Residue analysis of melamine in milk products by micellar electrokinetic capillary chromatography with amperometric detection

Food Chemistry 121 (2010) 215–219

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Residue analysis of melamine in milk products by micellar electrokinetic capillary chromatography with amperometric detection Jinyan Wang, Lianmei Jiang, Qingcui Chu *, Jiannong Ye * Department of Chemistry, East China Normal University, North Zhongshan Road, 3663, Shanghai 200062, People’s Republic of China

a r t i c l e

i n f o

Article history: Received 10 December 2008 Received in revised form 4 May 2009 Accepted 28 November 2009

Keywords: Micellar electrokinetic capillary chromatography Amperometric detection Melamine Milk Milk powder

a b s t r a c t A high-performance micellar electrokinetic capillary chromatography with amperometric detection (MECC–AD) method for the fast determination of melamine (MM), occasionally used to increase the apparent protein content of milk products, has been developed. Method development involved optimisation of the working electrode, the running buffer system and acidity, the concentration of sodium dodecyl sulphate (SDS), the separation voltage, the applied potential and the sample injection time. Under the optimum conditions, MM can be well separated from its co-existing interferences in real-world samples within 9 min at the separation voltage of 20 kV in a 8 mmol/L SDS/20 mmol/L H3BO3–Na2B4O7 running buffer (pH 7.4). Satisfactory recovery (83.3–105.5%), repeatability of the peak current (3.2%) and migration time (3.8%), and the limit of detection (2.1  10 6 g/mL) for the method were achieved. This proposed procedure has been successfully used for the determination of MM in milk products. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction Melamine (2,4,6-triamino-s-triazine, MM) is an important chemical intermediate used in the manufacture of plastics, in the production of melamine–formaldehyde resins for surface coatings, laminates, and adhesives, and in the production of flame retardants; it is also used as a fertilizer. MM is a metabolite of cyromazine, an approved insecticide used on a broad range of vegetable crops (Wang, 2008a). MM can be found as a contaminant in a variety of food from food-contact materials. There are no established regulatory limits for MM in any type of food. None of MM should be present in food because it is toxic at high dose exposure. MM might cause urolithiasis, subsequent tissue injury and even bladder cancer (European Food Safety Authority (EFSA), 2007, Pichon, Chen, Guenu, & Hennion, 1995). A study performed by the US Food and Drug Administration (FDA) has described the risk to human health associated with eating products from animals that have been fed with MM and its analogues (Food and Drug Administration (FDA), 2007). MM has also shown to be carcinogenic for male rats (National Toxicology Program, 1983), and the combination of MM and cyanuric acid is responsible for acute renal failure in cats (Birgit, Robert, Linda, Michael, & Patricia, 2007). Unfortunately, certain cereal-based ingredients (e.g. wheat flour, wheat gluten, corn flour, and rice protein concentrate) and * Corresponding authors. Fax: +86 21 6223 3528. E-mail addresses: [email protected], [email protected] (Q. Chu), [email protected] (J. Ye). 0308-8146/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.foodchem.2009.11.074

raw milk have been adulterated with MM to increase their apparent total nitrogen concentration and consequently the calculated protein content. On the one hand, the Kjeldahl method’s universality, precision and reproducibility have made it the internationallyrecognised method for estimating the protein content in food, and it is the standard method against which all other methods are judged. It does not, however, give a measure of true protein content, as it measures non-protein nitrogen in addition to the nitrogen in proteins. On the other hand, MM is not only a nitrogenrich chemical up to 66.7%, but also an inexpensive chemical that dramatically increases the apparent protein concentration. This can be mirrored from 2007’s pet food incident and the following year’s milk powder scandal when MM was added to raw material to fake high protein contents. The concentration found in some adulterated products reached 8% in wheat gluten and rice protein concentrates, 0.2% in animal feeds (European Food Safety Authority (EFSA), 2007) and 0.09–2563 mg/kg in infant formula milk powder (State Department, People’s Republic of China, 2008). Therefore, it is urgently necessary to develop some simple, economical and efficient methods for the analysis and quantitative measurement of MM in order to insure human health. A traditional method for MM is the subtractive method by directly measuring other impurities contents such as water, ash, alkali-soluble, and the veracity is comparatively poor. Although the sublimation and picric acid weight methods are widely used for the determination of MM because of the high veracity, the drawbacks such as complicated operation and long analysis time have made it unable to meet the requirement of food safety analyses

