Hindawi Publishing Corporation

International Journal of Analytical Chemistry

Volume 2016, Article ID 2879406, 8 pages
http://dx.doi.org/10.1155/2016/2879406

Research Article
Development and Validation of HPLC Method for
the Simultaneous Determination of Five Food Additives
and Caffeine in Soft Drinks

Bürge AGçJ, Fule Dinç Zor, and Özlem Aksu Dönmez

Department of Chemistry, Faculty of Science and Arts, Yildiz Technical University, Davutpasa, 34220 Istanbul, Turkey

Correspondence should be addressed to Şule Dinç Zor; sule dinc@yahoo.com

Received 9 December 2015; Revised 19 January 2016; Accepted 20 January 2016

Academic Editor: Josep Esteve-Romero

Copyright © 2016 Bürge Aşçı et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Box-Behnken design was applied to optimize high performance liquid chromatography (HPLC) conditions for the simultaneous
determination of potassium sorbate, sodium benzoate, carmoisine, allura red, ponceau 4R, and caffeine in commercial soft drinks.
The experimental variables chosen were pH (6.0–7.0), flow rate (1.0–1.4 mL/min), and mobile phase ratio (85–95% acetate buffer).
Resolution values of all peak pairs were used as a response. Stationary phase was Inertsil OctaDecylSilane- (ODS-) 3V reverse
phase column (250 × 4.6 mm, 5 𝜇m) dimensions. The detection was performed at 230 nm. Optimal values were found 6.0 pH,
1.0 mL/min flow rate, and 95% mobile phase ratio for the method which was validated by calculating the linearity (𝑟2 > 0.9962),
accuracy (recoveries ≥ 95.75%), precision (intraday variation ≤ 1.923%, interday variation ≤ 1.950%), limits of detection (LODs),
and limits of quantification (LOQs) parameters. LODs and LOQs for analytes were in the range of 0.10–0.19 𝜇g/mL and 0.33–
0.63 𝜇g/mL, respectively. The proposed method was applied successfully for the simultaneous determination of the mixtures of five
food additives and caffeine in soft drinks.

1. Introduction many analytical techniques have been developed for the

determination of various food additives in foods, there is no
Food additives are widely used in foodstuffs to prevent from report about simultaneous determination of this combination
spoilage and improve color, flavor, and texture of foods. in food samples. Among these analytical methods, HPLC
However, these additives in foods may affect individuals who coupled with UV/Vis or diode array detectors (DADs) are
are sensitive with some type of allergy, asthma, and hay
the most commonly used methods due to their sensitivity,
fever. Consequently, authorities have set threshold values for
selectivity, and high resolution. So, development of effec-
acceptable daily intake, varying from country to country. For
tive chromatographic separation method involves judicious
instance, the list of authorised food additives and maximum
permitted levels in European Union are laid down in the selection of experimental conditions that is suitable for the
annexes of council directive [1, 2]. separation of interested components at an adequate resolu-
To ensure food safety from farm to fork, it is also essential tion with reasonable run time. In this regard, experimental
to develop effective and reliable analytical methods for the design is a useful tool to simplify the laborious work [18].
monitoring of the additive levels in food [3]. Therefore, It not only is a timesaving method but also it has an ability
various analytical methods have been reported for the simul- to reveal possible interactions between variables [19, 20].
taneous determination synthetic food additives, such as thin Hence, experimental designs have been increasingly used
layer chromatography [4], UV-visible spectrophotometry [5, to determine the optimum conditions of chromatographic
6], voltammetry [7, 8], differential pulse polarography [9], separation of some analytes in food, drug, and biological fluid
capillary electrophoresis [10], HPLC-DAD [11–14], HPLC- samples with a minimum number of experiments for over the
MS [15], and HPLC-MS-MS [16, 17]. Until now, although past decade [21–28].
2 International Journal of Analytical Chemistry

