Accepted Manuscript Analytical Methods Fast and simultaneous determination of eleven synthetic color additives in flour and meat products by liquid chromatography coupled with diode-array detector and tandem mass spectrometry Ping Qi, Zhi-hao Lin, Gui-yun Chen, Jian Xiao, Zhi-an Liang, Li-ni Luo, Jun Zhou, Xue-wu Zhang PII: DOI: Reference:
S0308-8146(15)00265-4 http://dx.doi.org/10.1016/j.foodchem.2015.02.075 FOCH 17174
To appear in:
Food Chemistry
Received Date: Revised Date: Accepted Date:
29 July 2014 12 January 2015 14 February 2015
Please cite this article as: Qi, P., Lin, Z-h., Chen, G-y., Xiao, J., Liang, Z-a., Luo, L-n., Zhou, J., Zhang, X-w., Fast and simultaneous determination of eleven synthetic color additives in flour and meat products by liquid chromatography coupled with diode-array detector and tandem mass spectrometry, Food Chemistry (2015), doi: http://dx.doi.org/10.1016/j.foodchem.2015.02.075
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.
1
Fast and simultaneous determination of eleven synthetic
2
color additives in flour and meat products by liquid
3
chromatography coupled with diode-array detector and
4
tandem mass spectrometry
5 6 7
Ping Qi1, 2, Zhi-hao Lin2 , Gui-yun Chen2, Jian Xiao2 , Zhi-an Liang2 , Li-ni Luo2, Jun Zhou2 ,
8
Xue-wu Zhang1*
9
1College of Light Industry and Food Sciences, South China University of Technology,
10
Guangzhou, China
11
2GuangZhou Institute for Food Control, Guangzhou, China
12 13 14 15
Running title: Fast and simultaneous determination of eleven synthetic color additives
16 17 18 19 20
*To whom correspondence should be addressed: Dr XW Zhang, College of Light Industry and
21
Food Sciences, South China University of Technology, 381 Wushan Road, Guangzhou 510640,
22
China. Tel: 86 20 87110840; Fax: 86 20 87110840; E-mail:
[email protected].
23 24 25 26
27
Abstract:
28
In this study, an efficient, fast and sensitive method for the simultaneous determination of eleven
29
synthetic color additives (Allura red, Amaranth, Azo rubin, Brilliant blue, Erythrosine, Indigotine,
30
Ponceau 4R, New red, Sunset yellow, Quinoline yellow and Tartrazine) in flour and meat
31
foodstuffs is developed and validated using HPLC coupled with DAD and MS/MS. The color
32
additives were extracted with ammonia-methanol and was further purified with SPE procedure
33
using Strata-AW column in order to reduce matrix interference. This HPLC-DAD method is
34
intended for a comprehensive survey of color additives in foods. HPLC-MS/MS method was used
35
as the further confirmation and identification. Validation data showed the good recoveries in the
36
range of 75.2-113.8%, with relative standard deviations less than 15%. These methods are suitable
37
for the routine monitoring analysis of eleven synthetic color additives due to its sensitivity,
38
reasonable time and cost.
39
Keyword: color additives; HPLC-DAD; HPLC-MS/MS; food safety
40 41
1 Introduction
42
Color additives have been widely used as coloring agents in the food industry for many years.
43
They are usually classified as natural (or identical natural) and synthetic. Natural color additives
44
generally have a lower tinctorial strength than synthetic color additives, which are generally more
45
sensitive to light, temperature, pH, and redox agents. At present, it is more frequent that single or
46
mixtures of several synthetic color additives are used as food colorants in foodstuff to obtain
47
attractive colors of a product. However, the use of these synthetic color additives must be
48
permitted and controlled because they can occasionally produce allergy, asthma and other health
49
disorders in sensitized individuals (Boeniger, 1980; Amate, et al., 2010).
50
The Food Safety Law of the People’s Republic of China requires the application of synthetic color
51
additives to be kept under surveillance by the China Food and Drug Administration (CFDA) and
52
listed in Direct GB 2760-2011 of the Ministry of Health, in order to be legally used in food
53
markets in China. According to the Direct GB 2760-2011, eleven synthetic color additives are
54
listed as certifiable food color additives that can be added to food products. Permitted synthetic
55
food color additives are: Allura red, Amaranth, Azo rubin, Brilliant blue, Erythrosine, Indigotine,
56
Ponceau 4R, New red, Sunset yellow, Quinoline yellow and Tartrazine. Based on their chemical
57
structure, they can be divided into the azo(sunset yellow), triarylmethane (brilliant blue), xanthene
58
(erythrosine), and indigo (indigotine) colorant classes. These synthetic food color additives are
59
usually used as the water-soluble sodium salts. Their name, structure and properties were
60
summarized in Table S1. Moreover, the Direct GB 2760-2011 also regulates the fields of
61
application of the synthetic food color additives and the permitted maximum quantities allowed
62
for coloring foodstuffs. In China, the maximum amount allowed for most synthetic food color
63
additives is no more than 100mg/kg. Even it is non-permitted that these synthetic color additives
64
are used in several kinds of foods, such as stewed meat, roast meat and stream born products.
