2111
J. Sep. Sci. 2014, 37, 2111–2117
¨ ˘ Yılmaz1 Pelin Koseo glu Abdulselam Ertas¸2 Ufuk Kolak1 1 Department
of Analytical Chemistry, Faculty of Pharmacy, Istanbul University, Istanbul, Turkey 2 Department of Pharmacognosy, Faculty of Pharmacy, Dicle University, Diyarbakir, Turkey Received April 14, 2014 Revised May 13, 2014 Accepted May 18, 2014
Research Article
Simultaneous determination of seven phthalic acid esters in beverages using ultrasound and vortex-assisted dispersive liquid–liquid microextraction followed by high-performance liquid chromatography A sensitive, rapid, and simple high-performance liquid chromatography with UV detection method was developed for the simultaneous determination of seven phthalic acid esters (dimethyl phthalate, dipropyl phthalate, di-n-butyl phthalate, benzyl butyl phthalate, dicyclohexyl phthalate, di-(2-ethylhexyl) phthalate, and di-n-octyl phthalate) in several kinds of beverage samples. Ultrasound and vortex-assisted dispersive liquid–liquid microextraction method was used. The separation was performed using an Intersil ODS-3 column (C18 , 250 × 4.6 mm, 5.0 m) and a gradient elution with a mobile phase consisting of MeOH/ACN (50:50) and 0.2 M KH2 PO4 buffer. Analytes were detected by a UV detector at 230 nm. The developed method was validated in terms of linearity, limit of detection, limit of quantification, repeatability, accuracy, and recovery. Calibration equations and correlation coefficients (> 0.99) were calculated by least squares method with weighting factor. The limit of detection and quantification were in the range of 0.019–0.208 and 0.072–0.483 g/L. The repeatability and intermediate precision were determined in terms of relative standard deviation to be within 0.03–3.93 and 0.02–4.74%, respectively. The accuracy was found to be in the range of –14.55 to 15.57% in terms of relative error. Seventeen different beverage samples in plastic bottles were successfully analyzed, and ten of them were found to be contaminated by different phthalic acid esters. Keywords: Beverages / High-performance liquid chromatography / Liquid–liquid microextraction / Phthalic acid esters DOI 10.1002/jssc.201400408
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Phthalic acid esters (PAEs) are a class of dialkyl or alkyl aryl esters of 1,2-benzenedicarboxylic acid. They are widely used in food packaging materials, personal care products, children’s toys, school supplies, household cleaners, glues, paints, automobile parts, home d´ecor, and building materials for flexibilCorrespondence: Prof. Ufuk Kolak, Department of Analytical Chemistry, Faculty of Pharmacy, Istanbul University, 34116 Istanbul, Turkey E-mail: [email protected]; [email protected] Fax: +902124400254
Abbreviations: BBP, benzyl butyl phthalate; DBP, di-nbutyl phthalate; DCHP, dicyclohexyl phthalate; DEHP, di(2-ethylhexyl) phthalate; DLLME, dispersive liquid–liquid microextraction; DMP, dimethyl phthalate; DOP, di-n-octyl phthalate; DPP, dipropyl phthalate; PAE, phthalic acid ester; USVADLLME, ultrasound and vortex-assisted dispersive liquid–liquid microextraction
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ity, or to hold color and scent due to their excellent properties and compatibility with polymers [1–3]. All products may be contaminated with PAEs during production, manufacture, storage, and/or usage since there is no covalent bond between the PAEs and polymer chains. The migration of PAEs from plastic packages into the foods and beverages is one of the main human exposure sources. On the other hand, PAEs can also be easily transferred from a product onto the skin and released into the air. Since research on animals indicated that PAEs possessed some carcinogenic effects [2] and also caused fetal defects, uterine damage, disturbances in the male reproductive tract, and endocrine system disrupting activities [4, 5], some countries are prohibiting the usage of certain PAEs in plastic materials [6]. As Cinelli et al. reported, any limits of PAEs for foods and beverages have not been determined [7]. On the other hand, the European Food Safety Authority (EFSA) recommended the tolerable daily intakes of dibutyl phthalate (DBP), benzyl butyl phthalate (BBP), di-2-ethylhexyl phthalate (DEHP),
www.jss-journal.com
2112
J. Sep. Sci. 2014, 37, 2111–2117
P. K. Yılmaz et al.
Figure 1. Chemical structures of the determined PAEs.
di-isononyl phthalate (DiNP), and di-isodecyl phthalate (DiDP) as 0.01, 0.50, 0.05, 0.15, and 0.15 mg/kg body weight/day, respectively [8–12]. Considering the negative impacts of PAEs on human health, development of reliable analytical methods for the determination of PAEs is an important challenge. Because of the low concentration levels and the effects of complex sample matrices, different types of pretreatment techniques have been developed for the preconcentration of the PAEs. LLE [13–15] and SPE [2, 16] are the traditional pretreatment methods, but they have some disadvantages such as requirement of large volumes of sample and extraction solvent and are time consuming. Dispersive liquid–liquid microextraction (DLLME) is one of the recently developed techniques to overcome those disadvantages. DLLME is based on the extraction of the analyte by using an immiscible extracting solvent and a disperser solvent. The disperser solvent is soluble in both of the liquid phases and increases the contact between the two immiscible liquids [17]. Mousa et al. determined PAEs in bottled water samples using DLLME coupled with GC–MS [18]. The extracting solvents used for this purpose are generally chlorinated ones with serious toxic effects on human health. Recently ultrasound and vortex-assisted dispersive liquid–liquid microextraction (USVADLLME) procedures were developed to reduce the required volume of those hazardous extracting solvents. Ultrasonication provides a dispersed phase for the quantitative extraction of the analyte, and vortex prevents the formation of a biphasic system [7]. The detection of PAEs is generally performed by GC or HPLC. HPLC–UV is one of the commonly used techniques for determination of PAEs [19–21] followed by GC–MS [6, 22, 23], GC with flame ionization detection (FID) [24,25], and HPLC–MS [26, 27]. In this work, we aimed to develop a sensitive, rapid, and simple method for determination of seven PAEs (dimethyl phthalate (DMP), dipropyl phthalate (DPP), DBP, BBP, dicyclohexyl phthalate (DCHP), DEHP, and di-n-octyl phthalate (DOP)) in beverage samples using USVADLLME followed by HPLC–UV (Fig. 1). The developed method was validated in terms of linearity, LOD, LOQ, repeatability, accuracy, and recovery. Then the proposed method was used to detect the PAE contamination levels of 17 different beverage samples from Istanbul (Turkey). To the best of our knowledge, this
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
study could be considered as the first report on detection of PAEs in beverage samples from Turkey.
2 Materials and methods 2.1 Reagents and standards The standards of DMP, DPP, DBP, BBP, DCHP, DEHP, and DOP were purchased from Sigma (Darmstadt, Germany). The stock solution that contained each of seven PAEs at a concentration of 1000 mg/L was prepared with HPLCgrade methanol. The standard solutions were prepared daily by dilution with HPLC-grade methanol to desired concentrations. Methanol (MeOH) and acetonitrile (ACN) (HPLCgrade), dichloromethane (CH2 Cl2 ), chloroform (CHCl3 ), carbon tetrachloride (CCl4 ), sodium chloride (NaCl), and potassium dihydrogen phosphate KH2 PO4 were purchased from Merck (Darmstadt, Germany). All of the glassware used was washed with chloroform and dried at 90⬚C to avoid any contamination of PAEs. A blank analysis was also performed to check the purity of the chemicals used in terms of PAE contamination.