In this paper, a new RP-HPLC method was developed, Table 1: The experimental variables and levels of BBD.
using experimental design, for simultaneous determination
Level
of five synthetic food additives in soft drinks, including Variable
three synthetic colorants (carmoisine, allura red, and pon- −1 0 +1
ceau 4R), two preservatives (potassium sorbate and sodium pH (𝐴) 6.0 6.5 7.0
benzoate), and caffeine. For the optimization procedure, Box- Flow rate (𝐵) (mL min−1 ) 1.0 1.2 1.4
Behnken design (BBD) was used to construct mathematical Mobile phase ratio (𝐶) (%) 85 90 95
models that predict how changes input or controlled by
variables (pH, flow rate, and mobile phase ratio) affected
the resolution in defined experimental region. Further, the both soft drink samples and standard solutions were filtered
method validation has been carried out according to the through 0.45 𝜇m Millipore filters and then injected into
International Conference on Harmonization guidelines. The HPLC system.
optimized and validated method was successfully applied to
some commercial soft drinks containing potassium sorbate, 2.5. Chromatographic Procedure. The optimum separation of
sodium benzoate, carmoisine, allura red, ponceau 4R, and all analytes was achieved with 0.025 M sodium acetate/acetic
caffeine. acid buffer, pH 6.0, acetonitrile gradient that follows 0–5 min,
95 : 75 (v/v); 5–10 min, 70 : 30 (v/v). The mobile phase flow
2. Experimental rate was 1.0 mL/min and the injection volume was 20 𝜇L in
all the chromatographic runs. The detection was made with a
2.1. Apparatus. Chromatographic analyses were performed variable ultraviolet-visible detector fixed at 230 nm.
using a Shimadzu HPLC system (Kyoto, Japan) consisting
of a model LC20 AT pump unit, SPD-20A UV-Vis detector, 2.6. Optimization Procedure. A Box-Behnken design (BBD)
7725 20 𝜇L sample injection, a computer, and an Inertsil using three variables at three levels (coded levels: −1, 0,
OctaDecylSilane- (ODS-) 3V column (5 𝜇m, 250 mm × and +1) was used for the optimization of simultaneous
4.6 mm; GL Sciences, Tokyo, Japan). The statistical analysis determination of potassium sorbate, sodium benzoate, car-
for the analytical responses and validation data was evaluated moisine, allura red, ponceau 4R, and caffeine by HPLC. This
with Microsoft Excel 2000 software. The statistical software design was selected due to the small number of experiments
Statgraphics Centurion XV (StatPoint Inc., VA, USA) was required. The variables and levels selected for optimization
used for the graph plotting and for estimating the responses procedure were pH (𝐴; 6.0, 6.5, and 7.0), flow rate (𝐵; 1.0, 1.2,
of experimental variables. and 1.4), and mobile phase ratio (in terms of acetate buffer)
(𝐶; 85, 90, and 95) (Table 1). The proposed HPLC method
2.2. Chemicals and Reagents. All chemicals and solvents analyzed the compounds in two steps as mentioned above.
were of analytical reagent grade and used without further While the first step has an effect on the chromatographic
purification. Milli-Q water was used to prepare the solu- separation, the second step has an effect on the run time
tions and mobile phases (Millipore, Milford, MA, USA). of the method. Therefore, experimental variables of the
Sodium acetate trihydrate, glacial acetic acid, and HPLC- first step of HPLC method were taken into account. 15
grade acetonitrile were acquired from Merck (Darmstadt, experimental runs were performed at random and overall
Germany). Potassium sorbate (≥99.0% purity), sodium ben- resolution (𝑅) was chosen as the response for the separation
zoate (≥99.0%, purity), carmoisine (≥98.0% purity), allura of the compounds [19]. Experimental design matrix used and
red (≥98.0% purity), ponceau 4R (≥99.0% purity), and the results obtained by BBD were listed in Table 2.
caffeine (100.0% purity) were purchased from Sigma-Aldrich
(St. Louis, Missouri, USA). 2.7. Validation Procedure. In-house validation of the method
was performed according to International Conference on
2.3. Preparation of Standard Solutions. Standard stock solu- Harmonization guidelines (ICH Q2R1) [29]. Evaluated
tions of potassium sorbate, sodium benzoate, and caffeine parameters are linearity of calibration curve, limit of detec-
were prepared at a concentration of 250 𝜇g/mL. Standard tion (LOD), limit of quantification (LOQ), and precision,
stock solutions of carmoisine, allura red, and ponceau 4R accuracy, and stability. The linearity of the HPLC method
were prepared at a concentration of 100 𝜇g/mL. Fresh work- for the determination of five food additives and caffeine
ing solutions in the concentration range of 2–10 𝜇g/mL for was evaluated in a concentration range of 2–10 𝜇g/mL for
carmoisine, allura red, and ponceau 4R and 5–25 𝜇g/mL carmoisine, allura red, and ponceau 4R and 5–25 𝜇g/mL for
for caffeine, potassium sorbate, and sodium benzoate were potassium sorbate, sodium benzoate, and caffeine covering
prepared by the dilution of the standard stock solutions in the normal range of concentrations obtained when analyzing
Milli-Q water. soft drinks. Calibration equations were calculated by the least
squares treatment of the peak area of the food additives
2.4. Sample Preparation. Soft drink samples were purchased and caffeine. The limit of detection (LOD) and limit of
from local supermarkets in Istanbul, Turkey, and were quantitation (LOQ) were calculated as LOD 3𝑥𝜎/𝑆 and LOQ
degassed in an ultrasonic bath for 5 min. Then, 1 mL of 10𝑥𝜎/𝑆, where 𝜎 is the standard deviation of intercept and 𝑆
the sample was transferred to a 10 mL volumetric flask and is the slope. In order to test the prediction performance of the
diluted to the volume with Milli-Q water. Prior to the analysis, proposed methods, intraday (three times in a day operation
International Journal of Analytical Chemistry 3

Table 2: Experimental design matrix and the responses for BBD. Table 3: ANOVA results for optimization by BBD.