216

J. Wang et al. / Food Chemistry 121 (2010) 215–219

(Wu & Chen, 2005; Yuan, Ma, & Du, 2004). In recent years, high performance liquid chromatography (HPLC) and gas chromatography (GC) with UV or MS detection are also employed for the residue analysis of MM in animal feeds (Cai, OuYang, Qian, & Peng, 2008; Jiang, Fan, Lin, & Li, 2008; Li et al., 2008; Muniz-Valencia et al., 2008), pet food (Kim et al., 2008; Wang, 2008b; Wang et al., 2008) and human food (Chou, Hwang, & Lee, 2003; Ding et al., 2008; Ehling, Tefera, & Ho, 2007; Filigenzi, Tor, Poppenga, & Aston, 2007; Gong et al., 2008; Wendy et al., 2008), however, the above-mentioned instruments are very expensive, and the analysis of MM needs derivatization before GC operation (General Administration of Quality Supervision, Quarantine of People’s Republic of China. Determination of melamine in raw milk, & GB/ T 22388-, 2008; Food and Drug Administration (FDA)). Capillary electrophoresis (CE) is increasingly recognised as an important analytical separation technique because of its speed, efficiency, reproducibility, ultra-small sample volume, little consumption of solvent and ease of clearing up the contaminants. In addition, with amperometric detection (AD), CE–AD affords high sensitivity and good selectivity for electroactive species (Chu, Jiang, Tian, & Ye, 2008; Chu, Tian, Lin, & Ye, 2007; Wu, Guan, & Ye, 2007). In comparison with HPLC, CE is often a more efficient separation method without requiring complicated operation and high cost. However, to our knowledge, so far only a few publications have been reported about the analysis of MM resins by CE–MS method (Cook, Klampfl, & Buchberger, 2005; Nielen, Henk, & Van de Ven, 1996; Thuy et al., 2008). In this study, a sensitive, dependable and simple method for the fast determination of MM in milk products by employing micellar electrokinetic capillary chromatography with amperometric detection (MECC–AD) has been developed. Prior to CE, an easy samplepreparation procedure consisted in a leaching process using the running buffer of CE. The proposed method has been successively used for the residue analysis of MM in milk and infant formula milk powder.

2. Materials and methods 2.1. Apparatus The laboratory-built CE–AD system (Chu et al., 2007) was used in this work. A ±30 kV high-voltage power supply (Shanghai Institute of Nuclear Research, China) provided a voltage between the ends of the capillary. The inlet end of the capillary was held at a positive potential and the outlet end was maintained at ground. A 75 cm length of 25 lm i.d. fused-silica capillary (Polymicro Technologies, Phoenix, AZ, USA) was used for the separation. All samples were injected electrokinetically without preconcentration, applying 20 kV for 6 s. The design of CE–AD detection was based on the end-column approach in which the working electrode is simply placed at the outlet of the separation capillary and detection is carried out in the same solution reservoir that contains the grounding electrode for CE instrument. A carbon-disc electrode with 300 lm diameter was employed as the working electrode. Before use, the surface of the carbon-disc electrode was polished with emery sand paper, sonicated in deionized water, and then positioned carefully opposite the capillary outlet with the aid of an Oriel Corp. (Stratford, CT, USA) Model 14901 micropositioner. A three-electrode cell system consisting of a carbon-disc working electrode, a platinum auxiliary electrode and a SCE (saturated calomel electrode) reference electrode were used in combination with a BAS LC-4C amperometric detector (Biochemical System, West Lafayette, IN, USA). The electropherograms were recorded using a chart recorder (Shanghai Dahua Instrument factory, China).