2.2 Beverage samples The beverages in plastic bottles including lemon juice (LJ-1 and LJ-2), ice tea (IT-1 and IT-2), cherry juice (CJ-1 and CJ-2), vinegar (V-1 and V-2), turnip juice (TJ), lemon sauce (LS-1 and LS-2), mineral water (MW), coke (C-1 and C-2), soda (S-1 and S-2), and sports drink (SD) samples were purchased from a local market in Istanbul (Turkey) in 2013. All samples were kept at +4⬚C.
2.3 Instruments and analytical conditions A Shimadzu (Shimadzu, Kyoto, Japan) LC20AT HPLC system with UV detection was used for the quantitative analysis of PAEs. The separation of PAEs was accomplished with a GL Sciences (GL Sciences, Tokyo, Japan) Intersil ODS-3 column
www.jss-journal.com
J. Sep. Sci. 2014, 37, 2111–2117
(C18 , 250 × 4.6 mm, 5.0 m). The data obtained were analyzed by the LabSolutions software (version 1.25). A gradient program with a mobile phase system consisting of two parts as eluate A (0.2 M KH2 PO4 buffer, pH 2.6) and eluate B (MeOH/ACN 50:50, v/v) was established for the elution of PAEs. After the stabilization of the system with 75% eluent B flow, a linear gradient from 0.0 to 3.0 min from 75 to 90% of eluent B flow, from 3.0 to 5.0 min from 90 to 95% of eluent B flow, from 5.0 to 9.0 min from 95 to 100% of eluent B flow, from 9.0 to 11.0 min from 100 to 95% of eluent B flow, from 11.0 to 16.0 min from 95 to 75% of eluent B flow, and then for stabilization of the system an isocratic hold from 16.0 to 20.0 min at 75% eluent B flow were applied (Supporting Information Table S1). The total analysis time was 20 min for seven PAEs. The flow rate was set to 1 mL/min and the injection volume was 20 L. The column temperature was adjusted to 40⬚C. 2.4 Quantification PAEs were identified by comparing their retention times with those of the ones in the beverage samples and with the increase of the peak areas after spiking the beverage samples. The quantification was performed by the external standard method. The calibration curves were prepared in the concentration range of 0.05–60.00 g/L for DMP, DPP, DBP, BBP, and DCHP, whereas 0.10–60.00 g/L for DEHP and 0.50– 60.00 g/L for DOP were used. Standard curves were plotted as the analyte peak areas versus their concentrations with the data obtained from ten replicate analyses. The linear regression model of least-squares with weighting factor was used for the calibration and analysis of the results (LabSolutions, Version 1.25). 2.5 USVADLLME At first, the extraction of PAEs from their standard solutions of 2 g/L was tested with the USVADLLME method. Five milliliters of the standard solution was placed into a screwcap glass test tube with conical bottom. Two milliliters of MeOH and 300 L of CHCl3 were added as the disperser and the extraction solvent, respectively. Then, 1.0 g of NaCl was added for the salting out effect. The tube was placed into an Elma S15 ultrasonic bath (Elma Hans Schmidbauer, Siegen, Germany) for 30 s and then vortexed (Mixer UZUSIO VTX-3000L, Harmony, Tokyo, Japan) for 5 min. The tube was immediately centrifuged (VWR Compactstar CS4, VWR International, Leicestershire, UK) for 3 min at 4000 rpm, and then the chloroform phase was transferred into a glass test tube with a Hamilton microsyringe (Hamilton Bonaduz AG, Bonaduz, Switzerland). The solvent was evaporated under a N2 flow. The residue was dissolved with the mobile phase of the developed HPLC method and then filtered through 0.45 m syringe filter (Lubitech Technologies, China) before injection. In the extraction of the analytes from the beverage samples, these conditions were not suitable for a high recovery percentage, so that 2.5 mL of sample, 500 L of CHCl3, C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Liquid Chromatography
2113
and 0.5 g NaCl were used instead. The ultrasonication, vortex, and centrifugation times were kept the same. The standard solutions, original beverage samples, and spiked beverage samples were extracted in three replicates to check the repeatability of the developed method.
3 Results and discussion 3.1 Optimization of USVADLLME Several parameters as type and volume of the extraction solvent, volume of the disperser solvent, concentration of NaCl, ultrasonication, and vortex time were investigated to optimize the USVADLLME procedure. Extraction recovery was examined to determine the optimum conditions. 3.1.1 Selection of the extraction solvent type and volume The extraction solvent had to have a higher or lower density than the aqueous phase, low solubility in the aqueous phase, high extraction efficiency, and good chromatographic behavior. Also studies in the literature indicated that the chlorinated solvents had high extraction capabilities for PAEs from aqueous solutions [7,20]. Considering these criteria, three different chlorinated solvents, CH2 Cl2 , CHCl3, and CCl4 , were examined for the extraction of PAEs from the standard solution at 2.0 g/L (Supporting Information Fig. S1). CHCl3 had the highest recovery for all the PAEs. Different volumes of CHCl3 (100, 300, 500 L) were used to optimize the volume of the extraction solvent, and 300 L was selected since it provided sufficient recovery. A higher volume of CHCl3 had almost the same extraction performance (Supporting Information Table S2). The same volume of CHCl3 was used for the extraction of beverage samples but due to different and complex matrices of the samples, the extraction recoveries were low (results not shown here), so that 2.5 mL of sample and 500 L of CHCl3 was used for the extraction of PAEs from beverage samples. 3.1.2 Selection of the disperser solvent volume The disperser solvent provided the formation of the cloudy solution of the immiscible solvents (water and CHCl3 ). The volume of the disperser solvent had to be optimized since it determined the degree of dispersion. MeOH was selected as the disperser solvent considering its high dispersing property in CHCl3 /water mixture. Different volumes of MeOH (50, 100, 150 L) were examined (Supporting Information Table S3). The optimum disperser solvent volume was selected as 100 L, since lower volume was not enough for a complete dispersion and higher volume provided similar results. 3.1.3 Selection of the NaCl concentration NaCl was used to decrease the solubility of the PAEs in the aqueous phase (salting out effect). Two hundred www.jss-journal.com
2114
J. Sep. Sci. 2014, 37, 2111–2117
P. K. Yılmaz et al.
Figure 2. A representative chromatogram of the determined PAEs (standard solutions at 10 g/L).
grams NaCl/L was determined as the optimum concentration. Higher concentration provided almost the same effect (Supporting Information Table S4). 3.1.4 Selection of the ultrasonication and vortex time The effect of the ultrasonication (Supporting Information Table S5) and vortex time (Supporting Information Table S6) on the extraction recovery was investigated by different time ranges (ultrasonication time of 0, 10, 30, 60 s, and vortex time of 0, 1, 5, 10 min). Recoveries above 90% were obtained by using a ultrasonication time of 30 s with a vortex time of 5 min. Longer procedure times provided almost the same results.
3.2 Selection of HPLC conditions Different mobile phases and gradient systems were tested for the appropriate separation of the PAEs. All of these mobile phase systems were consisted of two parts as eluent A and B (Supporting Information Table S7). The mobile phase system
of A: 0.2M KH2 PO4 buffer (pH 2.6), B: 50:50 MeOH/ACN was selected considering a lower baseline drift and higher resolution. The analyte peaks were detected at 230, 256, and 280 nm. The highest analyte signals were obtained at 230 nm, and it was selected as the detection wavelength.
