Simultaneous determination of bisphenol A and tetrabromobisphenol A in tea using a modified QuEChERS sample preparation method coupled with liquid chromatography-tandem mass spectrometry

Guanwei Gao , Hongping Chen *, Li Zhu , Yunfeng Chai , Guicen Ma , Chen Wang , Zhenxia Hao , Xin Liu and Chengyin Lu *
Tea Research Institute, Chinese Academy of Agricultural Sciences, Key Laboratory of Tea Quality and Safety & Risk Assessment, Ministry of Agriculture, Hangzhou, 310008, China. E-mail: Lchy@mail.tricaas.com; Thean27@mail.tricaas.com; Fax: +86-571-86652004; Tel: +86-571-86650607

Received 6th September 2017 , Accepted 20th October 2017

First published on 23rd October 2017


A rapid and sensitive method for the simultaneous determination of bisphenol A (BPA) and tetrabromobisphenol A (TBBPA) in tea using liquid chromatography-tandem mass spectrometry (LC-MS/MS) was developed and validated. Sample preparation was based on a modified QuEChERS procedure through extraction of the target analytes using acetonitrile with 0.7% acetic acid followed by a dispersive solid phase extraction (d-SPE) clean up procedure using C18, MWCNTs and SCX as the adsorbent mixture to eliminate tea co-extracts. The linearity of the method with correlation coefficients (R2) higher than 0.99 was obtained. The limits of detection (LODs), and limits of quantitation (LOQs) were 8.0 and 20.0 μg kg−1 for BPA and 0.4 and 1.0 μg kg−1 for TBBPA, respectively. Satisfactory mean recoveries of BPA and TBBPA at three fortification levels ranged from 88% to 109% and 77% to 99%, respectively, while intra-day and inter-day precisions were below 15%. The developed method was successfully applied to determine BPA and TBPA in 136 tea samples.


Introduction

The control of risks associated with emerging food contaminants is one of the great challenges in food safety. More than 700 emerging pollutants and their metabolites and transformation products, are listed as present in the European aquatic environment.1 Among the emerging food contaminants, bisphenol A (BPA) and tetrabromobisphenol A (TBBPA) received the foremost attention because of their adverse ecological and human health effects. BPA is an organic compound and consists of two phenol molecules bonded by a methyl bridge and two methyl groups. As a synthetic monomer, BPA was widely used in many productions, including polycarbonate plastics, epoxy resin linings of canned foods and beverage containers.2 BPA could be introduced into food and beverages through migration from polycarbonate tools and containers or epoxy coatings.3 TBBPA is the most widely used brominated flame retardant (BFR).4 TBBPA is produced primarily from the primitive e-waste dismantling, manufacturing and consumption of products which are made up of TBBPA-based materials. Owing to the massive use and emissions of BPA-based and TBBPA-based materials, BPA and TBBPA can be released into the environment and cause adverse ecological and human health effects. BPA and TBBPA in soil, dust, air and other types of environmental media could be absorbed by crops and then enter the food chain.

Regarding the acute oral toxicity of BPA and TBBPA, their adverse effects on human health have aroused extensive concerns. Recent research indicates that BPA can disrupt the natural hormone balance in humans and can be particularly harmful to fetus, infants, and young children.5,6 TBBPA may increase the levels of circulating estrogens and produce uterine tumors.7 Furthermore, residues of BPA and TBBPA in tea may pose a threat to the health of tea drinkers.7,8 There is a tolerable daily intake (TDI) and reference dose (RfD) of 50 μg kg−1 day−1 estimated by the European Food Safety Authority (EFSA, 2007) and the US Environmental Protection Agency (EPA, 2009).9 The Environment Canada Domestic Substances list includes TBBPA as a priority substance for screening assessment.10