Run 𝐴 𝐵 𝐶 𝑅 Effect SS D.f. MS 𝐹-ratio 𝑃 value

1 6.0 1.0 90 1.041 𝐴 239422 1 239422 1.95 0.2216
2 7.0 1.0 90 0.622 𝐵 189543 1 189543 1.54 0.2694
3 6.0 1.4 90 0.553 𝐶 9489050 1 9489050 77.20 0.0003∗
4 7.0 1.4 90 0.128 𝐴𝐴 86144 1 86144 0.70 0.4407
5 6.0 1.2 85 0.000 𝐴𝐵 0 1 0 0.00 1.0000
6 7.0 1.2 85 0.027 𝐴𝐶 478304 1 478304 3.89 0.1056
7 6.0 1.2 95 3175.373 𝐵𝐵 86239 1 86239 0.70 0.4404
8 7.0 1.2 95 1792.207 𝐵𝐶 378422 1 378422 3.08 0.1397
9 6.5 1.0 85 0.046 𝐶𝐶 4374670 1 4374670 35.59 0.0019∗
10 6.5 1.4 85 0.000 Total error 614539 5 122908
11 6.5 1.0 95 2487.823 Total (corr.) 15999393 14
12 6.5 1.4 95 1257.455 SS: sum of squares; MS: mean squares; 𝐹-ratio: MS/MSerror ; 𝑃 value:
13 6.5 1.2 90 0.759 probability level; D.f.: degree of freedom.
𝑅2 = 0.961, 𝑅2 (adjusted for D.f.) = 0.892.
14 6.5 1.2 90 0.327 ∗
Significant factor at 𝛼 = 0.05.
15 6.5 1.2 90 0.918

2600
under the same conditions) and interday (four different days)
2100
studies were performed at three different concentrations
(Level 1: 10 𝜇g/mL; Level 2: 15 𝜇g/mL; Level 3: 20 𝜇g/mL for 1600
potassium sorbate, sodium benzoate, and caffeine; Level 1: R 1100
4 𝜇g/mL; Level 2: 6 𝜇g/mL; Level 3: 8 𝜇g/mL for carmoisine, 600
allura red, and ponceau 4R). Accuracy of the method was
100
ascertained by a recovery study by adding a known amount
of reference standards to the soft drink samples. Firstly, −400
−1.0 1.0 −1.0 1.0 −1.0 1.0
0.5 mL of the soft drink sample was transferred to a 10 mL
A B C
volumetric flask and the reference standards were added on
it at three different concentration levels. Then, added samples Figure 1: Analysis of the main variables in BBD.
were diluted to the volume with Milli-Q water, filtered, and
analyzed.
The experimental results are shown in Table 2. The
3. Results and Discussion regression model for the response was tested through analysis
of variance (ANOVA). From the results of ANOVA (Table 3),
3.1. Optimization of the HPLC Method. Chromatographic it can be deduced that linear contribution of mobile phase
optimization requires selecting suitable criteria for the eval- ratio (𝐶) and quadratic contribution of mobile phase ratio
uation of the resultant chromatograms in order to choose (𝐶𝐶) influence the resolution significantly. Interactions of
the optimum conditions. BBD is an independent, rotatable, the individual variables in this study are not significant
or nearly rotatable second-order design based on three-level to resolution in the selected range. Fitted quadratic model
incomplete factorial designs. It is more efficient compared equation is also presented in (2). Figure 1 shows the analysis
to other response surface designs, such as central composite of individual variables of experimental design. From Figure 1,
designs. It can also provide sufficient information to test it can be seen how the value of the resolution may increase if
the lack of fit, and therefore it is one of the best quadratic we take higher mobile phase ratio (𝐶). Also, we can infer that
models for response surface method and has been widely although pH (𝐴) and flow rate (𝐵) do not greatly influence
used in analytical fields. Because of the nonlinearity of the the resolution better resolutions are obtained for low values
model, a polynomial function to contain second-order model of pH and flow rate
is postulated to describe the evolution phenomenon:
𝑛 𝑛 𝑛 ̂ = 0.67 − 172.99𝐴 − 153.92𝐵 + 1089.10𝐶
𝑅
𝑦𝑖 = 𝑏0 + ∑𝑏𝑖 𝑥𝑖 + ∑𝑏𝑖𝑖 𝑥𝑖2 + ∑ 𝑏𝑖𝑗 𝑥𝑖 𝑥𝑗 + 𝜀𝑖 , (1)
𝑖=1 𝑖=1 1≤𝑖≤𝑗 + 152.74𝐴2 − 345.80𝐴𝐶 − 152.83𝐵2 − 307.58𝐵𝐶 (2)

