Journal of Chromatography B, 960 (2014) 87–91
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Simultaneous determination of 16 synthetic colorants in hotpot condiment by high performance liquid chromatography Bobin Tang a,b , Cunxian Xi b , Yun Zou c , Guomin Wang b , Xianliang Li b , Lei Zhang b , Dongdong Chen d , Jinzhong Zhang a,∗ a College of Resources and Environment, Southwest University, Key Laboratory of Eco-environments in Three Gorges Reservoir Region, Ministry of Education, Chongqing 400715, China b Chongqing Entry-Exit Inspection and Quarantine Bureau, Chongqing Engineering Technology Research Center of Import and Export Food Safety, Chongqing 400020, China c Chongqing Municipal Environmental Monitoring Center, Chongqing 401121, China d Chinese Academy of Inspection and Quarantine, Beijing 100123, China
a r t i c l e
i n f o
Article history: Received 21 September 2013 Accepted 10 April 2014 Available online 21 April 2014 Keywords: Hotpot condiment Synthetic colorant Simultaneous determination HPLC
a b s t r a c t A simultaneous determination method for 16 synthetic colorants in hotpot condiment was developed by high performance liquid chromatography. The samples were successively extracted with 2 mol/L carbamide solution containing 5% ammonia (dissolved in methanol) and methanol–acetone solution, and then the target analytes could be divided into two groups named as lipid-soluble and water-soluble colorants by ethyl acetate-cyclohexane with liquid–liquid extraction. The lipid-soluble and water-soluble colorants were puriﬁed by gel permeation chromatography and solid phase extraction column packed with polyamide resin, respectively. The obtained two eluates were combined, concentrated, and separated by C18 column and determined by diode array detector. Good linear relationships between peak areas and the concentrations of the synthetic colorants were obtained in the range of 0.01–50.0 mg/L with correlation coefﬁcients above 0.999 (n = 10). The limits of detection and quantitation were 1–3 and 10 g/kg for 16 synthetic colorants, respectively. The average recoveries at the spiked levels of 5, 10, 20 and 50 g/kg were in the range of 63.2–97.1% with relative standard deviations (n = 6) around 1.5–10.6%. This method is sensitive and reliable, and can be used to simultaneously determine 8 lipid-soluble and 8 water-soluble colorants in hotpot condiment. © 2014 Elsevier B.V. All rights reserved.
1. Introduction In general, synthetic colorants can be classiﬁed into watersoluble and lipid-soluble fractions based on their solubility. Some water-soluble colorants, such as new red, carmine, amaranth, allura red, erythrosine and sunset yellow, are permitted to add into food in China , due to their strong coloring ability, stability and low cost . However, many azo dyes can be converted into harmful derivatives in human body, e.g. carmine, allura red and sunset yellow can be reduced to aromatic amines by anaerobic bacteria from the human intestinal tract, which may cause frequent headaches for adults, and distraction and hyperaction for children . Most of lipid-soluble colorants (e.g. Sudan dyes and rhodamine B) and their metabolites have carcinogenic, teratogenic and mutagenic effects
∗ Corresponding author. Tel.: +86 23 68250484; fax: +86 23 68250444. E-mail address: [email protected]
(J. Zhang). http://dx.doi.org/10.1016/j.jchromb.2014.04.026 1570-0232/© 2014 Elsevier B.V. All rights reserved.