Different analytical methods have been developed for monitoring BPA and TBBPA in beverage and food consumption.11–15 Owing to the complex matrices, tedious sample preparation (extraction and cleanup) is required prior to the analysis by liquid chromatography (LC) coupled to ultraviolet (UV) absorbance or fluorescence detection. Compared with LC-UV absorbance or fluorescence detection, gas chromatography mass spectrometry (GC-MS) offers a higher degree of selectivity and sensitivity. However, GC-MS required a time consuming step of derivatization. A high performance liquid chromatography-electrospray (negative) ionization-tandem mass spectrometry (LC-MS/MS) based method has been successfully employed in the determination of BPA and TBBPA in milk, sediment and sludge samples.14,16

Tea, a product prepared from buds and two or three leaves of Camellia sinensis, is one of the most widely consumed functional beverages worldwide. Tea is a major cash crop and thus is an important economic resource in China, India, Sri Lanka, and other developing countries. Tea has increasingly received attention due to its health beneficial properties such as antioxidant, anti-carcinogenic, and anti-microbial properties.17 However, tea leaves inevitably absorb and accumulate certain environmental contaminants due to their long plantation time and complex processing. Thus, the residues of the environmental contaminants in tea should be monitored. The potential adverse effect of certain chemical contaminants on human health through tea has generated great concerns during the past few years.18 Studies have indicated that plants could uptake and accumulate four commonly occurring PPCP/EDCs, i.e., BPA.19 Tea plant leaves possess a high surface area, and so they could strongly accumulate pollutants from the environment.20,21 BPA and brominated BPA can be formed during the combustion.22 Therefore, the manufacturing process of tea which includes drying by combustion gases produced from burning of wood, oil, or coal may also lead to BPA and TBBPA in the tea product. Thus, it is of great significance to develop a proper extraction and quantitative method for the determination of BPA and TBBPA in tea samples.

Analysis of contaminants in tea samples is a challenge because of the presence of tea matrix co-extracts, e.g. pigments, alkaloids and polyphenols and other interferents, which can interfere with the analytes resulting in matrix interference and complicated extraction procedures. The QuEChERS (Quick, Easy, Cheap, Effective, Rugged and Safe) sample preparation method is a powerful alternative technique of sample preparation developed originally for pesticide residue analysis in food of plant origin.23 The procedure of the original QuEChERS method involves initial extraction with acetonitrile, liquid–liquid partitioning after addition of salt, followed by a simple clean-up step. However, the original QuEChERS method could not clean up the complex matrix efficiently. Thus, numerous versions of modified QuEChERS methods have been developed to extend this technique for contaminant residue analysis in some complex matrices.24,25 In our previous study, the modified QuEChERS method was successfully applied for determination of pesticides, phthalate esters (PAEs) and polycyclic aromatic hydrocarbons (PAHs) in tea samples.26–28 The modified QuEChERS method could clean up the complex tea matrix efficiently with simple pre-treatment and low consumption of organic solvents.

To the best of our knowledge, qualitative and quantitative analytical methods for the determination of BPA and TBBPA in tea samples have not been reported. Therefore, information on the contamination, source and health risk of BPA and TBBPA in tea is lacking. In the present work, we developed a modified QuEChERS method for the LC-MS/MS simultaneous determination of BPA and TBBPA in tea samples. The usefulness of the method was verified in terms of linearity, accuracy, precision, limits of detection (LODs), and limits of quantitation (LOQs) of BPA and TBBPA in different tea matrices.

Materials and methods

Chemicals

Standard bisphenol A (BPA) and tetrabromobisphenol A (TBBPA) were purchased from ANPEL Laboratory Technologies Inc. (Shanghai, China) and were of minimum 98% purity. The standards (BPA and TBBPA) were separately dissolved in methanol at 1000 μg mL−1. Two single standard solutions were mixed and diluted with methanol to 10 μg mL−1. Stock solutions were diluted to appropriate intermediate working solutions using methanol. Twelve calibration standard solutions were prepared at 0.1, 0.25, 0.5, 1, 2, 5, 10, 20, 50, 100, 200 and 500 μg L−1. These standard solutions were stored at 4 °C in the dark.

