Analytica Chimica Acta 716 (2012) 128–132

Contents lists available at SciVerse ScienceDirect

Analytica Chimica Acta
journal homepage:

Applicability of accelerated solvent extraction for synthetic colorants analysis in
meat products with ultrahigh performance liquid chromatography–photodiode
array detection
Qie Gen Liao, Wei Hong Li, Lin Guang Luo ∗
Jiangxi Academy of Agricultural Sciences Key Laboratory of Quality and Safety of Agricultural Products, Agricultural Product Quality Safety and Standards Institute, Jiangxi Academy
of Agricultural Sciences, Nanchang, 330200, PR China

a r t i c l e

i n f o

Article history:
Received 8 September 2011
Received in revised form
14 December 2011
Accepted 15 December 2011
Available online 24 December 2011
Ultrahigh performance liquid
chromatography (UHPLC)
Photodiode array detection (PDA)
Synthetic colorants
Accelerated solvent extraction (ASE)
Meat products

a b s t r a c t
Accelerated solvent extraction (ASE) coupled with ultrahigh performance liquid chromatography
(UHPLC) with photodiode array detection (PDA) has been used for the quantitative determination of
synthetic colorants in meat products. Samples were extracted with ethanol–water–ammonia with a
ratio of 75:24:1 (v/v/v) using ASE instrument at 85 ◦ C. As a result, all the colorants in meat products were
separated using an optimized gradient elution within 3.5 min. Detection and quantification limits of synthetic colorants were in the ranges of 0.01–0.02 mg kg−1 and 0.05 mg kg−1 , respectively. The intra-day
and inter-day precision of the synthetic colorants were ranged between 1.7% (E123) to 5.2% (E124) and
3.2% (E124) to 6.0% (E129), respectively. The intra-day and inter-day recoveries of the synthetic colorants
were ranged between 76.9% (E124) to 84.9% (E102) and 76.3% (E124) to 84.3% (E127), respectively. The
method has been applied for the determination of seven synthetic colorants in meat products.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction
Food colorants are very interesting classes of food additives.
Frequently color of a product determines its attractiveness to consumers. Incentive hues are associated with freshness, good taste,
and nutritional value. However, natural colorants are unstable and
easily undergo degradation during the food processing and storage.
Moreover, in poultry farming, because of ␤-carotene or carotenoid
deficiency or short feeding periods, poultry products may be of
lighter colors that do not meet the requirements of consumers. Synthetic colorants are a very important class of food additives, which
are widely used to compensate for the lack of natural colors. However, some researchers have confirmed their negative influence on
human organism. For instance, azodyes are decomposed by natural intestinal flora into aromatic amines [1], which cause frequent
headaches in adults and cause children to become distractible and
hyperactive [2]. Thus, content of synthetic dyes in food must be
strictly controlled. In China, the use of synthetic colorants in meat
products is strictly controlled by Directive GB 2760-2011 of the
Ministry of Health. Consequently, accurate and reliable methods

∗ Corresponding author. Tel.: +86 791 87090293; fax: +86 791 87090293..
E-mail address: (L.G. Luo).
0003-2670/$ – see front matter © 2011 Elsevier B.V. All rights reserved.

for the determination of synthetic colorants are required for the
safety assurance of meat products.
Several analytical techniques have been developed for the
identification and determination of various synthetic colorants,
including adsorptive voltammetry [3], differential pulse polarography [4], derivative spectrometry [5] and spectrophotometric
methods in combination with chemometrics [6,7]. Unfortunately,
these methods cannot be applied to complex colorant mixtures or
simultaneous determination of various synthetic colorants. Capillary electrophoresis [8–10] and micellar electrokinetic capillary
chromatography [11] have also been used, however, these methods have sensitivity issues due to their small injection volume.
High-performance ion chromatography [12], reversed-phase liquid chromatography [13–16] and ion-pair liquid chromatography
[17–23] coupled with UV or diode-array detectors, continue to be
the most preferred methods due to unrivalled resolution, sensitivity, and selectivity; however, analysis speeds need to be improved.
The use of ultrahigh-performance liquid chromatography (UHPLC)
could improve analysis speed [24]. Recently, analytical methods, based on liquid chromatography–mass spectrometry (LC–MS)
technique [25–27], have been developed in order to identify and
quantify synthetic colorants or their degradation [28].
The preparation of samples is also a very important problem
in the analysis of synthetic colorants. Extracting them from different types of matrices is not an easy task. Most of the above