3.3 Method validation The developed method was tested in terms of linearity, LOD, LOQ, repeatability, accuracy, and recovery. A representative chromatogram of seven PAEs at 10 g/L is shown in Fig. 2. The linearity of the developed method was determined by eight-point calibration curves for DMP, DPP, DBP, BBP, and DCHP, a seven-point calibration curve for DEHP, and a sixpoint calibration curve for DOP. The homoscedasticity of the calibration curves was tested by the F test (Supporting Information Table S8). All of the experimental F values were higher than F values from the F table at the confidence level of 99% for (n − 1) degrees of freedom (9, 9, 0.99). In the light of evidence of the heteroscedastic situation, the weighting factors were determined. The best weighting factors were
Table 1. Analytical performance of the developed method
Analytes
Calibration range (g/L)
tR (min)a)
DMP DPP DBP BBP DCHP DEHP DOP
0.05 – 60.00 0.05 – 60.00 0.05 – 60.00 0.05 – 60.00 0.05 – 60.00 0.10 – 60.00 0.50 – 60.00
4.315 6.622 7.615 7.945 9.763 13.976 15.198
± ± ± ± ± ± ±
0.003 0.005 0.007 0.008 0.009 0.016 0.042
Linear equation
r
LOD (g/L)
LOQ (g/L)
Resolution
Tailing factor
y = 34592.01 x + 737.2784 y = 31776.32x – 312.6182 y = 34568.99x + 468.8576 y = 39829.60x + 61.4829 y = 22522.20x + 28.6793 y = 25256.97x + 2983.6170 y = 17432.84x – 150.2135
0.9987 0.9973 0.9994 0.9981 0.9981 0.9991 0.9983
0.019 0.027 0.021 0.021 0.043 0.026 0.208
0.072 0.093 0.078 0.076 0.149 0.085 0.483
– 14.192 ± 6.375 ± 2.079 ± 10.698 ± 19.177 ± 3.190 ±
1.158 ± 0.006 1.119 ± 0.003 1.120 ± 0.003 1.106 ± 0.004 1.073 ± 0.003 1.032 ± 0.014 1.204 ± 0.052
0.151 0.081 0.024 0.126 0.225 0.021
a) Mean ± SD, n = 10
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 2111–2117
2115
Table 2. Repeatability (interday assays), intermediate precision (intraday assays), and recovery of the developed method
DMP Interday (n = 3)a)
Intraday (n = 3)
Recovery (%)b) (n = 3)
DPP
DBP
BBP
DCHP
DEHP
DOP
Concentration (g/L) 0.20 0.20 ± 0.00 4.00 3.68 ± 0.01 24.00 23.66 ± 0.01
0.18 ± 0.00 3.78 ± 0.02 24.52 ± 0.02
0.20 ± 0.01 3.68 ± 0.02 23.92 ± 0.02
0.19 ± 0.01 3.76 ± 0.02 24.42 ± 0.01
0.18 ± 0.01 3.74 ± 0.03 24.36 ± 0.03
0.20 ± 0.00 3.66 ± 0.06 23.98 ± 0.27
0.22 ± 0.01 3.42 ± 0.13 24.14 ± 0.04
Concentration (g/L) 0.20 0.21 ± 0.00 4.00 3.69 ± 0.01 24.00 23.81 ± 0.00
0.23 ± 0.01 3.83 ± 0.00 24.61 ± 0.00
0.21 ± 0.00 3.71 ± 0.01 23.88 ± 0.02
0.20 ± 0.01 3.76 ± 0.03 24.40 ± 0.03
0.21 ± 0.02 3.75 ± 0.02 24.42 ± 0.02
0.21 ± 0.00 3.76 ± 0.03 24.32 ± 0.02
0.18 ± 0.00 3.56 ± 0.14 23.72 ± 0.14
Concentration (g/L) 0.20 105.79 ± 0.14 4.00 107.10 ± 2.01 24.00 99.03 ± 0.03
116.90 ± 0.49 101.20 ± 0.48 102.56 ± 0.00
103.95 ± 3.09 101.00 ± 1.86 99.15 ± 0.13
95.16 ± 4.58 101.47 ± 1.78 101.64 ± 0.09
103.87 ± 0.46 103.41 ± 2.35 108.81 ± 6.50
95.41 ± 0.70 106.36 ± 1.47 112.65 ± 0.25
88.83 ± 0.81 95.76 ± 1.07 95.05 ± 0.34
a) Mean of the standard solutions ± SD b) Recovery from standard solutions ± SD Table 3. Recovery% of PAEs from the spiked beverage samplesa)
Samples
Analytes DMP
LJ-1 LJ-2 IT-1 IT-2 CJ-1 CJ-2 V-1 V-2 TJ LS-1 LS-2 MW C-1 C-2 S-1 S-2 SD
77.42 73.51 86.68 102.08 106.57 94.64 86.59 109.38 89.50 91.75 95.06 77.97 99.23 79.53 46.15 69.64 84.56
DPP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
2.55b) 3.39 1.09 0.09 2.51 0.97 4.59 3.76 6.61 4.69 5.44 0.04 2.11 4.70 4.46 2.98 0.07
90.50 86.87 94.23 104.54 112.60 114.45 107.68 121.56 107.86 106.27 109.66 115.16 116.35 95.72 58.68 88.16 118.77
DBP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1.14 5.51 1.52 1.39 1.76 4.03 5.68 4.07 3.17 5.49 6.12 0.39 5.06 5.58 5.68 3.77 0.27
86.47 79.59 87.11 101.83 105.13 105.87 100.27 112.16 99.30 95.75 90.12 99.80 97.55 88.12 55.40 83.23 114.80
BBP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
3.81 5.46 1.09 5.31 1.72 4.36 4.46 3.77 1.52 2.13 6.60 4.72 1.26 5.46 5.10 3.58 0.60
90.55 83.15 90.82 99.20 109.29 109.42 100.34 115.60 105.02 100.93 96.17 103.45 112.88 90.87 55.35 84.80 117.68
DCHP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
3.54 5.42 1.48 0.06 1.83 3.99 5.41 4.52 2.15 1.86 5.45 5.50 0.34 5.93 4.02 3.62 0.59
90.47 94.21 90.95 100.20 108.58 111.22 105.96 117.83 107.12 100.54 96.97 110.94 93.88 92.18 57.57 85.52 115.93
DEHP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1.68 3.97 1.69 0.04 1.70 4.21 5.19 4.12 2.57 1.46 6.50 0.47 3.97 5.77 5.47 3.62 1.09
85.01 86.09 93.34 103.51 113.00 107.03 103.60 116.90 104.84 94.11 86.46 110.63 91.96 94.74 58.05 87.36 112.67
DOP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1.29 4.53 2.86 0.16 2.60 4.91 5.15 2.42 3.24 1.30 3.25 0.16 0.95 6.09 5.76 3.76 3.19
78.37 78.13 84.04 92.82 105.89 104.00 101.80 103.88 97.71 104.34 98.83 103.99 97.59 89.90 54.47 83.32 101.74
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
4.73 3.87 2.66 0.12 1.91 3.99 5.50 3.57 4.11 3.99 5.74 3.38 1.73 6.48 5.62 3.52 5.49
a) Samples were spiked at 4.00 g/L. b) Mean ± SD
chosen considering the lowest sum of the relative error percentages (RE%) across the whole concentration ranges (Supporting Information Table S9). The calibration equations and correlation coefficients (r) were calculated by linear regression analysis based on least squares method with determined weighting factors. A good linearity was obtained with r values higher than 0.99 (Table 1). LODs and LOQs were calculated from the data of the standard solutions. S/N ratios were accepted as 3 and 10 for the determination of LODs and LOQs, respectively. The LODs and LOQs of seven PAEs were in the range of 0.019–0.208 and 0.072–0.483 g/L, respectively. The linear ranges, calibration equations, correlation coefficients, LODs, LOQs, resolutions, and tailing factors are given in Table 1. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The repeatability and the intermediate precision of the developed method were determined by standard solutions at low, middle, and high concentrations (0.20, 4.00, and 24.00 g/L). Three replicates of each standard solutions were injected consecutively in one day, and then in three separate analytical runs in three different days (Table 2). The precision was calculated as RSD%, and the accuracy was determined in terms of relative error ((observed concentration − nominal concentration)/nominal concentration × 100) (Supporting Information Table S10). All of the precision and accuracy results were lower than 20%, which is the acceptable limit for bioanalytical analysis [28]. The recovery of the USVADLLME method was determined as the percentage of the ratio of observed concentration www.jss-journal.com
2116
J. Sep. Sci. 2014, 37, 2111–2117
P. K. Yılmaz et al.
to nominal concentration of the standard solutions at concentration levels of 0.20, 4.00, and 24.00 g/L. The recovery of PAEs was between 95.05 and 116.90% (Table 2). In the literature, there are several HPLC–UV methods for the quantification of PAEs. Guo et al. developed a HPLC– UV method for the detection of six PAEs (DMP, DBP, DEP, DEHP, BBP, and DOP) in orange juice samples with a total analysis time of 28 min [29]. In the proposed study, it was aimed to develop a reliable, simple, sensitive, and less timeconsuming method that could be applied to real beverage samples. This study had the advantages of separation of seven PAEs (DMP, DPP, DBP, BBP, DCHP, DEHP, and DOP) with acceptable resolution values in 20 min with high precision and accuracy.