65
When they are consumed in excessive amounts, these substances and their metabolites also pose
66
potential health risk to human beings and may even be carcinogenic (Robens et al., 1980; Price, et
67
al., 1978). Therefore, CFDA usually makes the plan to monitor and investigate the levels of the
68
certified synthetic color additives in high consumption and risk products such as meat and flour
69
products every year. In response to the CFDA’s plan, a new, fast, accurate and robust method for
70
the quantitative determination of the certified synthetic color additives should be developed in
71
food products, particularly complex solid-matrix foods.
72
Based on HPLC, several types of methods have been reported for the determination of 3-40
73
colorants in food products (Alves et al., 2008; Dossi et al., 2006; Garcia-Falcón & Simal-Gandara
74
2005; Yuet-Wan Lok et al., 2010; Ma et al., 2006; Minioti et al., 2007; Yoshioka et al., 2008).
75
However, most of them only focused on liquid samples, like soft drinks and juice drinks, or
76
water-soluble foods such as fruit jelly, jam and confectionery because the colorants can be
77
analyzed directly with little sample preparation. Obviously, these methods are not suitable for the
78
complex solid-matrix foods. The other methods (Tavakoli et al., 2014; Tao et al., 2011; González
79
et al., 2003; Sun et al., 2013) reported for more complex foods use procedures for the
80
determination of the color additives that are not suitable for CFDA’s use, because they are either
81
very time-consuming or only applicable to the chili foods (chili powder, chili paste) or they
82
require the use of special cleanup materials and instruments. In addition, most researchers pay
83
much more attentions to the determination of illegal dyes (Zou et al., 2013; Zhu et al., 2014;
84
Chang et al., 2011; Enríquez-Gabeiras et al., 2012; Alesso et al., 2012), such as Sudan dyes,
85
Rhodamine B, Para red in foods. Not many researches were reported for the detection of permitted
86
synthetic color additives. However, the usage of permitted synthetic color additives sometimes
87
was above the authorized levels or beyond the scope of application in foods. Until now, to the best
88
of our knowledge, there are no reports in detail on the simultaneous determination of all the
89
certified synthetic color additives in solid-matrix foods, especially in animal origin foods. Animal
90
origin foods have very complex matrices. They typically contain high concentrations of fats,
91
proteins and other additives, which often caused the interference in the confirmation of color
92
additives.
93
Thus, in our study, a new method was developed and validated for the simultaneous determination
94
of eleven permitted synthetic color additives in high protein and fat content food products. Such
95
method will be employable for routine applications where high sample throughput is required
96
without affecting the accurateness and the sensitivity of the determinations. The analysis was
97
mainly performed with high performance liquid chromatography (HPLC), coupled diode array
98
detector (DAD) or tandem mass sepctrometery operated in negative electro-spray mode
99
(ESI-MS/MS). This HPLC-DAD method was intended for a comprehensive survey of color
100
additives in foods. However, it was not sufficient for the identification by the retention time and
101
spectrum because of the interference of food matrix.
102
chromatography−tandem mass spectrometry (HPLC-MS/MS) can distinguish and identify targets
103
from background matrix ions, which can increase sensitivity and specificity . Therefore,
104
HPLC-ESI-MS/MS method was chosen and developed for the further confirmation purposes to
105
assure accuracy of the results. The influences of extract preparation condition, mobile phase, SPE
106
condition, and MS parameters were investigated and optimized. The ionization behavior and the
107
MS/MS fragmentation behavior of dyes were researched. The proposed method can realize fast
108
separation of the 11 color additives in a 10-min gradient elution. The method was validated by
109
evaluating recovery, selectivity, linearity, accuracy and repeatability according to the China FDA
110
guideline GB/T 27404.
111
2 Materials and Methods
112
2.1 Reagents and materials
113
Certificated reference materials of Azo rubine, New red, Erythrosine and Quinoline yellow were
114
obtained from Dr. Ehrenstorfer GmbH (Augsburg, Germany). Appropriate amounts of powder of
The high-performance liquid
115
these color additives were dissolved in methanol/water (1:1 v/v) to give a concentration of 1.00
116
mg/mL. The other color additives (1.00 mg/mL) used as standards were purchased from Chemical
117
Metrology & Analytical Science Division (Beijing, China). Matrix-matched mixed working
118
standard solutions were prepared by adding desired volume of individual stock standard solutions
119
into the blank matrix. These solutions were stored at 4 °C in the dark. All working solutions for
120
the calibration were prepared fresh before use.