static time and cycles) are optimized to achieve the maximum extraction efficiencies for synthetic colorants. The proposed method has been applied to the analysis of meat products from local commercial and farmers’ markets. Each cell was locked with stainless steel screw caps equipped with Teflon O-ring seals. Amaranth (E123). ammonia–water–acetonitrile. ammonia and hexane were obtained from Shanghai chemicals company (Shanghai. The spiked samples were mixed 1 min by a vortex mixer. temperature. ammonia–water–ethanol. 1.70 min–2.e. and were set aside for a minimum of 1 h before extraction. All chemicals were of analytical reagent grade. The free space in the cell was filled up with hydromatrix. Conditions of the extraction were as follows: time heating cell. China). 2. In our preparatory experiments. Reagents Ammonium acetate was obtained from Riedel-de Haën (Seelze. The detection wavelengths were as follows: 0 min–1. acetonitrile containing 1% (v/v) ammonia. Accelerated solvent extraction (ASE) is a recent advancement in sample preparation for trace analytes.80 min–3. Erythrosine (E127) and Allura Red (E129) containing 0. 2. 625 nm.e. and candy) [12–23. 522 nm.31]. 429 nm. The aim of the current study is the development of a rapid UHPLC method for the simultaneous determination of seven synthetic colorants in meat products using an optimized gradient elution. / Analytica Chimica Acta 716 (2012) 128–132 129 Fig. 483 nm.90 min. USA). Germany). ASE uses conventional solvents at elevated pressures and temperatures to extract samples quickly [30. Approximately 5 g of the spiked blank/sample material mixed with 10 g of hydromatrix was packed in a 34 mL stainless steel extraction cell. the samples were stored below −20 ◦ C. the measures mentioned above were unsatisfactory in the extraction of the low concentration colorants. 5 min.G.0 cm diameter were placed above and below the packing. The HPLC grade organic solvents methanol (MeOH) and acetonitrile (ACN) were supplied by Merck (Darmstadt.Q.38 min.50 min. 1.50 min. Circular glass microfiber filters of 3.10 min–2.5 mg mL−1 were purchased from National Standard Material Center (Beijing.2. The solvent selected was ethanol–water–ammonia (75:24:1. v/v/v).. ethanol containing 1% (v/v) ammonia.1. and ultrasonic extraction were explored as reagent or extraction method.10 min. USA) was used throughout the study. 508 nm. California. Experimental 2. Samples The meat products were purchased from local commercial and farmers’ markets. Sub-samples of approximately 5 g were spiked with the right amount of the mixed standard solutions. 509 nm. unless otherwise stated. The ASE experimental parameters (i. 1. Liao et al.. and 2.00 min–2. China). Deionized water from Millipore (Milford. 2. soft drinks. Unfortunately. 2. ASE extraction Samples were extracted with a Dionex accelerated solvent extractor 350 (Dionex. Sunnyvale.3. Working standards of individual colorants were prepared by dilution of the stock solutions with deionized water. 2. Ponceau 4R (E124).50 min–2. MA. Tartrazine (E102).70 min. ASE may be a potential chance to reduce the adsorption. 2. Germany).00 min. These methods do not apply to pretreatment of samples with high protein content due to strong adsorption between the colorant and protein [29]. 529 nm. Sunset Yellow FCF (E110). mentioned methods have been developed for the determination of synthetic colorants in specific and simple foodstuffs (i. Brilliant Blue FCF (E133). After being homogenized in a high-speed food blender. Ethanol. Thus.25–27]. Representative UHPLC chromatograms of grilled chicken wings spiked at 0. solid juice powders. time of solvent .38 min–1. extraction solvent.5 mg kg−1 (A) and ham sausage (B) of each colorant extracted by ASE.