Acetonitrile and methanol were HPLC grade solvents and were purchased from Sigma-Aldrich (Merck, Germany). Ultrapure water was purified with a Milli-Q plus ultrapure water system (Millipore, Bedford, MA, USA). Acetic acid solution (51%) was supplied by Sigma-Aldrich (Merck, Germany). Tianjin Bonna-Agela Technologies (Tianjin, China) provided the primary secondary amine (PSA), octadecyl silica (C18), strong cation exchanger (SCX), strong anion exchanger (SAX), multi-walled carbon nanotubes (MWCNTs) and graphitised carbon black (GCB). Analytical-grade magnesium sulfate (MgSO4) and sodium chloride (NaCl) were supplied by Huadong Medicine (Zhejiang Medicine, China).

Sample collection

All of the tea samples used in this study were purchased from different local markets. A total of 136 tea samples including 29 green tea samples, 34 oolong tea samples, 54 non-smoked black tea samples and 19 smoked black tea samples were collected for BPA and TBBPA determination. Two organic tea samples free of BPA and TBBPA (green tea and oolong tea) were used as blanks and fortified samples for the optimization and method validation process. Before the method validation process, the samples were tested for the absence of BPA and TBBPA residues using this developed method. As BPA and TBBPA were commonly found in all black tea samples, one black tea sample containing low BPA and TBBPA residues was selected as the fortified sample. The prepared black tea sample contained BPA (4.0 ± 1.5 μg kg−1) and TBBPA (0.4 ± 0.1 μg kg−1). Smoked black tea is obtained when, after fermentation, the tea leaves are smoked by burning pine, spruce, or bamboo, giving a heavy smoky taste and smell to the tea leaves and its tea infusions.

Sample preparation

Briefly, 2.5 g of thoroughly homogenized tea sample was weighed into 50 mL centrifuge tubes, after which 5 mL of water was added. The mixture was vortexed for 1 min. Then, 10 mL acetonitrile with 0.7% acetic acid was used to extract BPA and TBBPA by using a rotary agitator (Dragon, China) at 70 rpm for 30 min. After the addition of 2 g of anhydrous MgSO4 and 4 g of NaCl, the tube was vigorously shaken for 1 min. Afterwards the extracts were centrifuged at 10[thin space (1/6-em)]000 rpm for 10 min. Two mL of the supernatants was transferred into a 5 mL tube containing 10 mg of MWCNTs, 100 mg of C18, and 200 mg of SCX. The mixture was vortexed for 1 min and centrifuged at 4000 rpm for 10 min. Finally, the extract (containing 0.25 g tea for every 1 mL solution) was filtered through 0.2 μm PTFE into an autosampler vial for LC-MS/MS analysis.

Instrumentation and analytical conditions

In this study, BPA and TBBPA were quantitatively and qualitatively determined on an Acquity Ultra-Performance LC system (consisting of a vacuum degasser, auto-sampler, and binary pump) (Waters, USA) equipped with an Acquity UPLC HSS T3 column (100 × 2.1 mm I.D., 1.8 μm particle size, Waters, USA). The mobile phase was composed of water (A) and methanol (B), and was pumped at a flow rate of 0.25 mL min−1. The gradient profile was 10% B at the beginning, held for 1.0 min, linearly increased to 25% B from 1.0–3.0 min, then linearly increased to 95% B from 3.0–5.0 min, 95% B from 5.0–9.0 min, and finally came back to the initial conditions, 10% B from 10 to 12 min. The elution program time for each injection was 12 min. The column compartment was maintained at 40 °C and a sample volume of 10 μL was injected into the separation system using an autoinjector.

An Applied Biosystems 5500 Q system (AB SCIEX, USA) equipped with an ion-spray interface operated in the ESI negative mode with selective reaction monitoring (SRM). The ion spray voltage and source temperature were held at 4.5 kV and 500 °C, respectively. The curtain (CUR) and collision gas (CAD) flow rates were set at 35 and 7 psi, respectively. Both the atomization air pressure (GS1) and auxiliary gas (GS2) were set at 50.0 psi. The MS/MS parameters were adjusted to optimize the performance for each compound and are summarized in Table 1.