2.2. Reagents and solutions MM was purchased from Sigma (St. Louis, MO, USA). The stock solution of MM (1.0  10 3 g/mL) was daily prepared in 8 mmol/L SDS/20 mmol/L H3BO3–Na2B4O7 buffer (pH 7.4) unless otherwise mentioned, and stored in a 4 °C refrigerator. A fresh working standard solution was prepared by diluting the stock solution with the above mentioned running buffer to the desired concentration and used for different studies. Before use, all solutions were filtered through 0.22 lm nylon filters. Other used chemicals were of analytical reagent grade. All experiments were performed at room temperature. 2.3. Sample preparation Milk and infant formula milk powder samples were purchased from supermarket (Shanghai, China). 1 mL milk or 0.3 g infant formula milk powder samples were accurately weighed and extracted with 5 mL running buffer for 30 min in an ultrasonic bath, and the buffer was 8 mmol/L SDS/20 mmol/L H3BO3–Na2B4O7 solution (pH 7.4) unless mentioned. Then each of the samples was filtered through filter paper first, followed by a 0.22 lm syringe filter. After filtration, the solutions were injected directly to the CE–AD system for analysis. Before use, all sample solutions were stored in a 4 °C refrigerator. 3. Results and discussion 3.1. Optimum of analytical procedure 3.1.1. Working electrode and buffer selection In this study, different working electrodes including copper-disc electrode and carbon-disc electrode were investigated in different buffer systems including Na2B4O7–KH2PO4, NaOH, Na2B4O7–H3BO3 and Na2B4O7–NaOH to obtain best electrochemical response, respectively. Experimental results showed that MM generated no electrochemical response at copper electrode or carbon electrode in NaOH or Na2B4O7–NaOH running buffers, respectively; MM has low current response at carbon electrode in Na2B4O7–KH2PO4 solution; and MM has higher electrochemical response at carbon electrode in Na2B4O7–H3BO3 solution. Therefore, the Na2B4O7– H3BO3 solution and carbon electrode were selected as the optimum running buffer and working electrode, respectively. 3.1.2. Effect of the potential applied to the working electrode In amperometric detection, the potential applied to the working electrode directly affects the sensitivity, detection limit and stability of this method. Therefore, hydrodynamic voltammetry experiment was investigated to obtain optimum detection results. As shown in Fig. 1, when the applied potential exceeded +750 mV (vs. SCE), MM could generate oxidation current at the working electrode, and the oxidation current increased rapidly with applied potential. When the applied potential was greater than +1200 mV (vs. SCE), both the baseline noise and the background current increased very strongly, resulting in an unstable baseline, which was a disadvantage for sensitive and stable detection. Therefore, the applied potential to the working electrode was maintained at +1150 mV (vs. SCE) where the background current was not too high and the S/N ratio was the highest. Moreover, the working electrode exhibited good stability and high reproducibility at this optimum potential. 3.1.3. Effects of the pH value and the SDS concentration In capillary zone electrophoresis (CZE), sample separation relies on the differential electrophoretic mobility of the analytes, i.e. re-