where 𝑛 is the number of variables, 𝑏0 is the constant term, 𝑏𝑖 , + 1088.49𝐶2 .

𝑏𝑖𝑖 , and 𝑏𝑖𝑗 represent the coefficient of the first-order terms,
quadratic terms, and interaction terms, respectively, and 𝜀𝑖 The regression models obtained were used to calculate
is a term that represents other sources of variability not the response surface for each variable separately. Figure 2
accounted for the estimation, such as background noise [30]. illustrates the response surface plots for the resolutions. In
4 International Journal of Analytical Chemistry

Table 4: The important parameters of the calibration equations for the proposed HPLC method for simultaneous determination of potassium
sorbate (SOR), sodium benzoate (BEN), carmoisine (CAR), allura red (ALU), ponceau 4R (PON), and caffeine (CAF).

Compounds Calibration range Regression equation 𝑆𝑎 𝑆𝑏 𝑟2 LOD LOQ

(𝜇g/mL) (𝑌 = 𝑎𝑋 + 𝑏) (𝜇g/mL) (𝜇g/mL)
3 4
SOR 5–25 𝑌 = 7.8097 × 104 𝑋 + 4.6484 × 104 1.391 × 10 2.3076 × 10 0.9990 0.12 0.40
BEN 5–25 𝑌 = 3.3313 × 104 𝑋 + 1.2421 × 104 7.52 × 102 1.2469 × 104 0.9985 0.10 0.33
CAR 2–10 4
𝑌 = 4.3383 × 10 𝑋 + 1.8834 × 10 4
6.86 × 102 4.555 × 103 0.9992 0.16 0.53
ALU 2–10 𝑌 = 5.5741 × 104 𝑋 − 7.427 × 103 9.22 × 102 1.5292 × 104 0.9992 0.11 0.35
3 4
PON 2–10 𝑌 = 6.043 × 103 𝑋 + 5.0389 × 104 2.150 × 10 1.4264 × 10 0.9962 0.17 0.56
4 4 2 4
CAF 5–25 𝑌 = 2.4885 × 10 𝑋 + 4.6613 × 10 4.81 × 10 3.1891 × 10 0.9989 0.19 0.63
𝑌: peak area; 𝑋: concentration (𝜇g/mL); 𝑆𝑎 : standard deviation of the slope; 𝑆𝑏 : standard deviation of the intercept.

3600 3500
2600 2500
R 1600 R 1500
600 500
−400 0.6 −500 0.6
−1 −0.2 −1 −0.2
−0.6 −0.2 0.2 0.6 −1 B −0.6 −0.2 0.2 0.6 −1 C
1 1
A A
(a) (b)

3600
2600
R 1600
600
−400 0.6
−0.2
−1 −0.6 −0.2 0.2 −1 C
0.6 1
B
(c)

Figure 2: Response surface plots for BBD: (a) pH (𝐴) versus flow rate (𝐵) (mobile phase ratio: 95%); (b) pH (𝐴) versus mobile phase ratio
(𝐶) (flow rate: 1.0 mL/min); (c) flow rate (𝐵) versus mobile phase ratio (𝐶) (pH: 6.0).

particular, the effect of pH (𝐴) and mobile phase ratio (𝐶) Detector A: 230 nm
on resolution is shown in Figure 2(b). This plot shows that 400
1 2 3 4
350
the highest resolution is obtained at greater values of the 300
mobile phase ratio. The relation between the effects of the 250
(mV)