Table 1 LC-MS/MS parameters for the determination of BPA and TBBPA
Compounds Retention time (min) Ion transitions (m/z) Collision voltage (eV) Ion transitions (m/z) Collision voltage (eV) Ion transitions (m/z) Collision voltage (eV)
a Ion transition for quantification.
BPA 7.30 227.1/133.0a 40 227.1/211.1 40 227.1/212.1 40
TBBPA 8.06 542.8/417.8a 55 542.8/445.7 43 542.8/78.9 130


Validation procedure

Blank assays were performed in each batch analysis to ensure the lack of BPA and TBBPA background contamination. Three tea samples (black, green and oolong tea) were analyzed by the procedure described above. Two homogenized tea samples (green and oolong tea) showing the absence of the target analytes were used as blank samples, and one black tea sample containing BPA and TBBPA near the LODs was used for recovery testing. For recovery testing, tea samples were fortified at 20, 50, and 100 μg kg−1 for BPA and 1, 10, and 100 μg kg−1 for TBBPA. Each level was analyzed six times (n = 6). The fortified samples were kept at room temperature for 2 h to allow BPA and TBBPA absorption into the tea samples, and then they were subjected to the entire sample preparation procedure. The precision of the method was verified by measuring the intra-day and inter-day precisions. The intra-day precisions were assessed for six parallel extractions at the same spiking level over a day. The inter-day precisions were assessed for three levels, and the spiked samples were extracted and analyzed independently on three successive days.

The calibration curves for BPA and TBBPA were evaluated at matrix-matched calibration standards of twelve different concentrations (0.1, 0.25, 0.5, 1, 2, 5, 10, 20, 50, 100, 200 and 500 μg L−1). The LODs and LOQs of BPA and TBBPA were estimated as the lowest matrix-matched standard, which provided a signal-to-noise ratio (S/N) higher than 3 and 10, respectively.

Results and discussion

LC-MS/MS optimization

The LC-MS/MS conditions were optimized using the individual standard solutions (1 mg L−1) of BPA and TBBPA in methanol. In negative ion mode (ESI−), the deprotonated molecular ions ([M[thin space (1/6-em)][thin space (1/6-em)] − H]) at m/z 227.1 and 542.8 for BPA and TBBPA, respectively, were acquired from the full-scan MS spectrum. Three transitions were selected for identification, and the collision voltages (CVs) were optimized based on their sensitivity at different CVs. For BPA, three abundant fragments at m/z 212.1, 211.1 and 133.0 were obtained from the deprotonated molecular ion. The product ion at m/z 133.0 was chosen as the quantitation ion because of the highest response. For TBBPA, although the intensity of the signal at m/z 542.8/78.9 was higher than that at m/z 542.8/417.8, 542.8/445.7 and 542.8/290.8, the signal-to-noise ratio (S/N) of the transition at 542.8/417.8 was considerably higher than those of the other transitions (Table 2). Therefore, the transition at m/z 542.8/417.8 was chosen for the quantitation ion.
Table 2 The signal-to-noise ratio (S/N) of the transitions for TBBPA obtained from spiked tea samples (4 μg kg−1)
Tea Transitions
542.8/417.8 542.8/445.7 542.8/290.8 542.8/78.9
Black tea 35.5 23.8 13.9 4.6
Green tea 40.9 49.3 28.1 13.4
Oolong tea 21.2 18.5 6.5 10.7


Previous studies illustrated that, compared with acetonitrile, methanol as the mobile phase could obtain higher LC-MS/MS response of BPA and TBBPA.16 Additionally, methanol acidified with acetic acid avoids the ionization of the target compounds in negative ESI mode, resulting in the reduction of the signal obtained from LC-MS/MS. In this study, we found that signals of BPA and TBBPA increased about 95% and 53% for methanol than acidified methanol as the mobile phase.