, and thus are not allowed to add into food and complex type ﬂavoring. To prevent excessive use of water-soluble colorants, some countries and regions have legislated laws and regulations to limit the types, purities, uses and amounts permitted to use in food and drinks. Recently, the maximum limits of allura red and carmine have been laid down as 300 and 500 mg/kg in complex type ﬂavoring by Codex Alimentarius Commission (CAC) , while that of allura red, erythrosine and sunset yellow as 40, 50 and 200 mg/kg by China , respectively, and other water-soluble colorants are forbidden to use. In USA, Japan and European Union (EU), synthetic colorants have been strictly prohibited to use in complex type ﬂavoring [6–8]. Due to different maximum limits applicable for raw materials and complex type ﬂavoring, people often ﬁnd that complex type ﬂavoring may not be safe if an individual material dyed by the permitted synthetic colorants below the maximum limits is safe. Chongqing hotpot is notable in China and abroad for its spicy and delicious taste, and bright red color. After the soup mixed with hotpot condiment being boiled, people can directly and instantly
B. Tang et al. / J. Chromatogr. B 960 (2014) 87–91
cook meat, vegetables, aquatic products, etc. Hotpot condiment is a complex type ﬂavoring which mainly contains beef tallow, vegetable oil, chilli, ginger, garlic, salt, Chinese prickly ash, broad-bean sauce and spice materials, these components are responsible for the natural color of hotpot soup. However, some synthetic colorants were found to excessively or illegally adding in food to enhance its visual esthetics and to promote sales, which leads to food safety incidents from time to time. It was reported that sunset yellow, allura red, Sudan I and IV were detected in 19 batches of export food owing to the illegal use during the period of 2009–2013 in China . In 2011, a large quantity of broad-bean sauce dyed by carmine and sunset yellow, and pepper and hotpot condiment dyed by rhodamine B were seized in Chongqing, China . Therefore, it is an urgent requirement to develop sensitive and reliable method to detect multiple synthetic colorants in hotpot condiment. Until now, the main analytical methods for synthetic colorants in food include capillary electrophoresis (CE) , spectrophotometry , thin-layer chromatography (TLC) , electrochemical detection [12–14], high performance liquid chromatography (HPLC) coupled with UV–Vis detector , diode array detector (DAD) [16,17], mass spectrometry (MS) [18,19], and tandem mass spectrometry (MS/MS) [20,21]. However, most of these methods can only be suitable for the same kind of synthetic colorants, water-soluble or lipid-soluble colorants, and simple matrix, e.g. meat, confectionery, soft drink, fruit juice, gelatine powder and chilli products. For a complicated matrix with high oil content, e.g. hotpot condiment, it is a great challenge to effectively extract and purify, and avoid matrix disturbance in simultaneous determination of multiple water-soluble and lipid-soluble colorants. Long et al.  successfully determined 3 water-soluble and 6 lipid-soluble colorants in chilli products by using HPLC-DAD with methanol extraction and molecularly imprinted polymer SPE column cleanup, and obtained satisfactory recoveries and detection limits. However, the preparation process of the molecularly imprinted polymer is very cumbersome and time-consuming, and chilli products are simple matrices versus hotpot condiment. Although LC–MS and LC–MS/MS possess high throughput for simultaneous determination of multiple synthetic colorants in food with high sensitivity and selectivity, they are often not applicable in ordinary labs. So a high throughput detection method based on HPLC coupled with DAD or UV–Vis detector may have more practical application in routine analysis. To the best of our knowledge, no report on the simultaneous determination of synthetic colorants in hotpot condiment can be found. In this work, the objectives were to optimize the conditions of extraction, puriﬁcation and detection for 8 lipid-soluble and 8 water-soluble colorants, and to develop a simultaneous determination method for them probably existed in hotpot condiment.
2. Material and methods 2.1. Chemicals and materials Solid standards of new red (92%), carmine (90%), amaranth (90%), allura red (90%), erythrosine (98%), sunset yellow (90%), acid red G (90%), acid scarlet GR (97%), rhodamine B (95%), para red (95.5%), Sudan I (90.5%), Sudan II (99%), Sudan III (97%), Sudan IV (91%), Sudan orange G (98.8%) and Sudan red 7B (99%) were purchased from Dr. Ehrenstorfer Co. (Augsburg, Germany). Methanol and acetone (HPLC grade) were purchased from Tedia (Fairﬁeld, OH, USA). Cyclohexane, ethyl acetate, carbamide, disodium hydrogen phosphate, anhydrous sodium dihydrogen phosphate, phosphoric acid and ammonium (analytical reagents) were obtained from Chuandong Chemical Co. (Chongqing, China).