02 mol L−1 ammonium acetate and 5. Intra-day and inter-day reproducibility of the method were assessed by performing replicate analyses. temperature. 5. and static time were investigated to acquire the most effective conditions.02 mol L−1 ammonium acetate or 0.5.8 min.5 min. which then linearly increased as follows: up to 30% B at 1. Ammonia-containing organic solvents. the excellent separation for all of compounds explored could be achieved with acetonitrile containing 0.1 min. temperature.5 min. and holding at 98% B until 3. a degasser.5.0 ␮L.7 ␮m. Identification was performed by matching the retention time and absorption spectra of standards. cycles. E127. 3. When a certain percentage of water was added to the ethanol containing 1% (v/v) ammonia.130 Q. the recovery significantly improved. recovery experiments were performed. and holding at 30% B until 1. the initial mobile phase was 95% A with 5% B. The UHPLC system consisted of a binary pump.02 mol L−1 ammonium acetate and 0. Intra-day precision was determined by analyzing six fortified samples on the same day. 2. 2A). In gradient-elution analysis. Effects of solvent proportion (A) and extraction temperature (B) on recovery [constant pressure (1500 psi) and static time (10 min) were maintained]. 2. acetonitrile containing 1% (v/v) ammonia. Validation procedure To check the sensitivity of the proposed method. 2.1 mm × 100 mm) column. The system was controlled by an Empower ChemStation. can be used to extract colorants with low extract recoveries. each resulting extract was evaporated to dryness under vacuum distillation at 75 ◦ C. 0.0 mg kg−1 . respectively. 1500 psi. respectively. Absolute recoveries of the extraction were measured by spiking different concentrations of the studied colorants to the blank samples. E124. time purging with nitrogen to expulse rest of solvent in the cell. Chromatographic analysis Samples were analyzed by UHPLC. a column oven and a photodiode array (PDA) detector (Waters.3 min. The injection volume was 5.9 min. and then transferred to a 50 mL centrifuge tube. The flow rate was set at 0.1. The matrix-match calibration curve of each colorant was used for the validation experiments and quantification. E123.1% acetic acid as the mobile phase. USA).000 rpm for 5 min. In order to evaluate the trueness of the method. pressure. 2.0.0 mL of 0.02 mol L−1 ammonium acetate and 0.1. and Fig. or disappeared with either methanol or aqueous phase containing 0. The best ratio of ethanol–water–ammonia was 75:24:1(v/v/v) (Fig. 3. E129 were broadened.G. as described above. 10 min (static time). Chromatographic conditions (mobile phase composition and flow-rate) were evaluated and optimized in an acquity UPLC® BEH C18 (1. The residue was redissolved with 5. After extraction. Rapid analysis was achieved within 3.8 min. The spiked samples were then extracted and quantified. Solvent A was 0. / Analytica Chimica Acta 716 (2012) 128–132 in contact with the sample. Optimization ASE procedure The influences of solvent proportion. The PDA detector was programmed to monitor the colorants at a range of 220–700 nm. Inter-day repeatability was determined by analyzing the fortified sample on six different days.2.02 mol L−1 ammonium acetate as the mobile phase system. an automated injector. the lower solutions were used for UHPLC analysis. The presence of 1% (v/v) ammonia could . 0. Under the optimal gradient elution conditions. whereas the column temperature was maintained at 30 ◦ C. After liquid–liquid extraction and high-speed centrifugation at 16. whereas solvent B was acetonitrile. the extraction UHPLC chromatograms of grilled chicken wings spiked at 0. Optimum separation conditions of synthetic dyes by UHPLC The separation behaviors using the following mobile phases were investigated: aqueous phases containing 0. 60%. 1 A and B respectively). The separation was also investigated under different gradient elution conditions. A return to the initial conditions was carried out at 3. organic phases containing acetonitrile or methanol. Therefore.02 mol L−1 of ammonium acetate aqueous solution. flushing volume respect to the cell size in percentage.05. 1. Results and discussion 3. static cycles. and up to 98% B at 2.1% acetic acid.5 mg kg−1 of each colorant and ham sausage showed that seven synthetic colorants properly separated with sharp symmetrical peak shapes (Fig. up to 70% B at 2. E110. calibration curves were also prepared with the mixed standards solutions spiking blank samples at concentration levels of 0.4. 2. The determination of each substance was conducted at the appropriate absorbance wavelengths. partially overlapped.2 min and held at 5% B until 3. it was adopted as the mobile phase in subsequent studies. E133. 85 ◦ C. such as ethanol containing 1% (v/v) ammonia. The limit of detection (LOD) and limit of quantification (LOQ) were determined by considering 3 and 10 times the standard deviation of the signals from the blank matrix.0 mL hexane. 120 s. However.3 mL min−1 . Liao et al.0 and 10. The peaks of E102.