Extraction and d-SPE clean up

Optimization of the extraction procedure. The QuEChERS method was modified for the determination of the target compounds in tea samples. Acetonitrile has a high dissolving ability for the target compound and has good permeability into tea samples. Acetonitrile extraction with d-SPE cleanup was efficient in the removal of complex co-extracts from tea samples. In addition, prior to acetonitrile extraction, ground tea powder was soaked in 5 mL water to obtain high extraction efficiency of the target compounds from the tea matrix.26,27

Previous studies indicated that pH could affect the extraction efficiency of several pesticides from vegetables and fruits.29 In this study, the recoveries of BPA and TBBPA using non-acidified acetonitrile extraction from the spiked tea samples ranged from 69% to 78%. The loss of BPA and TBBPA might occur during the extraction. Considering that BPA and TBBPA are weak acids with typical pKa values of 9.5 and 8.5, the pH of the extraction solution could have an important influence on the extraction efficiency of both compounds from the tea samples.30 In this study, in order to obtain high extraction efficiency, the solution pH of the extraction solvent was investigated. As shown in Fig. 1, the solution pH of the extraction solvent was maintained from 3 to 7 with acetic acid. The extraction solvents in the range of pH 3–7 had little significance on the recoveries of BPA. The recoveries of TBBPA increased from 88% to 95% when the solution pH changed from 3 to 4, but slightly deceased in the range of pH 4–7. The presence of the acid could speed up the hydrolysis of the ester bond linking these compound molecules between the target compounds and tea matrix.31 According to these results, acetonitrile with 0.7% acetic acid (pH = 4) was selected as the extraction solvent.


image file: c7ay02145c-f1.tif
Fig. 1 Effect of pH on the extraction recovery of BPA and TBBPA. Note: the solution pH of the extraction solvent was maintained from 3 to 7 with acetic acid.
Optimization of modified QuEChERS adsorbents. To achieve a satisfactory purification effect and recovery of BPA and TBBPA, several sorbents (PSA, C18, GCB, MWCNTs, SCX, and SAX) were investigated. In this study, 1 mL of the mixture of BPA and TBBPA standard solutions prepared with acetonitrile at 100 μg L−1 was added to six adsorbents individually. As shown in Fig. 2a, the recoveries of the target analytes were lower than 20% when 100 mg of PSA was used. The results of low recoveries of BPA and TBBPA were probably attributed to strong electrostatic interactions between PSA and the target analytes, which contain –OH groups.23 Being a strong anion exchange sorbent, SAX exhibited a strong retention ability for the target analytes, which resulted in low recoveries below 60% (Fig. 2a). For C18 (100 mg), SCX (100 mg), MWCNTs (10 mg), and GCB (50 mg), little adsorption of BPA and TBBPA was observed, and the recoveries of BPA and TBBPA ranged from 85% to 97% and 71% to 109%, respectively. MWCNTs and GCB have strong retention ability for organic molecules due to their high aspect ratio and surface area, and remarkably high mechanical strength. The obtained results in this study indicated that the clean-up capability of MWCNTs (50 mg) was higher than that of GCB (100 mg) and thus a lower amount of sorbent was required for the mixed d-SPE sorbents using MWCNTs. Moreover, several findings previously reported that the addition of more amounts of adsorbents resulted in a decrease of the recoveries of the target analytes.28,32 Therefore, the amounts of C18, SCX and MWCNTs were further evaluated to obtain satisfactory recovery.
image file: c7ay02145c-f2.tif
Fig. 2 The retaining ability of six types of adsorbents (a) and C18, MWCNTs, SCX at various amounts (b) with BPA and TBBPA at 100 μg L−1 prepared with acetonitrile solvent.