Solid phase extraction (SPE) column packed with polyamide resin was assembled as follows: a piece of thin adsorbent cotton was placed into the bottom of a glass column (2 cm × 15 cm), and ﬁlled with the thin paste formed by 1.0 g of polyamide resin (70–150 m, Fluka) and 10 mL of 0.1% phosphoric acid solution. Hydrophilic polyethersulfone membrane (0.45-m) was purchased from Tianjin Jinteng Instrument Co., Ltd. (Tianjin, China). 2.2. Instrumentation and conditions Sample analysis was carried out by Agilent 1200 Series HPLC system (USA), which was consisted of an HPLC pump operating at a ﬂow rate of 1.0 mL/min, and DAD monitoring the efﬂuent at 420 nm for Sudan Orange G, and 520 nm for the other synthetic colorants. The separation column was Venusil MP C18 column (4.6 × 250 mm, 5 m), and the column temperature was kept at 30 ◦ C. The injection volume was 10 L. The mobile phases were consisted of A (methanol) and B (0.01 mol/L phosphate buffer solution, pH 7.5). The gradient elution program was used as follows: 0–14 min, 10–100% A; 14–30 min, 100–10% A, and held for 5 min. Gel permeation chromatography (GPC, J2 Scientiﬁc Corporation, USA) was equipped with a puriﬁcation column (400 mm × 25 mm) packed with Bio-Beads (S-X3, 38–75 m). 2.3. Preparation and treatment of the samples 2.3.1. Extraction Eleven brands of hotpot condiment were purchased in a local supermarket in Chongqing, China. One hundred grams of sample for every brand were mixed well in a commercial laboratory blender, and stored into a 0–4 ◦ C fridge until analysis. Five grams of the homogenized sample and 20 mL of methanol–acetone (1:1, v/v) were transferred into a 50-mL polypropylene centrifuge tube, heated in 60 ◦ C water bath for 20 min, extracted with vortex mixing for 2 min, and then centrifuged at 5000 rpm for 5 min in a centrifuge (Hermle, Germany). The extraction was repeated with 20 mL of methanol–acetone (1:1, v/v), 2× 15 mL of 2 mol/L carbamide solution containing 5% ammonia (dissolved in methanol). Two extracts were combined and concentrated to dryness in a rotary evaporator (Büchi Rotavapor R-200, Switzerland) at 40 ◦ C, washed with 2× 5 mL of ethyl acetate–cyclohexane (1:1, v/v), 2× 5 mL of water, and transferred into a 50-mL polypropylene centrifuge tube, mixed well and centrifuged. The organic layer was transferred into a 10-mL glass tube, concentrated to dryness with N-Evap (Organomation Associates, USA) at 40 ◦ C, and dissolved in 5 mL of ethyl acetate–cyclohexane (1:1, v/v), the obtained solution was designated as solution 1. The aqueous phase was supplemented with 1 mL of 50% phosphoric acid solution to adjust pH to 3, and the obtained solution was designated as solution 2. 2.3.2. Puriﬁcation Solution 1 (4.5 mL) was transferred into GPC column, and eluted by ethyl acetate–cyclohexane (1:1, v/v) at a ﬂow rate of 4.7 mL/min. The eluate was collected from 13 to 20 min, and was designated as solution 3. Solution 2 was completely passed through the SPE column at a ﬂow rate of 3 mL/min, washed the column with 5 mL of 0.1% phosphoric acid and 5 mL of 80% methanol containing 0.1% phosphoric acid and dried under vacuum, and washed the column with 3 mL of 5% ammonia–methanol and dried again. The column was eluted with 7 mL of 5% ammonia–methanol, the eluate was transferred into a 10-mL volumetric ﬂask and diluted to 10 mL with methanol. This solution was designated as solution 4. 2.3.3. Concentration Solution 4 (9 mL) and solution 3 were combined in a rotary evaporation bottle, and then concentrated to dryness at 40 ◦ C. The bottle