121
All the water used was purified by Sartorius Arium 611 system with a resistance of 18.2 MΩ/cm.
122
HPLC grade methanol and ammonium acetate were purchased from Sigma-Aldrich (St. Louis,
123
MO, USA). Analytic grade ammonium hydroxide, ethanol and n-hexane were purchased from
124
Guangzhou Chemical Reagent Company (Guangzhou, China). SPE Strata-X-AW, Strata-X-C,
125
Strata-X and Strata-X-CW cartridges (200 mg, 6 mL) were obtained from Phenomenex (Torrance,
126
CA, USA), which were used in the purification step. Teflon membrane syringe filters (0.22µm)
127
were bought from Anpel Company (Shanghai, China).
128
2.2 Sample collection
129
All of the food samples such as corn steamed bun, barbecued pork and roasted duck were
130
purchased from local markets. The manufacturer or distributor declared that these products didn’t
131
contain any synthetic color additives. Prior to analysis, the products were mixed homogeneously
132
and stored in 50 mL PTFE centrifuge tubes at -20 °C. Spiked samples were prepared in a 50 mL
133
centrifuge tube by mixing 2.0 g of homogenized samples with a series of the 11 color additives
134
standard solutions at various concentrations.
135
2.3 Sample preparation
136
First, 2.0 g of homogenized sample and 10 mL of n-hexane were shaken by vortex mixer for 5 min,
137
and then the n-hexane layer was discarded to eliminate fat. Second, the sample was extracted with
138
10 mL of methanol-ammonia-water (80:2:18, V/V/V) for 10 min in an ultrasonic bath at 40 °C.
139
The super-extracts were collected. The above procedure was repeated 1 more time. Finally, the
140
pooled super-extracts were collected and evaporated to dryness by rotary evaporator at 35 °C. The
141
evaporation residues of pooled extracts were redissolved in 20 mL of deionized water as the
142
loading solution of SPE. Third, 10 mL of redissolved solution was loaded onto the Strata-X-AW
143
cartridge that was preconditioned with 6 mL of methanol and 6 mL of water. After washing with 6
144
mL of water/methanol (1:1, v/v), the retained constituents were eluted with 20 mL of ethanol that
145
contained 10% (v/v) ammonia−water, followed by the evaporation to dryness by rotary evaporator.
146
Finally, the evaporated residue was reconstitute in 5 mL of methanol−water (1:9, v/v), and then it
147
was filtered through a 0.22µm Teflon syringe filter for HPLC or HPLC-MS/MS analysis.
148
2.4 Apparatus
149
The HPLC-DAD method was developed using an Agilent 1260 HPLC system with binary-pump,
150
auto-sampler, temperature controlled column oven and DAD detector (Agilent Technologies,
151
Fermont, CA). HPLC-MS/MS analysis was performed on a Shimadzu Prominence Ultra Fast LC
152
(UFLC) system coupled with an API 3200 triple quadrupole mass spectrometer (AB SCIEX
153
Company, Canada).
154
2.5 HPLC conditions
155
Separation was carried out by Agilent XDB-C18 column (4.6×150mm, 5µm) with a C18 guard
156
column (4.6×12.5mm, 5µm) and a binary gradient which included 20mM ammonium acetate
157
buffer at pH 7 (mobile phase A) and methanol (mobile phase B). All separations were performed
158
at 40 °C temperature and flow rate was set at 1.0 mL/min. Running time was 10 min. The
159
optimized gradient for color additives separation was given as follows: 8% of B increased to 15%
160
in 1.0 min, 15-22% B at 1.0-1.8min, 22-34% B at 1.8-4.8min, 34-100% B at 4.8-7.0min, and kept
161
at 100% until to 8 min and then returned to the original proportion within 1 min. At last, another 1
162
min was needed to ensure the stability of baseline for the next injection. Injection volume was
163
20µl. The peaks of 11 color additives were individually measured at three wavelengths: 430 nm
164
for the yellows, 510 nm for the reds and 630 nm for the blues. Diode-array detector was
165
programmed to monitor the colorants from 240 to 800 nm. The color additives were identified and
166
quantified in the sample solutions by comparing their LC retention times and DAD absorption
167
spectra with those of the standards. Quantification was based on the external standard method. The
168
best fit standard curve was prepared by linear regression of peak areas versus concentration.
169
Typical HPLC chromatograms of the mixed standard solution at the three wavelengths are shown
170
in Figure 2b.