other variables on the resolution is also plotted in Figures 200

150 5
2(a)–2(c). 100
According to the results of the optimization proce- 50 6
dure, the optimum variables corresponded to pH, 6.0; flow 0
rate, 1.0 mL/min; mobile phase ratio, 95%. A typical chro- 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
matogram obtained under optimum conditions is shown in (min)
Figure 3.
Figure 3: Chromatogram of synthetic standard mixture containing
five food additives and caffeine recorded under optimized analysis
3.2. Validation of the HPLC Method. The results of the conditions (1: sodium benzoate (15 𝜇g/mL), 2: potassium sorbate
linearity, LODs, and LOQs are summarized in Table 4. A good (15 𝜇g/mL), 3: caffeine (15 𝜇g/mL), 4: ponceau 4R (6 𝜇g/mL), 5: allura
linear relationship is displayed between the corresponding red (6 𝜇g/mL), and 6: carmoisine (6 𝜇g/mL)).
peak areas and the concentrations of the compounds based
on the correlation coefficients (𝑟2 > 0.9962). The LODs of
the six compounds were in the range of 0.10–0.19 𝜇g/mL, and
the LOQs of the six compounds were in the range of 0.33– analytical method was sufficiently sensitive. A summary of
0.63 𝜇g/mL. So, these values demonstrated that the proposed intraday (the RSD of the recoveries of the nine samples) and
Table 5: Validation of the simultaneous determination of food additives and caffeine by developed method potassium sorbate (SOR), sodium benzoate (BEN), carmoisine (CAR), allura red
(ALU), ponceau 4R (PON), and caffeine (CAF).
International Journal of Analytical Chemistry

SOR (𝜇g/mL) BEN (𝜇g/mL) CAR (𝜇g/mL) ALU (𝜇g/mL) PON (𝜇g/mL) CAF (𝜇g/mL)
Levels
10.0 15.0 20.0 10.0 15.0 20.0 4.0 6.0 8.0 4.0 6.0 8.0 4.0 6.0 8.0 10.0 15.0 20.0
9.97 15.22 19.76 9.70 15.37 20.18 4.05 6.17 7.92 3.90 6.02 7.86 3.79 5.72 7.75 9.96 15.15 20.28
Intra-assay 9.80 15.30 20.01 10.02 15.09 20.21 4.15 6.33 8.01 3.83 5.97 7.82 3.89 5.92 7.91 9.99 15.50 20.02
9.91 15.31 19.86 10.03 14.79 20.38 4.18 6.20 8.11 3.87 6.06 7.70 3.92 5.87 8.04 10.03 15.16 20.20
Mean 9.89 15.28 19.88 9.92 15.08 20.26 4.13 6.23 8.01 3.87 6.02 7.79 3.87 5.84 7.91 9.99 15.27 20.16
RSDa % 0.869 0.321 0.634 1.895 1.923 0.533 1.646 1.364 1.186 0.904 0.747 1.060 1.757 1.781 1.643 0.350 0.310 0.689
Recovery% 98.90 101.87 99.40 99.20 100.53 101.30 103.25 103.83 100.13 96.75 100.33 97.83 96.75 97.33 98.88 99.90 101.80 100.80
9.92 15.09 19.66 9.84 14.76 20.71 4.14 6.01 7.81 3.91 6.01 7.96 4.10 6.01 8.20 10.35 15.10 20.09
9.84 15.23 19.59 9.85 14.95 20.74 4.18 6.06 8.14 3.85 5.84 7.90 4.08 6.18 8.26 10.01 15.32 20.43
Interassay
9.77 15.21 19.69 9.81 14.79 20.25 4.05 6.19 7.96 3.81 5.86 8.04 3.83 6.20 8.24 10.17 14.88 20.07
9.68 15.40 19.89 10.02 14.82 20.11 4.19 6.22 7.98 3.76 6.08 7.82 4.01 5.98 7.99 10.03 15.16 20.20
Mean 9.80 15.23 19.71 9.88 14.83 20.45 4.14 6.12 7.97 3.83 5.95 7.93 4.03 6.09 8.14 10.14 15.11 20.20
RSD% 1.041 0.840 0.624 0.961 0.566 1.565 1.546 1.650 1.694 1.644 1.950 1.173 1.919 1.855 1.518 1.548 1.204 0.817
Recovery% 98.00 101.53 98.55 98.80 98.87 102.25 103.50 102.00 99.62 95.75 99.17 99.13 100.75 101.50 102.12 101.40 100.73 101.00
a
Relative standard deviation, RSD (%), standard deviation/mean × 100.
5
6 International Journal of Analytical Chemistry

Table 6: Results of accuracy studies (mean value ± standard deviation, 𝑛 = 5).