217

J. Wang et al. / Food Chemistry 121 (2010) 215–219

6

MM 5

I/nA

4

3

2

1

0 700

800

900

1000

1100

1200

1300

E/mV

lies on the degree of dissociation and the molecular size. In MECC, however, both the difference in effective electrophoretic mobility, and the partition of analytes between the running buffer and the ‘‘pseudo-stationary phase” – micelles play an important role. Whether electrophoresis or partition mechanism dominates the separation process mainly depends on the acidity of the running buffer, hence, the pH value of the running buffer strongly influences the separation. The pH dependence of the migration time was investigated in the pH range of 7.4–9.2. It was found that MM always overlapped with the unknown compound peak in real samples from pH 7.4 to 9.2, and the typical electrophograms of standard solution and the real sample solution in CZE were shown in Fig. 2A and B, respectively. In considering the analysis time and stability, pH 7.4 was selected as the optimum pH value for the running buffer. Besides the pH value, SDS concentration is also an important parameter. It is well known that the SDS concentration is related to specific pseudo-retention factor of the analytes. As shown in Fig. 3, at a fixed pH value 7.4, the specific pseudo-retention factor of the analyte increased with increasing SDS concentration, resulting in longer migration time. When SDS concentration was greater than 8 mmol/L, baseline separation of MM can be obtained, and the interference of unknown compound peak in real samples can be avoided. Therefore, 8 mmol/L SDS was finally chosen for optimum separation and shorter analysis time.

3.1.4. Effects of the separation voltage For a given capillary length, the separation voltage determines the electric field strength, which affects both the velocity of electrosmotic flow (EOF) and the migration velocity of the analytes, which in turn determines the migration time of the analytes. As expected, higher separation voltage gave shorter migration time for all analytes. However, when the separation voltage exceeded 22 kV, baseline noise became larger. Therefore the optimum separation voltage selected was 20 kV, at which good separation can be obtained within 9 min.

Fig. 2. The electrophograms of MM standard solution (A) and milk sample (B) in CZE-AD. Working electrode potential was +1150 mV (vs. SCE); running buffer: 20 mmol/L H3BO3–Na2B4O7 buffer (pH 7.4); peak identification: (1) MM (1.0  10 4 g/mL), (2) unknown compound existing in milk sample; other experimental conditions were the same as in Fig. 1.

16

MM Unkown Compound

14

12

t/min

Fig. 1. Hydrodynamic voltammogram (HDV) of MM in MECC–AD. Fused-silica capillary: 25 lm i.d.  75 cm; working electrode: 300 lm diameter carbon-disc electrode; running buffer: 8 mmol/L SDS/20 mmol/L H3BO3–Na2B4O7 buffer (pH 7.4); separation voltage: 20 kV; injection time: 6 s (at 20 kV); concentration of MM, 5.0  10 5 g/mL.

10

8

6

4 0

4

8

12

16

SDS concentration/mmol/L Fig. 3. Effect of SDS concentration on migration time of MM. Working electrode potential was +1150 mV (vs. SCE); and other experimental conditions and peak identification were the same as in Fig. 2.

3.1.5. Effects of the injection time The injection time determining the amount of sampling affects both peak current and peak shape. The effect of injection time on peak current was studied by varying injection time from 2 s to 10 s at 20 kV. It was found that the peak current increased with increasing sampling time. When the injection time was longer than 6 s, peak current nearly leveled off and peak broadening became more severe. In this experiment, 6 s (20 kV) was selected as the optimum injection time.

218

J. Wang et al. / Food Chemistry 121 (2010) 215–219

Fig. 4. The electropherograms of MM standard solution (A), milk sample (B), infant formula milk powder sample (C) and infant formula milk powder sample added MM standard solution (2.0  10 5 g/mL) (D) in MECC–AD. Peak identification: (1) MM (5.0  10 5 g/mL), and (2) unknown compound existing in milk or infant formula milk powder sample; other experimental conditions were the same as in Fig. 3.

Table 1 Assay results of recovery with real-world samples (n = 3).a Samples Liquid milk

Infant formula Milk powder a b

Ingredient MM

MM

Original amount (g/mL) N.F.

b

N.F.

Added amount (g/mL) 2.0  10 4.0  10 6.0  10

5

2.0  10 4.0  10 6.0  10

5

5 5

5 5

Found (g/mL) 1.95  10 3.48  10 5.00  10 2.11  10 3.92  10 5.30  10

5 5 5 5 5 5

Recovery (%)

RSD (%)

97.5 87.0 83.3

3.4 4.5 7.8

105.5 98.0 88.3

2.8 3.5 6.0

MECC–AD conditions were the same as Fig. 4. N.F. meant MM was not found in the tested sample.