171
2.6 HPLC-MS/MS conditions
172
Shimadzu Inertsil ODS-3 C18 column (3.0×75mm, 2.2µm) was used for LC-MS/MS confirmative
173
analysis. The column temperature was 40℃. The flow rate was set at 0.3mL/min. The mobile
174
phase consisted of solutions A (methanol) and B (5mM ammonium acetate solution). The gradient
175
elution program was a little different from HPLC-DAD experiments. In gradient-elution analysis,
176
the initial mobile phase was 15% of solvent A, increased linearly to 95% in 7 min, and held at
177
95% for 1 min. A return to the initial conditions was carried out in 1 min. The triple quadrupole
178
tandem mass spectrometer operated under multiple reaction monitor mode (MRM) for quantitative
179
and qualitative analysis. Negatively charged ion species from the 11 color additives were
180
monitored. The MS/MS operation parameters of the analytes were optimized by introducing the
181
single standard solution of color additives. The optimized electrospray ionization conditions were:
182
gas temperature 650℃, capillary voltage -4500V; Curtain gas (N2) 20psig; Nebulizer gas (N2 )
183
55psi; and Auxiliary gas (N2) 50 psi. Detailed MRM settings were listed in Table S2. Applied
184
Biosystems/MDS Sciex Analyst software (versions 1.5.1) was used for data acquisition and
185
processing.
186
3. Results and discussion
187
3.1 Optimization of Extraction Procedures.
188
The complex matrices of the samples pose challenges for the sample extraction procedure. As
189
color additives mainly bond with celluloses, fibers, proteins, the development of an efficient
190
sample extraction procedure is critical for accurate determination of color additives in real sample
191
analysis. In this work, ultrasonic extraction was chosen as sample extraction method. For
192
eliminating the interference of fat, the high fat samples, such as roast pork, was mixed by the
193
solvent of n-hexane firstly and the n-hexane layer was discarded. After that, five extract solvents
194
were compared and evaluated in this experiment based on chemical properties of the 11 color
195
additives, including methanol-ammonia-water, 80:0:20 (V/V/V), 80:2:18 (V/V/V), 80:4:16
196
(V/V/V), 80:6:14 (V/V/V) and 80:8:12(V/V/V). The results (Figure 1) indicated that basic
197
conditions facilitated the release of the color additive from the food matrix, particularly for
198
tartrazine, new red and amaranth. All the extract solvents, except for the methanol-ammonia-water
199
(80:0:20, V/V/V), showed the similar extraction efficiency in terms of recovery and
200
reproducibility for the 11 color additives. So methanol-ammonia-water 80:2:18 (V/V/V) was
201
chosen as the finally extract solvent because the least amount of ammonia was used in the
202
solution.
203
3.2 Optimization of the SPE Procedure.
204
In this experiment, further purification of the extract solution with solid-phase extraction was also
205
studied. As we know, the efficiency of SPE depends on the types of sorbent. Four types of
206
Phenomenex cartridges (Strata-X-AW, Strata-X, Strata-X-C, Strata-X-CW) were investigated.
207
The effectiveness of SPE cartridges were evaluated initially to extract 11 color additives from the
208
spiked deionized water samples (spiked final concentration was 10mg/L). The results were
209
showed in Figure 2b. Strata-X-AW SPE columns showed the highest recovery: above 95% mean
210
recovery for all the analytes. This can be explained by the chemical structure of the 11 color
211
additives. All of them belong to the acid compounds, which contain one or several sulfonic or
212
acetate groups. Strata-X-AW SPE columns possess weak anion exchange forces that allows for
213
complete retention of acidic compounds. Therefore, X-AW SPE columns showed the best
214
recovery, and were chosen for the further purification.
215
3.3 HPLC-DAD method development and validation
216
3.3.1 Optimization of HPLC Condition
217
To establish the best HPLC conditions of the 11 color additives, the composition of the mobile
218
phase, pH, gradient elution program and detection wavelength on the response and separation of
219
the 11 color additives were studied. Because the HPLC-DAD method was intended for a
220
comprehensive survey of permitted color additives in foods where several hundred food products
221
will be analyzed for routine applications, the inexpensive mobile phase components will
222
significantly reduce the cost of the survey. So methanol rather than acetonitrile was chosen as
223
solvent B. Meanwhile, ammonium acetate aqueous solution was used for a solvent A, which was
224
compatible with LC/MS for the further confirmation analysis. On the other hand, the pH of the
225
mobile phase is an important parameter to be optimized as it has a significant effect on the peak
226
shape and separation of analytes. According to the report of Bonan (Bonan et al., 2013), all the
227
colorants are neutral species at pH 7. Thus, the pH of ammonium acetate aqueous solution
228
(solvent A) was adjusted to 7 in order to enhance the formation of neutral species from the ionized
229
analytes. Finally, methanol and ammonium acetate aqueous solution (0.02M and pH 7) were
230
chosen as mobile phase.
231
To baseline separate the 11 colorants, several kinds of the gradient program were studied. The best
232
gradient was described in 2.5 HPLC conditions. We chose three wavelengths for the identification
233
of the colorant additives: 430nm for the yellows, 510nm for the reds and 630nm for the blues.
