=11)

Su et al., 2013

0.005–0.1

3

89–104

15

4.6 (n=5)

Ma et al., 2013

TCA and CPE with Triton X-114 TCA and methylene chloride; SPE, solvent evaporation and redissolution TCA and lead acetate Fluorescence

UV-vis spectrophotometry Flow injectionChemiluminescence HPLC capillary electrophoresis MALDI-TOF MS Raman spectroscopy

TCA and chitosan-stabilized gold nanoparticles TCA and acetonitrile, SPE; solvent evaporation and redissolution TCA, lead acetate and SPE SDS and phosphate buffer pH 3 TCA and acetonitrile, SPE, solvent evaporation and redissolution TCA, acetonitrile and chloroform, SPE, solvent evaporation and redissolution acetonitrile, filtration and dilutions

a. Values estimated from the described procedures; b. plus 2 h for column washing; MALDI-TOF MS: Matrix-assisted laser-desorption ionization-time of flight mass spectrometry; SPE: solid-phase extraction; SDS: Sodium dodecyl sulphate; TCA: trichloroacetic acid.

22

Figure captions

Fig. 1. Emission spectra of Triton X-114 with excitation at 235 nm in the absence of melamine (a) or addition of 1.0 (b), 2.0, (c) 3.0, (d) 4.0, (e) 5.0 (f) and 6.0 mg L-1 (g) of melamine. The inset shows the corresponding Stern-Volmer graph. Values refer to fluorimetric measurements before optimization of the clean-up step.

Fig. 2. Effect of the concentration of trichloroacetic acid (A) and Triton X-114 (B) in fluorescence intensity of Triton X-114 in the absence (a) and presence (c) of melamine in the absence of milk and in the absence (b) and presence (d) of melamine in a medium containing milk. Error bars refer to the estimative of standard deviations (n=3).

23

500 1.6

a

1.5

400

g Iº/I

Fluorescence Intensity

1.4

300

1.3 1.2 1.1 1.0 1

200

2

3

4

5

6

-1

Melamine(mg L )

100

0 300

350

400

450

Wavelength (nm)

Figure 1

24

70

50

a

b

40 30

B

a

Fluorescence intensity

Fluorescence intensity

60

600

A

c d

20

500

b

400

300

c d

10 200 0.10

0.15

0.20

0.25 -1

Tricloroacetic acid (mol L )

0.30

6

9

12

15

18

-1

Triton X-114 (mmol L )

Figure 2 25

Highlights

- A simple, fast and inexpensive procedure for determination of melamine in milk;

- Melamine determination without complex sample treatment or organic solvents;

- Triton X-114 for sample clean-up by cloud point extraction and as a fluorophore;

- Quenching of Triton-X114 fluorescence pioneering exploited for analytical purposes.

26