Through the experiments above, the optimum conditions for determining MM have been decided. The applied potential to the working electrode was selected at +1150 mV (vs. SCE), the injection time was 6 s (20 kV), and MM can be detected within 9 min at the separation voltage of 20 kV in a 8 mmol/L SDS/20 mmol/L H3BO3– Na2B4O7 running buffer (pH 7.4). The typical electropherogram for a standard solution of MM was shown in Fig. 4A. 3.2. Linearity, detection limit and reproducibility of MM To determine the linearity of MM, a series of standard solutions from 2.0  10 6 g/mL to 1.0  10 4 g/mL were tested. The peak current and concentration of MM was subjected to regression analysis to calculate the calibration equations and correlation coefficients. The regression equation of MM was y = 6.05  104x 0.05 (correlation coefficient 0.9998) [the x value was the concentration of MM (g/mL), and the y value was the peak current (nA)]. The oxidation peak current of MM on carbon-disc electrode at +1150 mV showed good linear relationship in the concentration range of 1.0  10 4 5.0  10 6 g/mL. The detection limit of MM was 2.1  10 6 g/mL based on a signal-to-noise ratio of 3, which was comparable (5 lg/g) or better (65 lg/g) than those of reported methods (Ehling et al., 2007; Muniz-Valencia et al., 2008). The reproducibility of the peak current and migration time was estimated by making repetitive injections of a standard mixture solution (5.0  10 5 g/mL for each analyte) under the selected optimum conditions. The relative standard derivations (RSDs) of

the peak current and migration time were 3.2% and 3.8% for MM, respectively (n = 7). 3.3. Sample analysis and recovery Real-world samples including milk and infant formula milk powder were determined for evaluation of its feasibility for routine use, and the typical sample electropherograms of milk and infant formula milk powder were shown in Figs. 4B and C, respectively. By using standard addition method and comparing migration time of analyte with that of the standard solution (Fig. 4A), the compound MM in real-world samples can be identified and determined. From all the sample electropherograms, no signal peak of MM in the real samples could be observed, demonstrating that MM content in the newly produced milk samples was below the detection limit of the present method and the admissible limit (milk). To further evaluate the precision and accuracy of the method the recovery experiments under the optimum conditions were also conducted with real-world milk and infant formula milk powder samples (n = 3). Recovery was determined by standard addition method, and the results were listed in Table 1. The recovery was calculated by dividing the amount found by the amount added, then multiplying by 100%. The average recoveries were from 83.3% to 105.5% for high, middle and low concentration levels (n = 3) in this method, which indicated that this sample preparation method was suitable for the analysis of MM residue in real