Sample Food additive Sample concentration (𝜇g/mL) Added (𝜇g/mL) Found (𝜇g/mL) Recovery (%)
3 13.70 ± 0.048 100.67
Potassium sorbate 10.68 6 16.90 ± 0.057 103.67
9 20.00 ± 0.065 103.56
3 10.18 ± 0.045 102.33
Energy drink Caffeine 7.11 6 13.25 ± 0.054 102.33
9 16.61 ± 0.042 105.56
1 4.31 ± 0.038 105.00
Allura red 3.26 2 5.34 ± 0.037 104.00
3 6.39 ± 0.044 104.33
3 15.64 ± 0.059 97.67
Potassium sorbate 12.71 6 18.61 ± 0.103 98.33
9 21.74 ± 0.166 100.34
3 10.34 ± 0.040 103.60
Sodium benzoate 7.24 6 13.30 ± 0.054 101.20
9 16.45 ± 0.063 102.50
Pomegranate juice
1 2.25 ± 0.012 104.00
Carmoisine 1.21 2 3.30 ± 0.017 104.50
3 4.34 ± 0.022 104.40
1 5.11 ± 0.036 102.66
Ponceau 4R 4.08 2 6.16 ± 0.039 104.16
3 7.15 ± 0.045 102.22

interday precision (the RSD of the recoveries of the twelve Table 7: Analysis of soft drinks (mean value ± standard deviation,
samples) are listed in Table 5. The RSD values ranged from 𝑛 = 5).
0.310% to 1.950% for the HPLC method. These results show
Pomegranate Mandarin
that the proposed method is precise for the simultaneous Food additive Energy drink
juice juice
determination of these compounds. The recoveries of the six (𝜇g/mL)
(𝜇g/mL) (𝜇g/mL)
compounds to determine the accuracy of the method are Potassium sorbate 213.62 ± 0.34 254.13 ± 1.24 246.23 ± 1.76
summarized in Table 6. The proposed method resulted in
satisfactory recoveries for all additives and caffeine, ranging Sodium benzoate — 144.71 ± 2.38 148.67 ± 1.99
from 97.67% to 105.56%. The recoveries demonstrated that the Caffeine 142.20 ± 1.17 — —
matrixes have negligible effect on the quantification of these Allura red 65.28 ± 0.59 — —
compounds and the method is accurate within the desired Carmoisine — 24.26 ± 0.47 —
range. Under refrigerated and room temperature conditions, Ponceau 4R — 81.58 ± 1.51 —
all food additives and caffeine in mobile phase and water were
stable for at least 1 month.
These results show that the proposed method is precise,
accurate, and sensitive for the simultaneous determination of
the six compounds and can be used for routine analysis of 4. Conclusion
potassium sorbate, sodium benzoate, carmoisine, allura red, An efficient, accurate, and reliable method for the simulta-
ponceau 4R, and caffeine in soft drinks. neous determination of five food additives and caffeine in
soft drinks was developed using HPLC. Box-Behnken design
3.3. Application of the Method. The proposed HPLC method was applied to the optimization of the chromatographic sep-
was applied to the simultaneous determination of potassium aration conditions and this design reduced to the number of
sorbate, sodium benzoate, carmoisine, allura red, ponceau experiments required. It can be concluded that a slight change
4R, and caffeine in different soft drinks. Five replicates in mobile phase ratio has a direct effect on the resolution. All
determination was made and the results are summarized in the validation parameters were within the acceptance range.
Table 7. The concentration of food additives in soft drinks High percentage recovery data also shows that the proposed
ranged from 24.26 ± 0.47 𝜇g/mL to 254.13 ± 1.24 𝜇g/mL. method is free from the interference. Consequently, this study
The amounts of food additives and caffeine in all soft drink will provide a sensitive and rapid method for the detection of
samples were below the limit value defined in the legislation potassium sorbate, sodium benzoate, carmoisine, allura red,
on the food additives [1, 2]. ponceau 4R, and caffeine in soft drinks.
International Journal of Analytical Chemistry 7