234
Although those wavelengths are not the maximum absorption wavelengths for the individual color
235
additives, they were optimal for detecting all of the colorant additives in one analysis. Figure 2b
236
shows that the 11 color additives were completely separated within 8 min and do not interfere with
237
one another. The color additives were eluted from the column according to increasing polarity,
238
depending on the number of polar functional groups such as hydroxyl groups and sulfonate groups.
239
Azo groups (tartrazine, new red) generally tended to elute earlier than triarylmethane (brilliant
240
blue FCF) and xanthene (erythrosine) colorant classes. The first color additives eluted was the
241
more polar tartrazine , while the last was erythrosine which presents no sulfonated group and four
242
iodine substituents. Quinoline yellow consists essentially of sodium salts of a mixture of
243
disulfonates and monosulfonates (Kirschbaum et al., 2006). It showed two peaks corresponding to
244
relative isomers. The sum of the two peaks was used for determination of recoveries and precision
245
validation. The results indicated that this analytical method was more efficient than other methods
246
(Kirschbaum et al., 2006; Zou et al., 2013; Bonan et al., 2013; Petigara Harp et al., 2013) in terms
247
of analysis time and number of detected analytes.
248
3.3.2 Validation of the HPLC-DAD method
249
Linearity, limit of detection (LOD), limit of quantification (LOQ), precision and recovery were
250
determined to evaluate the validity of the HPLC method. Table S3 lists the results. Linearity was
251
studied by analyzing mixed standard working solutions of the 11 color additives at several
252
concentrations ranging from 0.1 to 10 mg/L in HPLC. All of the color additives showed
253
satisfactory linearity. Correlation coefficients (R) were 0.999 or higher for all color additives. The
254
LOD and LOQ, which were respectively defined as the minimum concentration based on 3 and 10
255
times the standard deviation of the signals from the negative blank matrix, were estimated by
256
analyzing 10 negative blank samples. Depending on the color additives involved, LODs and
257
LOQs were in the range of 0.007-0.096 and 0.023–0.32 mg/kg, respectively.
258
The color additives were spiked into corn steamed bun, roasted pork and roasted duck at three
259
levels (1, 5, 10mg/kg). Two replicates were tested for each concentration. The recoveries, spiked
260
levels and relatively standard deviations (RSD) were summarized in Table S4. The results show
261
that the recoveries varied from 85.2 to 108.3%, relative standard deviations (RSD) varied from 0.3
262
to 7.6%. The validation data demonstrated that the present HPLC method had a good overall
263
recovery, an excellent precision, and low LODs and LOQs, which was satisfied with the routine
264
monitoring purposes.
265
3.4 HPLC-MS/MS method development and validation
266
3.4.1 Optimization of MS/MS Parameters
267
The complexity of the food matrix may interfere with the exact measurements of the HPLC-DAD
268
method, so it is necessary that HPLC-MS/MS is used as a confirmative step to further identify the
269
existence of the color additives in the complex sample matrices. Negative ion mode is the
270
preferred mode when the compounds contain carboxyl, hydroxyl or sulfonate groups, due to their
271
deprotonated property. Therefore, the 11 color additives were analyzed in negative ion mode. The
272
MS/MS conditions were optimized individually for each subject by injecting 100 µg/mL standard
273
solution into MS/MS with a mobile phase of methanol and water (50:50, v/v, contain 5mmol/L
274
ammonium acetate) at a flow rate of 10 µ L/min. The precursor ions were found in scan mode
275
individually.
276
To achieve the highest selectivity and sensitivity, mass spectrometry parameters including
277
declustering potential voltage (DP), collision energy (CE), precursor ions and product ions were
278
optimized. The optimum MS parameters for each color additives are summarized in Table S2. The
279
results showed that the most abundant precursor ions of the colorants was closely related to the
280
number of sulfonate group in their molecular structure. Figure 3a-3d is the ESI(-) mass spectrum
281
of New red, Amaranth, Allura red and Azo rubine, respectively. As shown in Figure 3a-3d, the
282
most abundant precursor ions are their negatively two charged sodium adduct ions [M-2Na]2- or
283
[M-3Na+H]2- when the color additives contain two or three sulfonate group in their molecular
284
structure, such as New red and Allura red. This study indicated that the ionization behavior of
285
ESI-MS was determined by the most polar functional groups of molecules.
286
Furthermore, in order to obtain the best response, the concentration of ammonium acetate (5, 10,
287
20mM) in mobile phase B was optimized. The results were shown in Figure 4. Finally, 5mM
288
ammonium acetate was selected as aqueous solution in the mobile phase because the peak area,
289
peak shape and sensitivity were improved.