Different amounts of C18 (50, 100, 150, and 200 mg), SCX (50, 100, 150, and 200 mg) and MWCNTs (5, 10, 25, and 50 mg) were tested to evaluate their effects on the recoveries of BPA and TBBPA. As presented in Fig. 2b, recoveries of BPA and TBBPA ranged from 85% to 113% with amounts of MWCNTs less than or equal to 10 mg, but decreased below 80% when the amount of MWCNTs increased from 25 to 50 mg. We previously found that the color of the extracts became less intense with the amounts of MWCNTs and C18 increasing26,27 and this phenomenon was also observed when the amount of SCX was increased in this study. The results shown in Fig. 2b illustrate that satisfactory recoveries were observed at various amounts of SCX and C18 within 50–200 mg. Therefore, we chose 10 mg of MWCNTs and 200 mg of SCX as two components in the mixed d-SPE sorbents. The color of the extracts had little change when the amount of C18 increased from 100 to 200 mg. Taking into consideration the recoveries of BPA and TBBPA, color intensity of final extracts, and removal of co-extracts, the adsorbent mixture containing 100 mg of C18, 10 mg of MWCNTs, and 200 mg of SCX was selected to clean up the tea co-extracts.

In this study, tea extracts spiked with BPA and TBBPA at 50 μg L−1 were employed to evaluate the recoveries and removal of the tea matrix using the proposed mixed adsorbents (100 mg of C18, 10 mg of MWCNTs, and 200 mg of SCX). The color of the final extract became light yellow, while the recoveries of BPA and TBBPA obtained ranged from 91% to 99%. Therefore, the mixture of adsorbents containing 100 mg of C18, 10 mg of MWCNTs and 200 mg of SCX was used as the modified QuEChERS adsorbent.

Method validation

Matrix effects. Previous studies illustrated that signal suppression is a phenomenon in the determination of BPA and TBBPA in milk (20%) and children’s urine (83–89%), which indicated that BPA and TBBPA were easily subjected to ion suppression. In the present study, ME suppression ranged from 50% to 54% for BPA and 7–29% for TBBPA, although the modified QuEChERS extraction followed by the d-SPE clean up procedure was used for the removal of tea matrices. For this reason, matrix-matched calibration standards were prepared for quantification analysis.
Linearity, limit of detection and quantification. The linearity was examined with a calibration curve, obtained by measuring matrix-matched calibration standards of several concentrations levels ranging from 0.1 to 500 μg L−1. Linear regression analysis was carried out for peak areas versus analyte concentrations. The results are shown in Table 3. The matrix calibration curves were linear over the range of 5 μg L−1 to 500 μg L−1 and 0.25 μg L−1 to 500 μg L−1 for BPA and TBBPA, respectively. As shown in Table 3, good linearity with correlation coefficients (R2) was obtained in the range of 0.9919 to 0.9996. LODs and LOQs of the method obtained for the three tea matrices in this study were determined to be 8.0 and 20.0 μg kg−1 for BPA and 0.4 and 1.0 μg kg−1 for TBBPA, respectively.
Table 3 Linear range, correlation of coefficient (R2), recovery, relative standard deviations (RSD), matrix effect (ME), LOQ and LOD obtained for the target compounds in different tea matrices
Compounds Tea samples Linear range (μg L−1) R 2 ME (%) Recoveries, % Intra-day precision (RSD, %, n = 6) Inter-day precision (RSD, %, n = 3) LOD (μg kg−1) LOQ (μg kg−1)
Spiked levela Spiked level Spiked level
L1 L2 L3 L1 L2 L3 L1 L2 L3
a Spiked levels (L1, L2, L3) of BPA at 20, 50 and 100 μg kg−1 and TBBPA at 1, 10 and 100 μg kg−1.
BPA Black tea 5–500 0.9931 −54 93 91 95 10 6 10 4 3 5 8.0 20.0
Green tea 5–500 0.9955 −50 109 92 99 9 3 5 2 3 10 8.0 20.0
Oolong tea 5–500 0.9951 −52 99 88 97 7 9 5 3 4 6 8.0 20.0
TBBPA Black tea 0.25–500 0.9949 −29 83 98 95 7 3 8 5 4 5 0.4 1.0
Green tea 0.25–500 0.9996 −7 99 85 90 3 6 3 14 8 8 0.4 1.0
Oolong tea 0.25–500 0.9919 −28 77 82 90 3 5 4 6 4 3 0.4 1.0