B. Tang et al. / J. Chromatogr. B 960 (2014) 87–91
was washed with 5 mL methanol–acetone (1:1, v/v) for several times, and transferred into a test tube and concentrated to dryness with N-Evap below 40 ◦ C. The residue was dissolved with 1 mL of methanol–acetone (1:1, v/v), and ﬁltered through a hydrophilic polyethersulfone membrane for HPLC analysis. 3. Results and discussion 3.1. Optimization of sample extraction conditions 3.1.1. Extraction method In general, the content of beef tallow in hotpot condiment is in a range of 30–80%. As a solid substance at the normal temperature, the melting point of beef tallow is over 43 ◦ C , and thus the extraction solvent is difﬁcult to penetrate into the sample. In this work, the extraction efﬁciencies of several methods were examined, such as vortex mixing, ultrasound, homogeneity, and heating in water bath. The results showed that the samples could not fully contact with the solvent using vortex mixing and ultrasonic extraction. Although the samples could fully disperse using homogeneity, the recovery decreased because some samples were adhered to the homogenizer. The extraction efﬁciency of heating in water bath was the highest, because this method could melt beef tallow, let the samples and solvent fully contact, and made the target analytes to release easily. The recoveries of the synthetic colorants in hotpot condiment extracted by heating in water bath could reach above 70% except new red and amaranth. 3.1.2. Extraction solvent At present, some studies used methanol–water , water  and ammonia–ethanol  to extract water-soluble colorants, and methanol , acetonitrile , acetone  and hexane  to extract lipid-soluble colorants, and methanol  and dimethyl sulfoxide  to simultaneously extract water-soluble and lipidsoluble colorants. In this work, the extraction efﬁciencies of 5% ammonia–methanol, methanol–acetone (1:1, v/v), 5% ammonia methanol–acetone (1:1, v/v), methanol–acetone (1:1, v/v) and 2 mol/L carbamide solution containing 5% ammonia (dissolved in methanol) were examined. The results showed that the recoveries could reach 70–100% by methanol–acetone and 2 mol/L carbamide solution containing 5% ammonia with sequential extraction. 3.2. Optimization of the puriﬁcation conditions Because of the large difference in polarity of the target analytes, single puriﬁcation method cannot obtain satisfactory recovery. In this work, the synthetic colorants in hotpot condiment were grouped based on solubility difference with liquid–liquid extraction, and then puriﬁed individually. The extraction percents of 16 synthetic colorants in 4 mol/L carbamide solution were examined using ethyl acetate, hexane, dichloromethane and ethyl acetate–cyclohexane (1:1, v/v). The results indicated that the target analytes could be divided into water-soluble and lipid-soluble fractions by ethyl acetate–cyclohexane, and the recoveries of the lipid-soluble colorants are shown in Table 1. Water-soluble colorants are commonly puriﬁed using polyamide resin column , C18 column  and HLB column . In this work, we examined the retention capacities of C18 , HLB, MCX, HR-X and polyamide resin columns. The results showed that all of them had high retention capacities for the water-soluble colorants, but the retention capacities of MCX, C18 , HR-X and HLB columns were affected by methanol content and the acidity of the washing solution. The recoveries were lower than 15% when the pH of the washing solution was over 5 and the methanol content was over 80%. When the pH of the washing solution was adjusted to below 5 and the methanol content below