was easy to perform, effective in terms of extraction recovery, and suitable for various beverage matrices. The developed method enabled the quantification of seven PAEs in a total analysis time of 20 min with high sensitivity, precision, accuracy, and a wide linear range. The results indicated the requirement of determination and prevention of PAE contamination sources. This study was supported by the Research Fund of Istanbul University (Project number:40324). The authors have declared no conflict of interest.
5 References
3.4 Application of the developed method to the beverage samples
[1] Cao, X.-L., Compr. Rev. Food Sci. F. 2010, 9, 21–43.
The developed method was used for the quantification of seven PAEs in 17 beverage samples stored in plastic bottles. The extraction procedure was replicated three times for each of the original and spiked (4.00 g/L) beverage samples. All of the samples were analyzed in triplicate by HPLC–UV. The matrix effect was different for different kinds of PAEs and beverage samples. It could be concluded that all of the recovery values were higher than 80%, except for soda sample (Supporting Information Fig. S1 and Table 3). DEHP, which has been one of the most widely detected PAEs in foodstuffs [6, 7, 30], was determined in eight of the tested beverage samples within 0.09–1.44 g/L. DMP was detected in cherry juice samples (CJ-1 and CJ-2), vinegar sample (V-1), and sports drink sample (SD) at 0.14, 0.18, 0.18, and 0.02 g/L, respectively. In lemon juice sample (LJ-1), DPP was detected at 0.07 g/L. DCHP was determined only in SD at a concentration of 0.13 g/L. In none of the samples, DBP, BBP, and DOP were detected above the LOQ values (Supporting Information Table S11). Some researchers also reported the contamination of various kinds of beverages by PAEs. Khedr detected DEHP in water, soda, coke, and energy drink samples from Saudi Arabia [30]. In another study, DBP and DEHP were determined in 16 Chinese liquor samples [22]. Also a migration study of DMP, DEP, BBP, DBP, DEHP, and DOP to orange juice samples from China was performed by Guo et al. [29]. Only DEP and DEHP were detected with a storage time of three months, and their concentrations were found to be increased when the expiration date arrived.
[3] Phthalates, TEACH Chemical Summary, Toxicity and Exposure Assessment for Children’s Health, U.S. EPA, 2007.
4 Concluding remarks To the best of our knowledge, this study could be considered as the first report on PAEs analysis in beverage samples from Turkey. In this work, a USVADLLME–HPLC–UV method was developed for the quantification of seven PAEs in different kinds of beverage samples. This USVADLLME procedure C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[2] Ranjbari, E., Hadjmohammadi, M. R., Talanta 2012, 100, 447–453.
[4] Montuori, P., Jover, E., Morgantini, M., Bayona, J. M., Triassi, M., Food Addit. Contam. Part A 2008, 25, 511– 518. [5] LaFleur, A. D., Schug, K. A., Anal. Chim. Acta 2011, 696, 6–26. [6] Xu, D., Deng, X., Fang, E., Zheng, X., Zhou, Y., Lin, L., Chen, L., Wu, M., Huang, Z., J. Chromatogr. A 2014, 1324, 49–56. [7] Cinelli, G., Avino, P., Notardonato, I., Centola, A., Russo, M. V., Anal. Chim. Acta 2013, 769, 72–78. [8] European Food Safety Authority, EFSA J. 2005, 242, 1– 17. [9] European Food Safety Authority, EFSA J. 2005, 241, 1– 14. [10] European Food Safety Authority, EFSA J. 2005, 243, 1– 20. [11] European Food Safety Authority, EFSA J. 2005, 244, 1– 18. [12] European Food Safety Authority, EFSA J. 2005, 245, 1– 14. [13] Cai, Y., Cai, Y., Shi, Y., Liu, J., Mou, S., Lu, Y., Microchim. Acta 2007, 157, 73–79. [14] Yano, K., Hirosawa, N., Sakamoto, Y., Katayama, H., Moriguchi, T., Joung, K. E., Sheen, Y. Y., Asaoka, K., Bull. Environ. Contam. Toxicol. 2002, 68, 463–469. [15] Tsumura, Y., Ishimitsu, S., Saito, I., Sakai, H., Kobayashi, Y., Tonogai, Y., Food Addit. Contam. 2001, 18, 449–460. [16] Xu, Q., Yin, X., Wang, M., Wang, H., Zhang, N., Shen, Y., Xu, S., Zhang, L., Gu, Z., J. Agric. Food Chem. 2010, 58, 11311–11317. [17] Rezaee, M., Assadi, Y., Hosseini, M. R. M., Aghaee, E., Ahmadi, F., Berijani, S., J. Chromatogr. A 2006, 1116, 1– 9. [18] Mousa, A., Basheer, C., Al-Arfaj, A. R., J. Sep. Sci. 2013, 36, 2003–2009. [19] Sha, C., Yi-Sheng, Z., Shui-Yuan, C., Tian, Q., Hao, S., J. Sep. Sci. 2011, 34, 1503–1507.
www.jss-journal.com
J. Sep. Sci. 2014, 37, 2111–2117
[20] Hadjmohammadi, M. R., Ranjbari, E., Int. J. Environ. Anal. Chem. 2012, 92, 1312–1324. [21] Guo, L., Dong, H., Int. J. Environ. Anal. Chem. 2009, 89, 357–365. [22] Leng, G., Chen, W., Zhang, M., Huang, F., Cao, Q., J. Sep. Sci. 2014, 37, 684–690. [23] Zhou, J., Qi, Y., Wu, H., Diao, Q., Tian, F., Li, Y., J. Sep. Sci. 2014, 37, 650–657. [24] Farajzadeh, M. A., Sheykhizadeh, S., Khorram, P., J. Sep. Sci. 2013, 36, 939–946. [25] Hongyuan, Y., Xiaoling, C., Kuo, Y., Analyst 2012, 137, 4860–4866.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Liquid Chromatography
2117
´ ´ C., Barciela Alonso, M. C., Pena-V ˜ ´ [26] Perez Feas, azquez, E., Herbello Hermelo, P., Bermejo-Barrera, P., Talanta 2008, 75, 1184–1189. [27] Li, X., Xiong, W., Lin, H., Zhuo, L., Lv, S., Tang, X., Chen, M., Zou, Z., Lin, Z., Qiu, B., Chen, G., J. Sep. Sci. 2013, 36, 477–484. [28] US Department of Health and Human Services, Guidance for Industry, Bioanalytical Method Validation, FDA, CDER, Rockville, USA 2001. [29] Guo, Z., Wei, D., Wang, M., Wang, S., J. Chromatogr. Sci. 2010, 48, 760–765. [30] Khedr, A., J. Chromatogr. B 2013, 930, 98–103.