290
3.4.2 Validation of the HPLC-MS/MS method
291
The HPLC-MS/MS method has been validated including method selectivity, linearity, sensitivity,
292
recovery and relatively standard deviations (RSD). Selectivity was verified by comparing the TIC
293
of 1.0mg/L mixed colorant standards in pure solvents and in matrix. The results was shown in
294
Figure 5. No matrix interferences were observed in corresponding retention times of the target
295
compounds. Therefore, this method had high selectivity for all color additives.
296
The linearity of the method was studied by analyzing mixed standard working solutions of the 11
297
color additives at 5 levels ranging from 0.01 to 1.0 mg/L. All of the 11 color additives displayed
298
good linearity with correlation coefficients (R) exceeding 0.995 (Table S2). LODs and LOQs of
299
the 11 color additives were 0.0023-0.022 and 0.0076-0.071 mg/kg, respectively, which indicated
300
this method had a good sensitivity.
301
Recovery and precision were validated by the spiked blank flour and meat products at three levels
302
(1, 5, 10 mg/kg). The results was summarized in Table S4. It shows that the mean recoveries are in
303
the range of 75.2-113.8% at the three levels with the RSDs ranging from 2.3 to 15.1%.
304
Considering that these data were not corrected with the internal standard, the precision of
305
HPLC-MS/MS method was high enough for the confirmation and quantification of positive
306
samples
307
3.5 Application to real sample analysis
308
The validated method developed in this paper was applied to determine the color additives in real
309
samples. More than 100 samples of meat and flour products obtained from different local markets
310
have been analysed in routine work between May 2012 and November 2013. In order to assure the
311
quality of the results, a spiked blank sample (5mg/kg) was applied for every batch of samples.
312
Among them, only several positive meat samples were detected. The typical HPLC-DAD (430 nm)
313
and HPLC-MS/MS (MRM) chromatogram for one of the positive meat sample are shown in
314
Figure 6. It illustrates that the method developed in our study could be a suitable confirmatory
315
method.
316
4 Conclusion
317
In summary, the robust, fast, inexpensive and effective HPLC-DAD method with
318
ammonia-methanol extraction provided satisfactory sensitivity and precision for the simultaneous
319
determination of 11 color additives in complex solid-matrix foods, especially in animal origin
320
foods. Meanwhile, HPLC-MS/MS, which was developed and chosen for the further confirmation
321
purposes, showed good sensitivity and recovery under complex sample matrix situation. The SPE
322
clean-up could efficiently remove fats, proteins and natural pigments interferences from samples.
323
No interference peak was observed in the chromatograms. The lack of interference from matrix
324
effects demonstrated the broad applicability of the HPLC and HPLC-MS/MS methods. All of the
325
results showed that this method is suitable for the CFDA’s use to real samples.
326
Acknowledgements
327
We also would like to thank the Science & Technology Research Plan of Science and Information
328
technology of GuangZhou, China (NO. 2014J4100196) for the financial support.
329
Reference
330
Alesso, M., Bondioli, G., Talío, M. C., Luconi, M. O., & Fernández, L. P. (2012). Micelles
331
mediated separation fluorimetric methodology for Rhodamine B determination in
332
condiments, snacks and candies. Food Chemistry, 134, 513-517.
333
Alves, S. P., Brum, D. M., Branco de Andrade, É. C., & Pereira Netto, A. D. (2008).
334
Determination of synthetic dyes in selected foodstuffs by high performance liquid
335
chromatography with UV-DAD detection. Food Chemistry, 107, 489-496.
336
Amate, C. F., Unterluggauer, H., Fischer, R. J., Fernández-Alba, A. R., & Masselter, S. (2010).
337
Development and validation of a LC–MS/MS method for the simultaneous determination of
338
aflatoxins, dyes and pesticides in spices. Analytical and bioanalytical chemistry, 397,
339
93-107.
340 341
Boeniger, M. (1980). Carcinogenicity and metabolism of azo dyes, especially those derived from benzidine. NTIS, SPRINGFIELD, VA. 1980.
342
Bonan, S., Fedrizzi, G., Menotta, S., & Elisabetta, C. (2013). Simultaneous determination of
343
synthetic dyes in foodstuffs and beverages by high-performance liquid chromatography
344
coupled with diode-array detector. Dyes and Pigments, 99, 36-40.
345
Chang, X. C., Hu, X. Z., Li, Y. Q., Shang, Y. J., Liu, Y. Z., Feng, G., & Wang, J. P. (2011).
346
Multi-determination of Para red and Sudan dyes in egg by a broad specific antibody based
347
enzyme linked immunosorbent assay. Food Control, 22, 1770-1775.
348 349
Dossi, N., Toniolo, R., Susmel, S., Pizzariello, A., & Bontempelli, G. (2006). Simultaneous RP-LC determination of additives in soft drinks. Chromatographia, 63, 557-562.
350
Enríquez-Gabeiras, L., Gallego, A., Garcinuño, R. M., Fernández-Hernando, P., & Durand, J. S.