80%, polyamide resin column could keep high retention capacity, and most of natural colorants were washed away, and thus the recoveries spiked in all the blank sample extracts could reach above 95%. Lipid-soluble colorants are commonly puriﬁed with liquid–liquid extraction , neutral alumina column  and GPC [21,30]. Compared with liquid–liquid extraction and neutral alumina column, GPC can easily remove beef tallow and natural colorants, and the cleanup process is steady and easy to control. In this work, the recoveries of lipid-soluble colorants spiked in the blank extracts could reach 95–110%. 3.3. Optimization of the chromatographic conditions 3.3.1. Selection of the ﬁlter membrane Filter membrane can selectively adsorb colorants. In this study, hotpot condiment samples were dissolved in methanol–acetone (1:1, v/v), and the adsorption capacities of polypropylene, hydrophilic polyether sulfone and nylon 66 membranes to 16 synthetic colorants were examined. The results showed that nylon 66 membrane could strongly adsorb new red, amaranth and carmine, and the recoveries of the three standard solutions ﬁltered through the membrane were only 34%, 36% and 41%, respectively. However, polypropylene and hydrophilic polyether sulfone membranes could adsorb the target analytes very little, and the recoveries of 16 standard solutions ﬁltered through them were in the range of 97–103%. In this work, hydrophilic polyether sulfone membrane was used due to the low cost. 3.3.2. Selection of the mobile phases Water-soluble colorants are commonly eluted and separated by methanol–ammonium acetate [19,28] or acetonitrile–ammonium acetate [20,24] in the chromatographic column, while lipid-soluble colorants by acetonitrile–ammonium acetate [15,16] with gradient elution. In this work, the effects of methanol–0.02 mol/L ammonium acetate, methanol–0.01 mol/L phosphate and methanol–0.02 mol/L phosphate on the resolution and chromatographic peak were examined with C18 column. It was found that the pressure ﬂuctuation appeared using methanol–0.02 mol/L phosphate with gradient elution, erythrosine and Sudan orange G could not be easy to separate with methanol–0.02 mol/L ammonium acetate, and some tailing appeared in the chromatograms for new red and amaranth. The target analytes could be well separated, and good chromatographic peaks could be obtained using methanol–0.01 mol/L phosphate as mobile phases with gradient elution. 3.4. Validation of the method 3.4.1. Linear range and detection limit Under the optimum conditions of extraction, puriﬁcation and detection, good linear relationships between the peak areas (A) and the concentration (C) of 16 synthetic colorants were obtained in the range of 0.01–50.0 mg/L with correlation coefﬁcients above 0.999 (n = 10), the limits of detection (LODs, S/N = 3) and quantitation (LOQs, S/N = 10) were 1–3 and 10 g/kg, respectively, and the results are shown in Table 2. The chromatograms of the blank, spiked, and positive hotpot condiment samples are shown in Fig. 1. Compared with other studies using HPLC, the LODs obtained by this study were close to those of carmine, amaranth, para red and Sudan I–IV in pepper products [16,19], and 2–10 times lower than those of new red, sunset yellow, allura red and erythrosine in jam and meat products [18,24]. Thus, this method can meet the demand for the simultaneous determination of lipid-soluble and water-soluble colorants in food.