www.jss-journal.com
J. Sep. Sci. 2014, 37, 2111–2117
¨ ˘ Yılmaz1 Pelin Koseo glu Abdulselam Ertas¸2 Ufuk Kolak1 1 Department
of Analytical Chemistry, Faculty of Pharmacy, Istanbul University, Istanbul, Turkey 2 Department of Pharmacognosy, Faculty of Pharmacy, Dicle University, Diyarbakir, Turkey Received April 14, 2014 Revised May 13, 2014 Accepted May 18, 2014
Research Article
Simultaneous determination of seven phthalic acid esters in beverages using ultrasound and vortex-assisted dispersive liquid–liquid microextraction followed by high-performance liquid chromatography A sensitive, rapid, and simple high-performance liquid chromatography with UV detection method was developed for the simultaneous determination of seven phthalic acid esters (dimethyl phthalate, dipropyl phthalate, di-n-butyl phthalate, benzyl butyl phthalate, dicyclohexyl phthalate, di-(2-ethylhexyl) phthalate, and di-n-octyl phthalate) in several kinds of beverage samples. Ultrasound and vortex-assisted dispersive liquid–liquid microextraction method was used. The separation was performed using an Intersil ODS-3 column (C18 , 250 × 4.6 mm, 5.0 m) and a gradient elution with a mobile phase consisting of MeOH/ACN (50:50) and 0.2 M KH2 PO4 buffer. Analytes were detected by a UV detector at 230 nm. The developed method was validated in terms of linearity, limit of detection, limit of quantification, repeatability, accuracy, and recovery. Calibration equations and correlation coefficients (> 0.99) were calculated by least squares method with weighting factor. The limit of detection and quantification were in the range of 0.019–0.208 and 0.072–0.483 g/L. The repeatability and intermediate precision were determined in terms of relative standard deviation to be within 0.03–3.93 and 0.02–4.74%, respectively. The accuracy was found to be in the range of –14.55 to 15.57% in terms of relative error. Seventeen different beverage samples in plastic bottles were successfully analyzed, and ten of them were found to be contaminated by different phthalic acid esters. Keywords: Beverages / High-performance liquid chromatography / Liquid–liquid microextraction / Phthalic acid esters DOI 10.1002/jssc.201400408
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Phthalic acid esters (PAEs) are a class of dialkyl or alkyl aryl esters of 1,2-benzenedicarboxylic acid. They are widely used in food packaging materials, personal care products, children’s toys, school supplies, household cleaners, glues, paints, automobile parts, home d´ecor, and building materials for flexibilCorrespondence: Prof. Ufuk Kolak, Department of Analytical Chemistry, Faculty of Pharmacy, Istanbul University, 34116 Istanbul, Turkey E-mail: [email protected]; [email protected] Fax: +902124400254
Abbreviations: BBP, benzyl butyl phthalate; DBP, di-nbutyl phthalate; DCHP, dicyclohexyl phthalate; DEHP, di(2-ethylhexyl) phthalate; DLLME, dispersive liquid–liquid microextraction; DMP, dimethyl phthalate; DOP, di-n-octyl phthalate; DPP, dipropyl phthalate; PAE, phthalic acid ester; USVADLLME, ultrasound and vortex-assisted dispersive liquid–liquid microextraction
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ity, or to hold color and scent due to their excellent properties and compatibility with polymers [1–3]. All products may be contaminated with PAEs during production, manufacture, storage, and/or usage since there is no covalent bond between the PAEs and polymer chains. The migration of PAEs from plastic packages into the foods and beverages is one of the main human exposure sources. On the other hand, PAEs can also be easily transferred from a product onto the skin and released into the air. Since research on animals indicated that PAEs possessed some carcinogenic effects [2] and also caused fetal defects, uterine damage, disturbances in the male reproductive tract, and endocrine system disrupting activities [4, 5], some countries are prohibiting the usage of certain PAEs in plastic materials [6]. As Cinelli et al. reported, any limits of PAEs for foods and beverages have not been determined [7]. On the other hand, the European Food Safety Authority (EFSA) recommended the tolerable daily intakes of dibutyl phthalate (DBP), benzyl butyl phthalate (BBP), di-2-ethylhexyl phthalate (DEHP),
www.jss-journal.com
2112
J. Sep. Sci. 2014, 37, 2111–2117
P. K. Yılmaz et al.
Figure 1. Chemical structures of the determined PAEs.
di-isononyl phthalate (DiNP), and di-isodecyl phthalate (DiDP) as 0.01, 0.50, 0.05, 0.15, and 0.15 mg/kg body weight/day, respectively [8–12]. Considering the negative impacts of PAEs on human health, development of reliable analytical methods for the determination of PAEs is an important challenge. Because of the low concentration levels and the effects of complex sample matrices, different types of pretreatment techniques have been developed for the preconcentration of the PAEs. LLE [13–15] and SPE [2, 16] are the traditional pretreatment methods, but they have some disadvantages such as requirement of large volumes of sample and extraction solvent and are time consuming. Dispersive liquid–liquid microextraction (DLLME) is one of the recently developed techniques to overcome those disadvantages. DLLME is based on the extraction of the analyte by using an immiscible extracting solvent and a disperser solvent. The disperser solvent is soluble in both of the liquid phases and increases the contact between the two immiscible liquids [17]. Mousa et al. determined PAEs in bottled water samples using DLLME coupled with GC–MS [18]. The extracting solvents used for this purpose are generally chlorinated ones with serious toxic effects on human health. Recently ultrasound and vortex-assisted dispersive liquid–liquid microextraction (USVADLLME) procedures were developed to reduce the required volume of those hazardous extracting solvents. Ultrasonication provides a dispersed phase for the quantitative extraction of the analyte, and vortex prevents the formation of a biphasic system [7]. The detection of PAEs is generally performed by GC or HPLC. HPLC–UV is one of the commonly used techniques for determination of PAEs [19–21] followed by GC–MS [6, 22, 23], GC with flame ionization detection (FID) [24,25], and HPLC–MS [26, 27]. In this work, we aimed to develop a sensitive, rapid, and simple method for determination of seven PAEs (dimethyl phthalate (DMP), dipropyl phthalate (DPP), DBP, BBP, dicyclohexyl phthalate (DCHP), DEHP, and di-n-octyl phthalate (DOP)) in beverage samples using USVADLLME followed by HPLC–UV (Fig. 1). The developed method was validated in terms of linearity, LOD, LOQ, repeatability, accuracy, and recovery. Then the proposed method was used to detect the PAE contamination levels of 17 different beverage samples from Istanbul (Turkey). To the best of our knowledge, this
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
study could be considered as the first report on detection of PAEs in beverage samples from Turkey.
2 Materials and methods 2.1 Reagents and standards The standards of DMP, DPP, DBP, BBP, DCHP, DEHP, and DOP were purchased from Sigma (Darmstadt, Germany). The stock solution that contained each of seven PAEs at a concentration of 1000 mg/L was prepared with HPLCgrade methanol. The standard solutions were prepared daily by dilution with HPLC-grade methanol to desired concentrations. Methanol (MeOH) and acetonitrile (ACN) (HPLCgrade), dichloromethane (CH2 Cl2 ), chloroform (CHCl3 ), carbon tetrachloride (CCl4 ), sodium chloride (NaCl), and potassium dihydrogen phosphate KH2 PO4 were purchased from Merck (Darmstadt, Germany). All of the glassware used was washed with chloroform and dried at 90⬚C to avoid any contamination of PAEs. A blank analysis was also performed to check the purity of the chemicals used in terms of PAE contamination.