351
(2012). Interference-free determination of illegal dyes in sauces and condiments by matrix
352
solid phase dispersion (MSPD) and liquid chromatography (HPLC–DAD). Food Chemistry,
353
135, 193-198.
354
Garcia-Falcón, M. S., & Simal-Gandara, J. (2005). Determination of food dyes in soft drinks
355
containing natural pigments by liquid chromatography with minimal clean-up. Food Control,
356
16, 293-297.
357
González, M., Gallego, M., & Valcárcel, M. (2003). Liquid chromatographic determination of
358
natural and synthetic colorants in lyophilized foods using an automatic solid-phase
359
extraction system. Journal of agricultural and food chemistry, 51, 2121-2129.
360
Kirschbaum, J., Krause, C., & Brückner, H. (2006). Liquid chromatographic quantification of
361
synthetic colorants in fish roe and caviar. European Food Research and Technology, 222,
362
572-579.
363
Ma, M., Luo, X., Chen, B., Su, S., & Yao, S. (2006). Simultaneous determination of water-soluble
364
and
365
chromatography–diode array detection–electrospray mass spectrometry. Journal of
366
Chromatography A, 1103, 170-176.
367
fat-soluble
synthetic
colorants
in
foodstuff
by
high-performance
liquid
Minioti, K. S., Sakellariou, C. F., & Thomaidis, N. S. (2007). Determination of 13 synthetic food
368
colorants
in
water-soluble
foods
by
reversed-phase
high-performance
liquid
369
chromatography coupled with diode-array detector. Analytica Chimica Acta, 583, 103-110.
370
Petigara Harp, B., Miranda-Bermudez, E., & Barrows, J. N. (2013). Determination of Seven
371
Certified Color Additives in Food Products Using Liquid Chromatography. Journal of
372
agricultural and food chemistry, 61, 3726-3736.
373
Price, P. J., Suk, W. A., Freeman, A. E., Lane, W. T., Peters, R. L., Vernon, M. L., & Huebner, R.
374
J. (1978). In vitro and in vivo indications of the carcinogenicity and toxicity of food dyes.
375
International Journal of Cancer, 21, 361-367.
376
Robens, J. F., Dill, G. S., Ward, J. M., Joiner, J. R., Griesemer, R. A., & Douglas, J. F. (1980).
377
Thirteen-week subchronic toxicity studies of Direct Blue 6, Direct Black 38, and Direct
378
Brown 95 dyes. Toxicology and applied pharmacology, 54, 431-442.
379
Sun, H., Sun, N., Li, H., Zhang, J., & Yang, Y. (2013). Development of Multiresidue Analysis for
380
21 Synthetic Colorants in Meat by Microwave-Assisted Extraction–Solid-Phase
381
Extraction–Reversed-Phase
382
Analytical Methods, 6, 1291-1299.
Ultrahigh
Performance
Liquid
Chromatography.
Food
383
Tavakoli, M., Shemirani, F., & Hajimahmoodi, M. (2014). Magnetic mixed hemimicelles
384
solid-phase extraction of three food colorants from real samples. Food Analytical Methods,
385
7, 100-108.
386
Tao, Y., Chen, D., Chao, X., Yu, H., Yuanhu, P., Liu, Z., & Yuan, Z. (2011). Simultaneous
387
determination of malachite green, gentian violet and their leuco-metabolites in shrimp and
388
salmon by liquid chromatography–tandem mass spectrometry with accelerated solvent
389
extraction and auto solid-phase clean-up. Food Control, 22, 1246-1252.
390
Yuet-Wan Lok, K., Chung, W. Y., Benzie, I. F., & Woo, J. (2010). Colour additives in snack
391
foods consumed by primary school children in Hong Kong. Food Additives and
392
Contaminants, 3, 148-155.
393
Yoshioka, N., & Ichihashi, K. (2008). Determination of 40 synthetic food colors in drinks and
394
candies by high-performance liquid chromatography using a short column with photodiode
395
array detection. Talanta, 74, 1408-1413.
396
Zou, T., He, P., Yasen, A., & Li, Z. (2013). Determination of seven synthetic dyes in animal feeds
397
and meat by high performance liquid chromatography with diode array and tandem mass
398
detectors. Food chemistry, 138, 1742-1748.
399
Zhu, Y., Zhao, B., Xiao, R., Yun, W., Xiao, Z., Tu, D., & Chen, S. (2014). Simultaneous
400
determination of 14 oil-soluble synthetic dyes in chilli products by high performance liquid
401
chromatography with a gel permeation chromatography clean-up procedure. Food
402
chemistry, 145, 956-962.
403 404
Legends
405
Figure 1. Average recovery of each color additives using different extract solutions
406
Figure 2. (a) Effect of different SPE sorbent on the recovery of each color additives. (b) Typical
407
HPLC-DAD chromatograms of mixed color additives standard solutions (1.0µ g/ml).