9 2.9 4.0 4.879 3.4 83. The maximum absorption wavelengths were monitored as detection wavelengths. 10.3 84.5 2.05 0.6 80. 12.0 E124 0.2 2.96 × 103 + (4.1 83.5 4.3 84.4 79.6 3. Matrix-matched calibration curves were prepared daily by spiking blank control samples of meat products Table 2 Recoveries and precision of each colorant spiked in the grilled chicken wings. Flush percentage refers to the amount of solvent flushed through the cell following the static heating step.6 5.9 5.0 78.63 × 103 + (1.227 2.7 2. indicating the acceleration of protein denaturation. 100%) did not significantly affect the extraction efficiencies of the analytes. The number of extraction cycle was 5 LOD (mg kg−1 ) LOQ (mg kg−1 ) R2 0.9 83.02 0.7 83. Liao et al.2 3. Increasing the static extraction time from 5 min to 10 min resulted in a recovery improvement at about 10%. static time of 10 min was selected in order to minimize the time of analysis.5 2.0 77.6 3.Q.953 2.0 5.3 84.62 × 105 )x y = 4.2 3.9999 0.9 3.6 5. However.3 82.0 5. linear range. To evaluate if extraction time could influence extraction efficiency.6 4.5 2.3 78.5 3. Due to increased viscosity by water.3 77. which probably led to the lower recoveries [32].5 2.0 5.05–10. 7.5 83.0 0. 3.9 77.05–10. Therefore. and causing protein denaturation. and then slightly decreased at higher temperatures (Fig.05 0.0 0.8 5.0 5.9 2.3 84. Consequently.9994 0.5 4.032 2.9 3.2 E110 0.0 83. Colorant Spiked level (mg kg−1 ) Intra-day recovery (n = 6) Inter-day recovery (n = 6) Mean (%) R.05 0. (%) Mean (%) R. The number of the extraction cycles was tested to assure a rapid extraction.0 0.0 78.5 84.036  (nm) 429 522 509 483 508 625 529 Linear range (mg kg−1 ) Linear equation 0.01 0.3 4.0 83.3 84. Colorant E102 E123 E124 E110 E129 E133 E127 tR (min) 1.5 2.5 2.3.9998 0. Increasing the flush volume allowed more solvent to pass through the sample.02 0.7 4.0 5.4 82. After two cycles of extraction of the same matrix.7 83. as well as high recovery.01 × 102 + (1.3 83.0 79.4 83.85 × 105 )x y = −1.5 79.5 82.05–10.5 84.0 0.29 × 103 + (1.1 3.9 5.5 81.8 4.D.8 4. 15 min) were used.25 × 105 )x y = 1.5 83.02 0.21 × 105 )x y = 1. retention time.05 0. a percentage higher than 25% (v/v) of water in the extraction solvent also resulted in lower recoveries.05 0.05 0.9 5.1 E129 0.5 3.8 4. / Analytica Chimica Acta 716 (2012) 128–132 131 Table 1 The detection wavelength. 60%.7 80.1 E127 0.3 77.6 2. linear equation.9992 varied between one and three.3 84.7 3.5 4.05 0.0 0.05–10.5 2.7 . different extraction times (5.0 y = 6.3 80. but also increased the final volume of the extract.313 1.05 0. 85 ◦ C was adopted as the extraction temperature.9998 0.4 81.8 5.D.94 × 10 + (1.0 76.05 0.2 2.50 × 10 )x y = 6.87 × 105 )x y = 5.5 1.3 4.3 2.7 84.6 763 76.5 3.0 80.3 4.4 81. expressed as a percentage of the cell volume.1 79. for the colorants and protein are negatively charged in the alkaline medium.33 × 103 + (2.5 2.9997 0. Different flush volumes (40%.7 4.0 5. 2B). A small amount of aggregation from meat products was observed when heated at higher than 85 ◦ C.02 0. the flush volume was set at their default values (60%). On the other hand.3 4. the recovery could no longer increase.S.3 80.9 3.7 E133 0.9 81.9 79.2 2.5 80. On the other hand. Validation of the method Each colorant has a different wavelength of maximum absorption.2 3. 80%.0 E123 0. The remaining solvent could contain colorants. Static cycles introduce fresh solvent during the extraction process and are useful for matrices that are difficult to penetrate. Consequently.05 0.01 0. Recovery of the colorants increased with increasing extraction temperature from 55 ◦ C to 85 ◦ C. the flushing and purging steps did not completely expel the extraction solvent from the extraction cell.9 4.3 84.05 0.5 3.05–10.9 81. limit of detection (LOD) and coefficients of determination (R2 ) of all colorants.4 82.5 2.5 82.458 1.9998 0.S.3 4.05 0.18 × 103 + (1.05 0. and the aggregation could wrap or adsorb on the synthetic colorants.5 2.10 × 105 )x 2 reduce the electrostatic adsorption between the colorants and protein.3 6.0 80.3 82.05–10.G.9 3. the presence of ethanol could reduce hydrophobic force between the colorants and protein.4 2.01 0.0 4.5 84.05–10.5 3. (%) E102 0.0 4.0 5. the extraction efficiencies were affected by extraction temperature.05 0.0 80.0 0.