2.2 Beverage samples The beverages in plastic bottles including lemon juice (LJ-1 and LJ-2), ice tea (IT-1 and IT-2), cherry juice (CJ-1 and CJ-2), vinegar (V-1 and V-2), turnip juice (TJ), lemon sauce (LS-1 and LS-2), mineral water (MW), coke (C-1 and C-2), soda (S-1 and S-2), and sports drink (SD) samples were purchased from a local market in Istanbul (Turkey) in 2013. All samples were kept at +4⬚C.
2.3 Instruments and analytical conditions A Shimadzu (Shimadzu, Kyoto, Japan) LC20AT HPLC system with UV detection was used for the quantitative analysis of PAEs. The separation of PAEs was accomplished with a GL Sciences (GL Sciences, Tokyo, Japan) Intersil ODS-3 column
www.jss-journal.com
J. Sep. Sci. 2014, 37, 2111–2117
(C18 , 250 × 4.6 mm, 5.0 m). The data obtained were analyzed by the LabSolutions software (version 1.25). A gradient program with a mobile phase system consisting of two parts as eluate A (0.2 M KH2 PO4 buffer, pH 2.6) and eluate B (MeOH/ACN 50:50, v/v) was established for the elution of PAEs. After the stabilization of the system with 75% eluent B flow, a linear gradient from 0.0 to 3.0 min from 75 to 90% of eluent B flow, from 3.0 to 5.0 min from 90 to 95% of eluent B flow, from 5.0 to 9.0 min from 95 to 100% of eluent B flow, from 9.0 to 11.0 min from 100 to 95% of eluent B flow, from 11.0 to 16.0 min from 95 to 75% of eluent B flow, and then for stabilization of the system an isocratic hold from 16.0 to 20.0 min at 75% eluent B flow were applied (Supporting Information Table S1). The total analysis time was 20 min for seven PAEs. The flow rate was set to 1 mL/min and the injection volume was 20 L. The column temperature was adjusted to 40⬚C. 2.4 Quantification PAEs were identified by comparing their retention times with those of the ones in the beverage samples and with the increase of the peak areas after spiking the beverage samples. The quantification was performed by the external standard method. The calibration curves were prepared in the concentration range of 0.05–60.00 g/L for DMP, DPP, DBP, BBP, and DCHP, whereas 0.10–60.00 g/L for DEHP and 0.50– 60.00 g/L for DOP were used. Standard curves were plotted as the analyte peak areas versus their concentrations with the data obtained from ten replicate analyses. The linear regression model of least-squares with weighting factor was used for the calibration and analysis of the results (LabSolutions, Version 1.25). 2.5 USVADLLME At first, the extraction of PAEs from their standard solutions of 2 g/L was tested with the USVADLLME method. Five milliliters of the standard solution was placed into a screwcap glass test tube with conical bottom. Two milliliters of MeOH and 300 L of CHCl3 were added as the disperser and the extraction solvent, respectively. Then, 1.0 g of NaCl was added for the salting out effect. The tube was placed into an Elma S15 ultrasonic bath (Elma Hans Schmidbauer, Siegen, Germany) for 30 s and then vortexed (Mixer UZUSIO VTX-3000L, Harmony, Tokyo, Japan) for 5 min. The tube was immediately centrifuged (VWR Compactstar CS4, VWR International, Leicestershire, UK) for 3 min at 4000 rpm, and then the chloroform phase was transferred into a glass test tube with a Hamilton microsyringe (Hamilton Bonaduz AG, Bonaduz, Switzerland). The solvent was evaporated under a N2 flow. The residue was dissolved with the mobile phase of the developed HPLC method and then filtered through 0.45 m syringe filter (Lubitech Technologies, China) before injection. In the extraction of the analytes from the beverage samples, these conditions were not suitable for a high recovery percentage, so that 2.5 mL of sample, 500 L of CHCl3, C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Liquid Chromatography
2113
and 0.5 g NaCl were used instead. The ultrasonication, vortex, and centrifugation times were kept the same. The standard solutions, original beverage samples, and spiked beverage samples were extracted in three replicates to check the repeatability of the developed method.
3 Results and discussion 3.1 Optimization of USVADLLME Several parameters as type and volume of the extraction solvent, volume of the disperser solvent, concentration of NaCl, ultrasonication, and vortex time were investigated to optimize the USVADLLME procedure. Extraction recovery was examined to determine the optimum conditions. 3.1.1 Selection of the extraction solvent type and volume The extraction solvent had to have a higher or lower density than the aqueous phase, low solubility in the aqueous phase, high extraction efficiency, and good chromatographic behavior. Also studies in the literature indicated that the chlorinated solvents had high extraction capabilities for PAEs from aqueous solutions [7,20]. Considering these criteria, three different chlorinated solvents, CH2 Cl2 , CHCl3, and CCl4 , were examined for the extraction of PAEs from the standard solution at 2.0 g/L (Supporting Information Fig. S1). CHCl3 had the highest recovery for all the PAEs. Different volumes of CHCl3 (100, 300, 500 L) were used to optimize the volume of the extraction solvent, and 300 L was selected since it provided sufficient recovery. A higher volume of CHCl3 had almost the same extraction performance (Supporting Information Table S2). The same volume of CHCl3 was used for the extraction of beverage samples but due to different and complex matrices of the samples, the extraction recoveries were low (results not shown here), so that 2.5 mL of sample and 500 L of CHCl3 was used for the extraction of PAEs from beverage samples. 3.1.2 Selection of the disperser solvent volume The disperser solvent provided the formation of the cloudy solution of the immiscible solvents (water and CHCl3 ). The volume of the disperser solvent had to be optimized since it determined the degree of dispersion. MeOH was selected as the disperser solvent considering its high dispersing property in CHCl3 /water mixture. Different volumes of MeOH (50, 100, 150 L) were examined (Supporting Information Table S3). The optimum disperser solvent volume was selected as 100 L, since lower volume was not enough for a complete dispersion and higher volume provided similar results. 3.1.3 Selection of the NaCl concentration NaCl was used to decrease the solubility of the PAEs in the aqueous phase (salting out effect). Two hundred www.jss-journal.com
2114
J. Sep. Sci. 2014, 37, 2111–2117
P. K. Yılmaz et al.
Figure 2. A representative chromatogram of the determined PAEs (standard solutions at 10 g/L).
grams NaCl/L was determined as the optimum concentration. Higher concentration provided almost the same effect (Supporting Information Table S4). 3.1.4 Selection of the ultrasonication and vortex time The effect of the ultrasonication (Supporting Information Table S5) and vortex time (Supporting Information Table S6) on the extraction recovery was investigated by different time ranges (ultrasonication time of 0, 10, 30, 60 s, and vortex time of 0, 1, 5, 10 min). Recoveries above 90% were obtained by using a ultrasonication time of 30 s with a vortex time of 5 min. Longer procedure times provided almost the same results.
3.2 Selection of HPLC conditions Different mobile phases and gradient systems were tested for the appropriate separation of the PAEs. All of these mobile phase systems were consisted of two parts as eluent A and B (Supporting Information Table S7). The mobile phase system
of A: 0.2M KH2 PO4 buffer (pH 2.6), B: 50:50 MeOH/ACN was selected considering a lower baseline drift and higher resolution. The analyte peaks were detected at 230, 256, and 280 nm. The highest analyte signals were obtained at 230 nm, and it was selected as the detection wavelength.