408
Figure 3. Mass spectrum of New red(a), Amaranth(b), Allura Red(c) and Azo rubine(d)
409
Figure 4. HPLC-MS/MS response of each color additives in ESI(-) with different concentrations
410
of NH4AC in Mobile Phase A
411
Figure 5. Comparison of TIC of 0.1 µ g/ml color additives standards in pure solvent and in matrix,
412
respectively: (a) in pure solvent, (b) in flour-matrix, (c) in meat-matrix
413
Figure 6. HPLC and MRM Chromatograms for one positive meat sample
414
415
416 417 418 419 420 421 422 423 424
Figure 1. Average recovery of each color additives using different extract solutions: (a) methanol-ammonia-water, 80:0:20 (V/V/V); (b) methanol-ammonia-water, 80:2:18 (V/V/V); (c) methanol-ammonia-water, 80:4:16 (V/V/V); (d) methanol-ammonia-water, 80:6:14 (V/V/V) and (e) methanol-ammonia-water, 80:8:12(V/V/V).
425 426 427
Figure 2a. Effect of different SPE sorbent on the recovery of of each color additives.
428 430 nm
429
510 nm
430 630 nm
431 432 433 434
Figure 2b. Typical HPLC-DAD chromatograms of mixed color additives standard solutions (1.0µ g/ml).
435
436 437 438 439 440 441
Figure 3. Mass spectrum of New red(a), Amaranth(b), Allura Red(c) and Azo rubine(d).
442 443 444
Figure 4. HPLC-MS/MS response of each color additives in ESI(-) with different concentrations of NH4AC in Mobile Phase A
445 Allura red Erythrosine 446 447 448 Brilliant blue 449 a Quinoline yellow 450 disulfonate Azo rubine 451 New red m 452 Indigotin Sunset yellow 453 Quinoline yellow Poncean 4R 454 monosulfonate 455 Amaranth Tartrazin 456 457 458 459 460 461 462 b 463 464 465 466 467 468 469 470 471 472 7.80 6.21 473 1.9e5 1.8e5 474 1.7e5 5.28 475 1.6e5 1.5e5 476 1.4e5 c 477 1.3e5 1.2e5 478 1.1e5 6.44 479 1.0e5 480 9.0e4 8.0e4 481 7.0e4 482 6.0e4 4.67 483 5.0e4 4.23 4.0e4 4.01 484 3.0e4 7.00 1.73 3.25 2.0e4 2.43 485 1.0e4 486 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 Time, min 487 488 Figure 5. Comparison of TIC of 0.1 µg/ml color additives standards in pure solvent and in matrix, 5.25
1.6e5 1.5e5 1.4e5
7.72
1.3e5 1.2e5 1.1e5
9.0e4
6.18
8.0e4 7.0e4 6.0e4 5.0e4 4.0e4
6.39
3.0e4
4.01
2.0e4
4.64
1.72
1.86
2.40
3.24
6.94
1.0e4
0.0
1.0
2.0
3.0
4.0
5.0 Time, min
6.0
7.0
9.0
8.0
5.24
1.8e5 1.7e5 1.6e5
7.72
1.5e5 1.4e5 1.3e5 1.2e5
6.18
Intensity, cps
1.1e5 1.0e5 9.0e4 8.0e4 7.0e4 6.0e4
6.39
5.0e4 4.0e4
4.63
6.09
4.01
3.0e4
4.22
2.0e4
1.70 1.83
2.37
6.94
3.23
1.0e4
0.0
1.0
2.0
3.0
4.0
5.0 Time, min
6.0
7.0
9.0
8.0
Intensity, cps
Intensity, cps
1.0e5
8.5
9.0
489
respectively: (a) in pure solvent, (b) in flour-matrix, (c) in meat-matrix. 430 nm
Sunset Yellow
490 4.64
2.9e4 2.8e4
Sunset Yellow
2.6e4 2.4e4 2.2e4 2.0e4
Intensity, cps
1.8e4
M/Z 203.2 > 207.1
1.6e4 1.4e4
M/Z 203.2 > 171.2
1.2e4 1.0e4 8000.0 6000.0 4000.0 2000.0
7.29
4.01
491 492 493
0.0
1.0
2.0
3.0
4.0
8.15 8.37
5.0
6.0
7.0 Time, min
8.0
Figure 6. HPLC and MRM Chromatograms for one positive meat sample.
9.23 9.0
10.0
11.0
12.0
4 5 6 7 8 9 10 11
Highlights
A fast method for the simultaneous determination of 11 additives. HPLC coupled with DAD and MS/MS methods are used. Good recoveries in the range of 75.2-113.8%. Suitable for the routine monitoring analysis of 11 additives.