Schenone.D. Gosetti. C. S. Krause. Chromatographia 57 (S) (2003) 115. S. Gao.Y. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] 4. Chromatogr. Baydan. Chromatogr. Huang. Bai. LC–GC N.2% (E124) to 6. coefficients of determination (R2 ).F.V.N.P. E.W. Chiu. Chen. For ASE extraction step.3% (E127).G. N. B. LeBlanc.O. F. 28 (2005) 923.D.E. Garcia-Falcon.B. Pfalzgraf.9% (E102) and 76. M. Di Nezio. S.R. Lett.E. Trenerry.T.C. H. Durbeck. Food Chem.L. J.F. Fernández Band.99. Psych. D.0 mg kg−1 were shown in Table 2. Anal. C. A 1103 (2006) 170. The results of precision indicated that intra-day and inter-day repeatability (R. Application to real samples The proposed method was applied to meat products from local commercial and farmers’ market. Gallego. Talanta 46 (1998) 933. González. 5 (1976) 129.J. Conclusions In the current study. Polati. Dagnac. T. González. C. A. K. Ichihashi. F. C. D. S.C.M. Hou. Bronze. A. M. Mouvet. Dinc.C. To reduce matrix effects. J. A 957 (2002) 89. only a few samples tested positive for E102.S. X. Y. Zhang. Chromatogr. A 1054 (2004) 379.S. Marengo. Liao et al.V.P. J. F.A. respectively. M. A 1136 (2006) 231. 40 (2007) 3080. Chromatogr. N. Q. In the eighty samples. M. Chromatogr. Food Sci. Wang. M.Y. Zhang.7% (E123) to 5. their concentrations were found to be in the range of 2.C. L.C. Chem. Chen. Gennaro. E123. Raffi.2% (E124) and 3.P. J. A.0% (E129). detection wavelength.M. Toxicol. L. C. Gong.J. H. Ni. Onur.01–0. J. J. Shih. J. matrix-match calibrations curves were chosen as reference curves throughout the study. J. S. Chromatogr. Brockner. Acta 583 (2007) 103. Y. S. limit of detection and limit of quantification of all colorants are presented in Table 1. Food Chem. Calibration equations of mixed standard solutions. A 995 (2003) 29. S. Brum.C. O. S. Acta 354 (1997) 163.3% (E124) to 84. A 775 (1997) 187. Chia. J. Cheng. Sue.4. H. Ni. Vittori. M.E. Chim. S.F. Andrade.M. J. Am. Chen. the meat products.F. L. J. The results from recovery experiments in the same matrices at levels ranging from 0. Bononi. Jorissen. linear range. Orthomol. Riviello. K. Correlation coefficients (R2 ) were better than 0. C. Huang. ethanol–water–ammonia with a ratio of 75:24:1 (v/v/v) and extraction temperature of 85 ◦ C were found to be the most effective.J.5–10. É. J. Chromatogr. The results of recovery indicated that the intra-day and inter-day recoveries of synthetic colorants were between 76. B. A 827 (1998) 73. Berzas Nevado. 