B. Tang et al. / J. Chromatogr. B 960 (2014) 87–91
Table 1 Extraction percents of lipid-soluble colorants using different solvents. Compound
Sudan orange G Rhodamine B Para red Sudan I Sudan II Sudan III Sudan red 7B Sudan IV
Extraction percent (%) Ethyl acetate
99.6 99.7 99.9 99.8 99.6 99.8 99.7 97.7
99.9 99.7 99.8 99.8 99.7 99.9 99.7 99.7
99.7 99.9 99.8 99.8 99.9 99.9 99.8 82.6
99.5 99.6 99.8 99.9 99.8 99.9 99.8 38.2
Table 2 Linear ranges, regression equation, correlation coefﬁcients and LODs of 16 synthetic colorants. Compound New red Amaranth Carmine Sunset yellow Acid red G Allura red Acid scarlet GR Erythrosine Rhodamine B Sudan I Para red Sudan II Sudan III Sudan red 7B Sudan IV Sudan orange G
Linear range (mg/L) 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0 0.05–50.0
A = 27.11C + 1.57 A = 20.04C + 0.32 A = 17.59C + 0.34 A = 15.00 C + 1.27 A = 22.51C + 0.11 A = 27.25C + 0.40 A = 33.05 C − 0.13 A = 40.64C + 0.53 A = 45.64 C − 0.22 A = 22.99C + 0.06 A = 19.44C + 0.31 A = 30.63 C − 0.76 A = 47.99 C − 1.13 A = 40.94 C − 1.65 A = 53.73 C − 2.91 A = 73.65C + 0.85
0.9999 1.000 0.9999 0.9998 0.9999 0.9999 0.9996 0.9999 0.9999 0.9999 0.9999 0.9999 0.9999 0.9997 0.9998 0.9999
0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.003 0.003 0.003 0.003 0.003 0.003 0.001
3.4.2. Recovery The standard solutions were spiked into the blank hotpot condiment samples at four concentration levels (5, 10, 20 and 50 g/kg), the obtained average recoveries were in the range of 63.2–97.1% with relative standard deviations (RSDs, n = 6) around 1.5–10.6%
shown in Table 3 (Supplementary Data). The results in this study were consistent with those of amaranth, carmine, allura red, para red and Sudan orange G, and Sudan I–IV in chilli determined by HPLC with external standard quantitation . However, the compared study  did not determine new red, sunset yellow, erythrosine, acid red G, acid scarlet GR, rhodamine B and Sudan 7B. 3.4.3. Analysis of real samples Synthetic colorants in 11 brands of hotpot condiment were determined by the developed method, and the obtained results are shown in Table 4 (Supplementary Data). Among them, sunset yellow could be detected in Brand 9, Sudan I in Brand 4, and rhodamine B in Brand 2, 9 and 10. However, sunset yellow content (3.4 g/kg) in Brand 9 was much lower than the maximum limit (200 mg/kg) in complex type ﬂavoring (China standard, Reference 1). As the forbidden colorants, Sudan I and rhodamine B contents (below 4 g/kg) in positive samples were lower than their LOQs (10 g/kg), which should be further conﬁrmed by LC–MS/MS. LC–MS/MS results indicated that Sudan I was not found in Brand 4, and rhodamine B was not found in Brand 2 and 10, whereas rhodamine B was detected as 3.1 g/kg in Brand 9. Rhodamine B at such low level may originate from the raw materials or environmental pollution rather than illegal addition. 4. Conclusions
Fig. 1. Chromatograms of the blank, spiked, and positive hotpot condiment samples. (a) Blank hotpot condiment; (b) spiked with 5 g/kg of standards; (c) positive sample. (1) New red; (2) amaranth; (3) carmine; (4) sunset yellow; (5) allura red; (6) acid red G; (7) acid scarlet GR; (8) erythrosine; (9) Sudan orange G; (10) rhodamine B; (11) para red; (12) Sudan I; (13) Sudan II; (14) Sudan III; (15) Sudan red 7B; (16) Sudan IV.
Synthetic colorants in hotpot condiment were sequentially extracted with 2 mol/L carbamide solution containing 5% ammonia (dissolved in methanol) and methanol–acetone, separated by liquid–liquid extraction, puriﬁed, and simultaneously determined by HPLC with diode array detector. This method can effectively reduce matrix interference, and the recoveries, precision and detection limits can meet the demand for residue analysis of lipid-soluble and water-soluble colorants in food, and is hopeful to be used in
B. Tang et al. / J. Chromatogr. B 960 (2014) 87–91
simultaneous determination of 16 synthetic colorants in hotpot condiment. Acknowledgements This work was supported by the scientiﬁc program of General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China (No. 2009IK185). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jchromb. 2014.04.026. References  Standards for Use of Food Additives (GB 2760-2011), Ministry of Health, People’s Republic of China.  H.Y. Huang, Y.C. Shih, Y.C. Chen, J. Chromatogr. A 959 (2002) 317.  F. Raﬁi, J.D. Hall, E.D. Cerniglia, Food Chem. Toxicol. 35 (1997) 897.  K. Golka, S. Kopps, Z.W. Myslak, Toxicol. Lett. 151 (2004) 203.  WTO/TBT-SPS Notiﬁcation and Enquiry of China, http://www.tbt-sps.gov. cn/Pages/home.aspx  Commission Decision 2005/402/EC, Ofﬁcial Journal of the European Union, L135/34.  Food May be Illegal to Add Non-edible Substance Abuse and Easy List of Food Additives, Ministry of Health, People’s Republic of China.  European Parliament and Council Directive 94/36/EC, Ofﬁcial Journal of the European Communities, L237/13.