J. Wang et al. / Food Chemistry 121 (2010) 215–219

samples. The recoveries obtained in this sample pre-treatment procedure were similar or better than the values reported in the literatures for the analysis of MM in real-world samples. The literature values of recovery reported were 93–105% (Muniz-Valencia et al., 2008), 82–105.6% (Wang, 2008b), 49.9–90.3% (Wendy et al., 2008), and 70.2–80.2% (Gong et al., 2008), respectively. Furthermore, the recovery electropherogram of infant formula milk powder sample after adding MM standard solution (2.0  10 5 g/ mL) was shown in Fig. 4D, which indicated that MM can be baseline separated from the unknown compound often existing in milk samples. 4. Conclusions The above experimental results demonstrated the capability and the advantages of the developed MECC–AD system for the fast determination of MM in real-world milk and infant formula milk powder samples with simple pre-treatment. Good recoveries ranging from 83.3% to 105.5% of MM were determined. The MECC–AD condition was optimised with a 8 mmol/L SDS/20 mmol/L H3BO3–Na2B4O7 running buffer (pH 7.4), and the separation was completed within 9 min. This research work is mainly focused on establishing a new analytical method for the determination of MM in milk products. The proposed method could be recommended as a routine method for the analysis of MM in real-world milk samples if its detection sensitivity can be further improved by means of sample pre-treatment such as MM preconcentration. Acknowledgements The authors were grateful for the financial support provided by the National Science Foundation of China (Grant No. 20875032), the Basic Research Fund of the Science and Technology Commission of Shanghai Municipality (Grant Nos. 08JC1409300 and 09zR1409700), the Daxia Foundation of ECNU (Ky 2009-124) and the opening foundation of the Key Laboratory of Green Pesticide and Agricultural Bioengineering, Ministry of Education, Guizhou University (Grant No. 2009GDGP0102). References Birgit, P., Robert, H. P., Linda, J. L., Michael, S. F., & Patricia, A. P. (2007). Assessment of melamine and cyanuric acid toxicity in cats. Journal of Veterinary Diagnostic Investigation, 19(6), 616–624. Cai, Q. R., OuYang, Y. Y., Qian, Z. J., & Peng, Y. F. (2008). Determination of melamine residue in feeds by ultra performance liquid chromatography coupled with electrospray tandem mass spectrometry. Chinese Journal of Chromatography, 26(3), 339–342. Chou, S. S., Hwang, D. F., & Lee, H. F. (2003). High performance liquid chromatographic determination of cyromazine and its derivative melamine in poultry meats and eggs. Journal of food and drug analysis, 11(4), 290–295. Chu, Q. C., Jiang, L. M., Tian, X. H., & Ye, J. N. (2008). Rapid determination of acetaminophen and p-aminophenol in pharmaceutical formulations using miniaturized capillary electrophoresis with amperometric detection. Analytical Chimica Acta, 606, 246–251. Chu, Q. C., Tian, X. H., Lin, M., & Ye, J. N. (2007). Application of CE in studying flavonoid/phenolic profiles of honeybee-collected pollen. Journal of Agricultural and Food Chemistry, 55(22), 8864–8869. Cook, H. A., Klampfl, C. W., & Buchberger, W. (2005). Analysis of melamine resins by capillary zone electrophoresis with electrospray ionization-mass spectrometric detection. Electrophoresis, 26(7–8), 1576–1583. Ding, T., Xu, J. Z., Li, J. Z., Shen, C. Y., Wu, B., Chen, H. L., et al. (2008). Determination of melamine residue in plant origin protein powders using high performance liquid chromatography–diode array detection and high performance liquid