2010CJJ012). J.J. and calculated from the measurement of six standard solutions (concentration levels of 0. Netto.132 Q. S. M. Method 3 (2010) 188. M. V. 26 (2008) 567. Talanta 56 (2002) 663. Kirschbaurn. Yoshioka. J. J.S. M. A 1216 (2009) 7063. Chromatogr. Y. Kanbur. M. 75 (2003) 685. C. Rel. Mou. Agric. M. N. H. Chromatogr. Talanta 58 (2002) 579. X. M. R. F. Fuh. M. Y. Calibration equations were calculated using the peak area of the colorants. Gallone. respectively. 35 (1997) 897. Anal. Culzoni. Acta 329 (1996) 65.C. 107 (2008) 489. J. E. F. Precision and recovery were investigated in order to research the influence of the matrix. Anal. Detection and quantification limits of synthetic colorants were in the ranges of 0.C. J. E. Acknowledgements This work was supported by the fund from nonprofit industryspecific agricultural research projects (no. .S. Bristeau.9 mg kg−1 . Valcárcel. of synthetic colorants were between 1.C. Jin. K. Tateo. J. R. Combeau. Valcárcel. A 1067 (2005) 225. Goicoechea. respectively.9% (E124) to 84. 3. R. an efficient and accurate analytical method for the simultaneous determination of seven synthetic colorants in a single run by UHPLC-PDA was developed and optimized. Chromatogr.R. M.W. M. Bushway. Kreisselmeier. C. H. Zhang. Chromatogr. J. Hall. A 704 (1995) 195. E. Godoy.A. M. Technol. M.S. Prado. Thomaidis. J. The applicability of the proposed method was verified by the determination of colorants in [24] [25] [26] [27] [28] [29] [30] [31] [32] F. Chromatogr.0 mg kg−1 ). Minioti. S. M. Food Anal. Rodriguez Flores.E. H. Ni. Food Control 16 (2005) 293. A 1164 (2007) 120. Yao. J.B. N.W. Talanta 74 (2008) 1408.J. 28 (2010) 427. M. D. Ai. The method permits the detection of colorants at very low concentrations (mg kg−1 range). V. A. Luo. Alves. Chromatogr. and/or E124. Buckley. Llamas. The results demonstrate that the precision and accuracy of the present method are acceptable for routine monitoring purposes. Talanta 56 (2002) 115. Sakellariou.D. 200903055) and innovation fund of Jiangxi Academy of Agricultural Sciences (no. Chatelut. Gallego.05 mg kg−1 . / Analytica Chimica Acta 716 (2012) 128–132 with mixtures colorants. H. J. L.05 mg kg−1 to 5. V. Marengo. J. N. Z. Angioi. Gosetti. L.S. Guiberteau Cabanillas. Perkins. Y. The developed method could also be potentially applied to other products with high protein content. X. Villasenor Llerena. Czech. Gianotti. J. S. Contento Salcedo.05–10. Ma. Food Chem. Anal. Su. Cerniglia. Chim. Thompson. Simal-Gandara. 51 (2003) 2121.Garrido. Baran. The results also indicate that the method is suitable for determining low amounts of synthetic colorants in meat Angioi. A 959 (2002) 317. M. Gianotti. Anal. F. Gennaro. B. Boas. Y. Sun.02 mg kg−1 and 0. Kim.L. Hawley. Jeannot. Chim. Sun.