Conflict of Interests in ham sausage by HPLC-DAD method,” Journal of Liquid

Chromatography & Related Technologies, vol. 35, no. 2, pp. 268–
The authors declare that there is no conflict of interests 279, 2012.
regarding the publication of this paper. [14] K. Ma, Y. N. Yang, X. X. Jiang, M. Zhao, and Y. Q. Cai,
“Simultaneous determination of 20 food additives by high
Acknowledgment performance liquid chromatography with photo-diode array
detector,” Chinese Chemical Letters, vol. 23, no. 4, pp. 492–495,
This work was financially supported by Yildiz Technical Uni- 2012.
versity Research Foundation (Project no. 2010-01-01-GEP04). [15] M. Ma, X. Luo, B. Chen, S. Su, and S. Yao, “Simultaneous deter-
mination of water-soluble and fat-soluble synthetic colorants
in foodstuff by high-performance liquid chromatography-diode
References array detection-electrospray mass spectrometry,” Journal of
Chromatography A, vol. 1103, no. 1, pp. 170–176, 2006.
[1] Regulation (EC) no 1333/2008 of the European Parliament and
of the Council on Food Additives, 2008. [16] H. Gao, M. Yang, M. Wang, Y. Zhao, Y. Cao, and X. Chu,
“Determination of 30 synthetic food additives in soft drinks
[2] European Commission, Commission Directive 2008/84/EC,
by HPLC/electrospray ionization-tandem mass spectrometry,”
Laying down Specific Purity Criteria on Food Additives Other
Journal of AOAC International, vol. 96, no. 1, pp. 110–115, 2013.
Than Colours and Sweeteners, 2008.
[17] F. Feng, Y. Zhao, W. Yong, L. Sun, G. Jiang, and X. Chu,
[3] A. K. Malik, C. Blasco, and Y. Pico, “Liquid chromatography-
“Highly sensitive and accurate screening of 40 dyes in soft
mass spectrometry in food safety,” Journal of Chromatography
drinks by liquid chromatography–electrospray tandem mass
A, vol. 1217, no. 25, pp. 4018–4040, 2010.
spectrometry,” Journal of Chromatography B, vol. 879, no. 20,
[4] I. Baranowska, M. Zydroń, and K. Szczepanik, “TLC in the pp. 1813–1818, 2011.
analysis of food additives,” Journal of Planar Chromatography—
Modern TLC, vol. 17, no. 1, pp. 54–57, 2004. [18] R. Noguerol-Cal, J. M. López-Vilariño, M. V. González-
Rodrı́guez, and L. F. Barral-Losada, “Development of an
[5] E. Dinç, E. Baydan, M. Kanbur, and F. Onur, “Spectrophotomet-
ultraperformance liquid chromatography method for improved
ric multicomponent determination of sunset yellow, tartrazine
determination of additives in polymeric materials,” Journal of
and allura red in soft drink powder by double divisor-ratio spec-
Separation Science, vol. 30, no. 15, pp. 2452–2459, 2007.
tra derivative, inverse least-squares and principal component
regression methods,” Talanta, vol. 58, no. 3, pp. 579–594, 2002. [19] R. Gheshlaghi, J. M. Scharer, M. Moo-Young, and P. L. Douglas,
“Application of statistical design for the optimization of amino
[6] S. Altinöz and S. Toptan, “Simultaneous determination of
acid separation by reverse-phase HPLC,” Analytical Biochem-
Indigotin and Ponceau-4R in food samples by using Vierordt’s
istry, vol. 383, no. 1, pp. 93–102, 2008.
method, ratio spectra first order derivative and derivative UV
spectrophotometry,” Journal of Food Composition and Analysis, [20] D. C. Montgomery, Design and Analysis of Experiments, John
vol. 16, no. 4, pp. 517–530, 2003. Wiley & Sons, New York, NY, USA, 2004.
[7] A. H. Alghamdi, “Determination of allura red in some food [21] S. L. C. Ferreira, R. E. Bruns, E. G. P. da Silva et al., “Statistical
samples by adsorptive stripping voltammetry,” Journal of AOAC designs and response surface techniques for the optimization
International, vol. 88, no. 5, pp. 1387–1393, 2005. of chromatographic systems,” Journal of Chromatography A, vol.
[8] X. Yang, H. Qin, M. Gao, and H. Zhang, “Simultaneous 1158, no. 1-2, pp. 2–14, 2007.
detection of Ponceat 4R and tartrazine in food using adsorptive [22] D. B. Hibbert, “Experimental design in chromatography: a
stripping voltammetry on an acetylene black nanoparticle- tutorial review,” Journal of Chromatography B: Analytical Tech-
modified electrode,” Journal of the Science of Food and Agricul- nologies in the Biomedical and Life Sciences, vol. 910, no. 1, pp.
ture, vol. 91, no. 15, pp. 2821–2825, 2011. 2–13, 2012.
[9] S. Chanlon, L. Joly-Pottuz, M. Chatelut, O. Vittori, and J. L. [23] F. van de Velde, M. E. Pirovani, M. S. Cámara, D. R. Güemes,
Cretier, “Determination of carmoisine, allura red and ponceau and C. M. D. H. Bernardi, “Optimization and validation of a
4R in sweets and soft drinks by differential pulse polarography,” UV–HPLC method for vitamin C determination in strawberries
Journal of Food Composition and Analysis, vol. 18, no. 6, pp. 503– (Fragaria ananassa Duch.), using experimental designs,” Food
515, 2005. Analytical Methods, vol. 5, no. 5, pp. 1097–1104, 2012.
[10] M. A. Prado, L. F. V. Boas, M. R. Bronze, and H. T. Godoy, [24] N. Garcı́a-Villar, J. Saurina, and S. Hernández-Cassou, “High-
“Validation of methodology for simultaneous determination of performance liquid chromatographic determination of bio-
synthetic dyes in alcoholic beverages by capillary electrophore- genic amines in wines with an experimental design optimiza-
sis,” Journal of Chromatography A, vol. 1136, no. 2, pp. 231–236, tion procedure,” Analytica Chimica Acta, vol. 575, no. 1, pp. 97–
2006. 105, 2006.
[11] N. Dossi, R. Toniolo, S. Susmel, A. Pizzariello, and G. Bontem- [25] T. Sivakumar, R. Manavalan, C. Muralidharan, and K. Valliap-
pelli, “Simultaneous RP-LC determination of additives in soft pan, “Multi-criteria decision making approach and experimen-
drinks,” Chromatographia, vol. 63, no. 11-12, pp. 557–562, 2006. tal design as chemometric tools to optimize HPLC separation
[12] N. O. Can, G. Arlı, and Y. Lafçı, “A novel RP-HPLC method for of domperidone and pantoprazole,” Journal of Pharmaceutical
simultaneous determination of potassium sorbate and sodium and Biomedical Analysis, vol. 43, no. 5, pp. 1842–1848, 2007.
benzoate in soft drinks using C18-bonded monolithic silica [26] M. Medenica, B. Jancic, D. Ivanovic, and A. Malenovic, “Experi-
column,” Journal of Separation Science, vol. 34, no. 16-17, pp. mental design in reversed-phase high-performance liquid chro-
2214–2222, 2011. matographic analysis of imatinib mesylate and its impurity,”
[13] Q. H. Yan, L. Yang, H. R. Zhang, and L. Y. A. Niu, “A sensitive Journal of Chromatography A, vol. 1031, no. 1-2, pp. 243–248,
and validated method for determination of four additives 2004.
8 International Journal of Analytical Chemistry