3.3 Method validation The developed method was tested in terms of linearity, LOD, LOQ, repeatability, accuracy, and recovery. A representative chromatogram of seven PAEs at 10 g/L is shown in Fig. 2. The linearity of the developed method was determined by eight-point calibration curves for DMP, DPP, DBP, BBP, and DCHP, a seven-point calibration curve for DEHP, and a sixpoint calibration curve for DOP. The homoscedasticity of the calibration curves was tested by the F test (Supporting Information Table S8). All of the experimental F values were higher than F values from the F table at the confidence level of 99% for (n − 1) degrees of freedom (9, 9, 0.99). In the light of evidence of the heteroscedastic situation, the weighting factors were determined. The best weighting factors were
Table 1. Analytical performance of the developed method
Analytes
Calibration range (g/L)
tR (min)a)
DMP DPP DBP BBP DCHP DEHP DOP
0.05 – 60.00 0.05 – 60.00 0.05 – 60.00 0.05 – 60.00 0.05 – 60.00 0.10 – 60.00 0.50 – 60.00
4.315 6.622 7.615 7.945 9.763 13.976 15.198
± ± ± ± ± ± ±
0.003 0.005 0.007 0.008 0.009 0.016 0.042
Linear equation
r
LOD (g/L)
LOQ (g/L)
Resolution
Tailing factor
y = 34592.01 x + 737.2784 y = 31776.32x – 312.6182 y = 34568.99x + 468.8576 y = 39829.60x + 61.4829 y = 22522.20x + 28.6793 y = 25256.97x + 2983.6170 y = 17432.84x – 150.2135
0.9987 0.9973 0.9994 0.9981 0.9981 0.9991 0.9983
0.019 0.027 0.021 0.021 0.043 0.026 0.208
0.072 0.093 0.078 0.076 0.149 0.085 0.483
– 14.192 ± 6.375 ± 2.079 ± 10.698 ± 19.177 ± 3.190 ±
1.158 ± 0.006 1.119 ± 0.003 1.120 ± 0.003 1.106 ± 0.004 1.073 ± 0.003 1.032 ± 0.014 1.204 ± 0.052
0.151 0.081 0.024 0.126 0.225 0.021
a) Mean ± SD, n = 10
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 2111–2117
2115
Table 2. Repeatability (interday assays), intermediate precision (intraday assays), and recovery of the developed method
DMP Interday (n = 3)a)
Intraday (n = 3)
Recovery (%)b) (n = 3)
DPP
DBP
BBP
DCHP
DEHP
DOP
Concentration (g/L) 0.20 0.20 ± 0.00 4.00 3.68 ± 0.01 24.00 23.66 ± 0.01
0.18 ± 0.00 3.78 ± 0.02 24.52 ± 0.02
0.20 ± 0.01 3.68 ± 0.02 23.92 ± 0.02
0.19 ± 0.01 3.76 ± 0.02 24.42 ± 0.01
0.18 ± 0.01 3.74 ± 0.03 24.36 ± 0.03
0.20 ± 0.00 3.66 ± 0.06 23.98 ± 0.27
0.22 ± 0.01 3.42 ± 0.13 24.14 ± 0.04
Concentration (g/L) 0.20 0.21 ± 0.00 4.00 3.69 ± 0.01 24.00 23.81 ± 0.00
0.23 ± 0.01 3.83 ± 0.00 24.61 ± 0.00
0.21 ± 0.00 3.71 ± 0.01 23.88 ± 0.02
0.20 ± 0.01 3.76 ± 0.03 24.40 ± 0.03
0.21 ± 0.02 3.75 ± 0.02 24.42 ± 0.02
0.21 ± 0.00 3.76 ± 0.03 24.32 ± 0.02
0.18 ± 0.00 3.56 ± 0.14 23.72 ± 0.14
Concentration (g/L) 0.20 105.79 ± 0.14 4.00 107.10 ± 2.01 24.00 99.03 ± 0.03
116.90 ± 0.49 101.20 ± 0.48 102.56 ± 0.00
103.95 ± 3.09 101.00 ± 1.86 99.15 ± 0.13
95.16 ± 4.58 101.47 ± 1.78 101.64 ± 0.09
103.87 ± 0.46 103.41 ± 2.35 108.81 ± 6.50
95.41 ± 0.70 106.36 ± 1.47 112.65 ± 0.25
88.83 ± 0.81 95.76 ± 1.07 95.05 ± 0.34
a) Mean of the standard solutions ± SD b) Recovery from standard solutions ± SD Table 3. Recovery% of PAEs from the spiked beverage samplesa)
Samples
Analytes DMP
LJ-1 LJ-2 IT-1 IT-2 CJ-1 CJ-2 V-1 V-2 TJ LS-1 LS-2 MW C-1 C-2 S-1 S-2 SD
77.42 73.51 86.68 102.08 106.57 94.64 86.59 109.38 89.50 91.75 95.06 77.97 99.23 79.53 46.15 69.64 84.56
DPP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
2.55b) 3.39 1.09 0.09 2.51 0.97 4.59 3.76 6.61 4.69 5.44 0.04 2.11 4.70 4.46 2.98 0.07
90.50 86.87 94.23 104.54 112.60 114.45 107.68 121.56 107.86 106.27 109.66 115.16 116.35 95.72 58.68 88.16 118.77
DBP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1.14 5.51 1.52 1.39 1.76 4.03 5.68 4.07 3.17 5.49 6.12 0.39 5.06 5.58 5.68 3.77 0.27
86.47 79.59 87.11 101.83 105.13 105.87 100.27 112.16 99.30 95.75 90.12 99.80 97.55 88.12 55.40 83.23 114.80
BBP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
3.81 5.46 1.09 5.31 1.72 4.36 4.46 3.77 1.52 2.13 6.60 4.72 1.26 5.46 5.10 3.58 0.60
90.55 83.15 90.82 99.20 109.29 109.42 100.34 115.60 105.02 100.93 96.17 103.45 112.88 90.87 55.35 84.80 117.68
DCHP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
3.54 5.42 1.48 0.06 1.83 3.99 5.41 4.52 2.15 1.86 5.45 5.50 0.34 5.93 4.02 3.62 0.59
90.47 94.21 90.95 100.20 108.58 111.22 105.96 117.83 107.12 100.54 96.97 110.94 93.88 92.18 57.57 85.52 115.93
DEHP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1.68 3.97 1.69 0.04 1.70 4.21 5.19 4.12 2.57 1.46 6.50 0.47 3.97 5.77 5.47 3.62 1.09
85.01 86.09 93.34 103.51 113.00 107.03 103.60 116.90 104.84 94.11 86.46 110.63 91.96 94.74 58.05 87.36 112.67
DOP ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
1.29 4.53 2.86 0.16 2.60 4.91 5.15 2.42 3.24 1.30 3.25 0.16 0.95 6.09 5.76 3.76 3.19
78.37 78.13 84.04 92.82 105.89 104.00 101.80 103.88 97.71 104.34 98.83 103.99 97.59 89.90 54.47 83.32 101.74
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
4.73 3.87 2.66 0.12 1.91 3.99 5.50 3.57 4.11 3.99 5.74 3.38 1.73 6.48 5.62 3.52 5.49
a) Samples were spiked at 4.00 g/L. b) Mean ± SD
chosen considering the lowest sum of the relative error percentages (RE%) across the whole concentration ranges (Supporting Information Table S9). The calibration equations and correlation coefficients (r) were calculated by linear regression analysis based on least squares method with determined weighting factors. A good linearity was obtained with r values higher than 0.99 (Table 1). LODs and LOQs were calculated from the data of the standard solutions. S/N ratios were accepted as 3 and 10 for the determination of LODs and LOQs, respectively. The LODs and LOQs of seven PAEs were in the range of 0.019–0.208 and 0.072–0.483 g/L, respectively. The linear ranges, calibration equations, correlation coefficients, LODs, LOQs, resolutions, and tailing factors are given in Table 1. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The repeatability and the intermediate precision of the developed method were determined by standard solutions at low, middle, and high concentrations (0.20, 4.00, and 24.00 g/L). Three replicates of each standard solutions were injected consecutively in one day, and then in three separate analytical runs in three different days (Table 2). The precision was calculated as RSD%, and the accuracy was determined in terms of relative error ((observed concentration − nominal concentration)/nominal concentration × 100) (Supporting Information Table S10). All of the precision and accuracy results were lower than 20%, which is the acceptable limit for bioanalytical analysis [28]. The recovery of the USVADLLME method was determined as the percentage of the ratio of observed concentration www.jss-journal.com
2116
J. Sep. Sci. 2014, 37, 2111–2117
P. K. Yılmaz et al.
to nominal concentration of the standard solutions at concentration levels of 0.20, 4.00, and 24.00 g/L. The recovery of PAEs was between 95.05 and 116.90% (Table 2). In the literature, there are several HPLC–UV methods for the quantification of PAEs. Guo et al. developed a HPLC– UV method for the detection of six PAEs (DMP, DBP, DEP, DEHP, BBP, and DOP) in orange juice samples with a total analysis time of 28 min [29]. In the proposed study, it was aimed to develop a reliable, simple, sensitive, and less timeconsuming method that could be applied to real beverage samples. This study had the advantages of separation of seven PAEs (DMP, DPP, DBP, BBP, DCHP, DEHP, and DOP) with acceptable resolution values in 20 min with high precision and accuracy.