219

chromatography-electrospray ionization tandem mass spectrometry. Chinese Journal of Chromatography, 26(1), 6–9. Ehling, S., Tefera, S., & Ho, I. P. (2007). High-performance liquid chromatographic method for the simultaneous detection of the adulteration of cereal flours with melamine and related triazine by-products ammeline, ammelide, and cyanuric acid. Food Additives & Contaminants, 24(12), 1319–1325. European Food Safety Authority (EFSA) (2007) . Question No. EFSA-Q-2007-093. Food and Drug Administration (FDA). GC–MS screen for the presence of melamine, ammeline, ammelide and cyanuric acid. Version 2.1. . Food and Drug Administration (FDA) (2007). Interim melamine and analogues safety/risk assessment. May 25. Filigenzi, M. S., Tor, E. R., Poppenga, R. H., & Aston, L. A. (2007). The determination of melamine in muscle tissue by liquid chromatography/tandem mass spectrometry. Rapid Communications in Mass Spectrometry, 21(24), 4027–4032. General Administration of Quality Supervision, Inspection and Quarantine of People’s Republic of China. Determination of melamine in raw milk and dairy products, GB/T 22388-2008. Gong, X. M., Dong, J., Sun, J., Wang, H. T., Zhao, S. Y., & Zhu, L. Q. (2008). Determination of melamine in vegetable materials by high performance liquid chromatography. Food Science, 29(4), 321–323. Jiang, C. Y., Fan, J., Lin, D. Q., & Li, H. N. (2008). Determination of melamine residue in feeds by GC–MS. Feed Industry, 29(8), 48–50. Kim, B., Perkins, L. B., Bushway, R. J., Nesbit, S., Fan, T., Sheridan, R., et al. (2008). Determination of melamine in pet food by enzyme immunoassay, highperformance liquid chromatography with diode array detection, and ultraperformance liquid chromatography with tandem mass spectrometry. Journal of AOAC International, 91(2), 408–413. Li, A. J., Zhang, D. H., Ma, S. M., Li, M., Zhang, X. Z., & Yao, T. L. (2008). Determination of melamine residues in feed by liquid chromatography tandem mass spectrometry. Chinese Journal of Analytical Chemistry, 36(5), 699–701. Muniz-Valencia, R., Ceballos-Magana, S. G., Rosales-Martinez, D., GonzaloLumbreras, R., Santos-Montes, A., Cubedo-Fernandez-Trapiella, A., et al. (2008). Method development and validation for melamine and its derivatives in rice concentrates by liquid chromatography. Application to animal feed samples. Analytical Bioanalytical Chemistry, 392, 523–531. National Toxicology Program. (1983). TR-245 Carcinogenesis bioassay of melamine (CAS No. 108-78-1) in F344/N rats and B6C3F1 mice (feed study). . Nielen, M. W. F., Henk, J. F. M., & Van de Ven (1996). Characterization of (methoxymethyl) melamine resin by combined chromatographic/mass spectrometric techniques. Rapid Communications in Mass Spectrometry, 10(1), 74–81. Pichon, V., Chen, L., Guenu, S., & Hennion, M. C. (1995). Comparison of sorbents for the solid-phase extraction of the highly polar degradation products of atrazine (including ammeline, ammelide and cyanuric acid). Journal of Chromatography A, 711, 257–267. State Department, People’s Republic of China. (2008). 22 Milk plants 69 batches of products found melamine. 17 September. . Thuy, D. T. V., Markus, H., Manuela, H., Wolfgang, B., Clemens, S., & Christian, W. K. (2008). Improved analysis of melamine–formaldehyde resins by capillary zone electrophoresis–mass spectrometry using ion-trap and quadrupole-time-offlight mass spectrometers. Journal of Chromatography A, 1213(1), 83–87. Wang, H., Liu, Y. Q., Cao, H., Yang, H. M., Liu, X. L., & Yan, L. B. (2008). Determination of melamine in pet food by solid phase extraction coupled with high performance liquid chromatography. Chinese Journal of Analytical Chemistry, 36(2), 273. Wang, J. M. (2008a). Current production status and market analysis of melamine. Chemical Engineering Design, 18(2), 47–51. Wang, Z. (2008b). Determination of melamine and the related analogs in animal food by gas chromatography–mass spectrometry. Fujian Analytical Testing, 17(2), 1–4. Wendy, C. A., Sherri, B. T., Christine, M. K., Susan, B. C., Mark, R. M., Charles, M. G., et al. (2008). Determination and confirmation of melamine residues in catfish, trout, tilapia, salmon, and shrimp by liquid chromatography with tandem mass spectrometry. Journal of Agricultural and Food Chemistry, 56(12), 4340–4347. Wu, M. L., & Chen, C. H. (2005). Determination of melamine in cyanamide by high performance liquid phase chromatography (HPLC). Ningxia Petroleum and Chemical Industry, 2, 24–26. Wu, T., Guan, Y. Q., & Ye, J. N. (2007). Determination of flavonoids and ascorbic acid in grapefruit peel and juice by capillary electrophoresis with electrochemical detection. Food Chemistry, 100(4), 1573–1579. Yuan, L. Y., Ma, Z. W., & Du, Y. H. (2004). Fast determination of the content of melamine in solution. Henan Chemical Industry, 4, 42.