[27] G. Iriarte, N. Ferreirós, I. Ibarrondo et al., “Optimization

via experimental design of an SPE-HPLC-UV-fluorescence
method for the determination of valsartan and its metabolite
in human plasma samples,” Journal of Separation Science, vol.
29, no. 15, pp. 2265–2283, 2006.
[28] E. Nemutlu, S. Kır, D. Katlan, and M. S. Beksaç, “Simultaneous
multiresponse optimization of an HPLC method to separate
seven cephalosporins in plasma and amniotic fluid: application
to validation and quantification of cefepime, cefixime and
cefoperazone,” Talanta, vol. 80, no. 1, pp. 117–126, 2009.
[29] International Conference on Harmonization (ICH) of Tech-
nical Requirements for Registration of Pharmaceuticals for
Human Use, Topic Q2 (R1): Validation of Analytical Procedures:
Text and Methodology, International Conference on Harmo-
nization (ICH), Geneva, Switzerland, 2005.
[30] J.-Z. Song, C.-F. Qiao, S.-L. Li, Y. Zhou, M.-T. Hsieh, and H.-
X. Xu, “Rapid optimization of dual-mode gradient high perfor-
mance liquid chromatographic separation of Radix et Rhizoma
Salviae Miltiorrhizae by response surface methodology,” Journal
of Chromatography A, vol. 1216, no. 42, pp. 7007–7012, 2009.
 Photoenergy
International Journal of
International Journal of Organic Chemistry International Journal of Advances in
Medicinal Chemistry
Hindawi Publishing Corporation
International
Hindawi Publishing Corporation Hindawi Publishing Corporation
Analytical Chemistry
Hindawi Publishing Corporation
Physical Chemistry
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

International Journal of

Carbohydrate Journal of
Chemistry
Hindawi Publishing Corporation
Quantum Chemistry
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Submit your manuscripts at

http://www.hindawi.com

Journal of
The Scientific Analytical Methods
World Journal
Hindawi Publishing Corporation
in Chemistry
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of International Journal of International Journal of Journal of Bioinorganic Chemistry

Spectroscopy
Hindawi Publishing Corporation
Inorganic Chemistry
Hindawi Publishing Corporation
Electrochemistry
Hindawi Publishing Corporation
Applied Chemistry
Hindawi Publishing Corporation
and Applications
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of  Chromatography   Journal of Journal of International Journal of

Theoretical Chemistry
Hindawi Publishing Corporation
Research International
Hindawi Publishing Corporation
Catalysts
Hindawi Publishing Corporation
Chemistry
Hindawi Publishing Corporation
Spectroscopy
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014