 http://biz.ifeng.com/city/chongqing/zhongqingxinwen/detail 2011 06/15/ 40848 0.shtml  F. Capitán, L.F. Capitán-Vallvey, M.D. Fernández, I. de Orbe, R. Avidad, Anal. Chim. Acta 331 (1996) 141.  F. Soponar, A.C. Mot¸, C. Sârbu, J. Chromatogr. A 1188 (2008) 295.  S. Chanlon, L. Joly-Pottuz, M. Chatelut, O. Vittori, J.L. Cretier, J. Food Compos. Anal. 18 (2005) 503.  T. Gan, J.Y. Sun, W. Meng, L. Song, Y.X. Zhang, Food Chem. 141 (2013) 3731.  R.A. Medeiros, B.C. Lourencao, R.C. Rocha-Filho, O. Fatibello-Filho, Talanta 99 (2012) 883.  C.Y. Long, Z.B. Mai, X.F. Yang, B.H. Zhu, X.M. Xu, X.D. Huang, X.Y. Zou, Food Chem. 126 (2011) 1324.  C.Y. Long, Z.B. Mai, Y.F. Yang, B.H. Zhu, X.M. Xu, L. Lu, X.Y. Zou, J. Chromatogr. A 1216 (2009) 8379.  Y.H. Zhu, B. Zhao, R.Q. Xiao, W. Yun, Z.J. Xiao, D.W. Tu, S.Q. Chen, Food Chem. 145 (2014) 956.  M.R. Fuh, K.J. Chia, Talanta 56 (2002) 663.  M. Ma, X.B. Luo, B. Chen, S.P. Su, S.Z. Yao, J. Chromatogr. A 1103 (2006) 170.  T.T. Zou, P.L. He, A. Yasen, Z. Li, Food Chem. 138 (2013) 1742.  M. Oplatowska, P.J. Stevenson, C. Schulz, L. Hartig, C.T. Elliott, Anal. Bioanal. Chem. 401 (2011) 1411.  G.H. Li, C.T. Wang, Y.J. Zhang, J. Li, J. Henan Univ. Technol. (Natural Science Edition) 31 (2010) 30 (in Chinese).  N. Yoshioka, K. Ichihashi, Talanta 74 (2008) 1408.  G.L. Qie, W.H. Li, L.G. Luo, Anal. Chim. Acta 716 (2012) 128. ˜ P. Fernández-Hernando, J.S.  L. Enríquez-Gabeiras, A. Gallego, R.M. Garcinuno, Durand, Food Chem. 135 (2012) 193.  C.K. Zacharis, F.S. Kika, P.D. Tzanavaras, P. Rigas, E.R. Kyranas, Talanta 84 (2011) 480.  P. Qi, T. Zeng, Z.J. Wen, X.Y. Liang, X.W. Zhang, Food Chem. 125 (2011) 1462.  M. González, M. Gallego, M. Valcárcel, Anal. Chem. 75 (2003) 685.  F. Feng, Y.S. Zhao, W. Yong, L. Sun, G.B. Jiang, X.G. Chu, J. Chromatogr. B 879 (2011) 1813.  O. Pardo, V. Yusà, N. León, A. Pastor, Talanta 78 (2009) 178.