was easy to perform, effective in terms of extraction recovery, and suitable for various beverage matrices. The developed method enabled the quantification of seven PAEs in a total analysis time of 20 min with high sensitivity, precision, accuracy, and a wide linear range. The results indicated the requirement of determination and prevention of PAE contamination sources. This study was supported by the Research Fund of Istanbul University (Project number:40324). The authors have declared no conflict of interest.
5 References
3.4 Application of the developed method to the beverage samples
[1] Cao, X.-L., Compr. Rev. Food Sci. F. 2010, 9, 21–43.
The developed method was used for the quantification of seven PAEs in 17 beverage samples stored in plastic bottles. The extraction procedure was replicated three times for each of the original and spiked (4.00 g/L) beverage samples. All of the samples were analyzed in triplicate by HPLC–UV. The matrix effect was different for different kinds of PAEs and beverage samples. It could be concluded that all of the recovery values were higher than 80%, except for soda sample (Supporting Information Fig. S1 and Table 3). DEHP, which has been one of the most widely detected PAEs in foodstuffs [6, 7, 30], was determined in eight of the tested beverage samples within 0.09–1.44 g/L. DMP was detected in cherry juice samples (CJ-1 and CJ-2), vinegar sample (V-1), and sports drink sample (SD) at 0.14, 0.18, 0.18, and 0.02 g/L, respectively. In lemon juice sample (LJ-1), DPP was detected at 0.07 g/L. DCHP was determined only in SD at a concentration of 0.13 g/L. In none of the samples, DBP, BBP, and DOP were detected above the LOQ values (Supporting Information Table S11). Some researchers also reported the contamination of various kinds of beverages by PAEs. Khedr detected DEHP in water, soda, coke, and energy drink samples from Saudi Arabia [30]. In another study, DBP and DEHP were determined in 16 Chinese liquor samples [22]. Also a migration study of DMP, DEP, BBP, DBP, DEHP, and DOP to orange juice samples from China was performed by Guo et al. [29]. Only DEP and DEHP were detected with a storage time of three months, and their concentrations were found to be increased when the expiration date arrived.
[3] Phthalates, TEACH Chemical Summary, Toxicity and Exposure Assessment for Children’s Health, U.S. EPA, 2007.
4 Concluding remarks To the best of our knowledge, this study could be considered as the first report on PAEs analysis in beverage samples from Turkey. In this work, a USVADLLME–HPLC–UV method was developed for the quantification of seven PAEs in different kinds of beverage samples. This USVADLLME procedure C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[2] Ranjbari, E., Hadjmohammadi, M. R., Talanta 2012, 100, 447–453.
[4] Montuori, P., Jover, E., Morgantini, M., Bayona, J. M., Triassi, M., Food Addit. Contam. Part A 2008, 25, 511– 518. [5] LaFleur, A. D., Schug, K. A., Anal. Chim. Acta 2011, 696, 6–26. [6] Xu, D., Deng, X., Fang, E., Zheng, X., Zhou, Y., Lin, L., Chen, L., Wu, M., Huang, Z., J. Chromatogr. A 2014, 1324, 49–56. [7] Cinelli, G., Avino, P., Notardonato, I., Centola, A., Russo, M. V., Anal. Chim. Acta 2013, 769, 72–78. [8] European Food Safety Authority, EFSA J. 2005, 242, 1– 17. [9] European Food Safety Authority, EFSA J. 2005, 241, 1– 14. [10] European Food Safety Authority, EFSA J. 2005, 243, 1– 20. [11] European Food Safety Authority, EFSA J. 2005, 244, 1– 18. [12] European Food Safety Authority, EFSA J. 2005, 245, 1– 14. [13] Cai, Y., Cai, Y., Shi, Y., Liu, J., Mou, S., Lu, Y., Microchim. Acta 2007, 157, 73–79. [14] Yano, K., Hirosawa, N., Sakamoto, Y., Katayama, H., Moriguchi, T., Joung, K. E., Sheen, Y. Y., Asaoka, K., Bull. Environ. Contam. Toxicol. 2002, 68, 463–469. [15] Tsumura, Y., Ishimitsu, S., Saito, I., Sakai, H., Kobayashi, Y., Tonogai, Y., Food Addit. Contam. 2001, 18, 449–460. [16] Xu, Q., Yin, X., Wang, M., Wang, H., Zhang, N., Shen, Y., Xu, S., Zhang, L., Gu, Z., J. Agric. Food Chem. 2010, 58, 11311–11317. [17] Rezaee, M., Assadi, Y., Hosseini, M. R. M., Aghaee, E., Ahmadi, F., Berijani, S., J. Chromatogr. A 2006, 1116, 1– 9. [18] Mousa, A., Basheer, C., Al-Arfaj, A. R., J. Sep. Sci. 2013, 36, 2003–2009. [19] Sha, C., Yi-Sheng, Z., Shui-Yuan, C., Tian, Q., Hao, S., J. Sep. Sci. 2011, 34, 1503–1507.
www.jss-journal.com
J. Sep. Sci. 2014, 37, 2111–2117
[20] Hadjmohammadi, M. R., Ranjbari, E., Int. J. Environ. Anal. Chem. 2012, 92, 1312–1324. [21] Guo, L., Dong, H., Int. J. Environ. Anal. Chem. 2009, 89, 357–365. [22] Leng, G., Chen, W., Zhang, M., Huang, F., Cao, Q., J. Sep. Sci. 2014, 37, 684–690. [23] Zhou, J., Qi, Y., Wu, H., Diao, Q., Tian, F., Li, Y., J. Sep. Sci. 2014, 37, 650–657. [24] Farajzadeh, M. A., Sheykhizadeh, S., Khorram, P., J. Sep. Sci. 2013, 36, 939–946. [25] Hongyuan, Y., Xiaoling, C., Kuo, Y., Analyst 2012, 137, 4860–4866.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Liquid Chromatography
2117
´ ´ C., Barciela Alonso, M. C., Pena-V ˜ ´ [26] Perez Feas, azquez, E., Herbello Hermelo, P., Bermejo-Barrera, P., Talanta 2008, 75, 1184–1189. [27] Li, X., Xiong, W., Lin, H., Zhuo, L., Lv, S., Tang, X., Chen, M., Zou, Z., Lin, Z., Qiu, B., Chen, G., J. Sep. Sci. 2013, 36, 477–484. [28] US Department of Health and Human Services, Guidance for Industry, Bioanalytical Method Validation, FDA, CDER, Rockville, USA 2001. [29] Guo, Z., Wei, D., Wang, M., Wang, S., J. Chromatogr. Sci. 2010, 48, 760–765. [30] Khedr, A., J. Chromatogr. B 2013, 930, 98–103.
www.jss-journal.com