Accuracy and precision. The intra-day and inter-day accuracy and precision data are shown in Table 3. The accuracy of the method was validated by measuring the recovery of analytes from the control samples for black tea, green tea and oolong tea with spiked levels of 20, 50, and 100 μg kg−1 for BPA and 1, 10, and 100 μg kg−1 for TBBPA. Considering that BPA and TBBPA were commonly found in black tea samples, a background value was measured in the present study from the selected fortified sample via the proposed method, and a compensation method was adopted by deducting the background. Satisfactory recoveries of BPA and TBBPA in all matrices ranged from 88% to 109% and 77% to 99%, respectively. The intra-day and inter-day precisions were investigated by analysing the spiked tea samples (20, 50, 100 μg kg−1 for BPA and 1, 10, 100 μg kg−1 for TBBPA) on the same day and during three successive days, respectively. The intra-day precision (RSD) for BPA ranged from 3% to 10%, and from 3% to 8% for TBBPA. The inter-day precision (RSD) ranged from 2% to 10% and from 3% to 14% for BPA and TBBPA, respectively.

Comparison of analytical performance with other methods

Table 4 compares the analytical performance of the proposed method (recovery and LOQs) with other methods reported for the determination of BPA and TBBPA in different samples. It was observed that by using internal standard, mean recoveries of BPA and TBBPA from spiked milk, seafood, sediment and sludge samples ranged from 67% to 107%. In the present study, similar recoveries of BPA and TBBPA (77–109%) were achieved without using internal standard, while intra-day and inter-day precisions were below 15%. The results suggested that the accuracy of the optimized method achieved acceptable recoveries in line with criteria formulated by European Union (EU) guidelines.33
Table 4 Comparison of analytical performance of the proposed method with other methods using LC-MS/MS reported for the determination of BPA and TBBPA in different samples
Analytes Sample Analytical technique Recovery (%) LOQs References
BPA Seafood LC-MS/MS 67–71 0.5–2.5 μg kg−1 (dry weight) 11
BPA Children urine LC-MS/MS 77.2–102.6 0.1 ng mL−1 13
BPA Milk LC-MS/MS 83–106 2.2 ng mL−1 14
BPA Sediment and sludge LC-MS/MS 70–105 0.15 μg kg−1 (dry weight) 34
BPA Human urine LC-MS/MS 101–105 0.4 ng L−1 35
BPA Placental tissue LC-MS/MS 98–99 0.5 μg kg−1 36
BPA Tea LC-MS/MS 88–109 20.0 μg kg−1 This study
TBBPA Seafood LC-MS/MS 77–107 0.25–1.0 μg kg−1 (dry weight) 11
TBBPA Sediment and sludge LC-MS/MS 70–105 0.03 μg kg−1 (dry weight) 34
TBBPA Human serum LC-MS/MS 83.3–103.8 4.1 μg kg−1 37
TBBPA Tea LC-MS/MS 77–99 1.0 μg kg−1 This study


As can be seen from Table 4, the LOQs of BPA and TBBPA in seafood, sediment and sludge ranged from 0.15 to 2.5 μg kg−1 and 0.03 to 1.0 μg kg−1 (dry weight), respectively. The LOQ of TBBPA obtained from the present method was similar with those previously reported in different samples. The LOQ of BPA in tea matrices achieved in this study is higher than those in different samples reported previously. The reason might be that the presence of the tea matrix resulted in significant signal suppression of BPA (50–54%) and TBBPA (7–29%). Additionally, the LOQ of BPA in tea matrices (20 μg kg−1) is adequate for safety control of food products since it is much lower than the specific migration limit (SML) of 600 μg kg−1 established by the European Food Safety Authority (EFSA).34 Therefore, the LOQs of BPA and TBBPA in the proposed method were deemed satisfactory for the residue determination of BPA and TBBPA in tea samples.

Real samples

The proposed method was applied to analyze tea samples to further evaluate its performance. The samples included 29 green tea samples, 34 oolong tea samples, and 73 black tea samples (including 54 non-smoked black tea samples and 19 smoked black tea samples). Fig. 3 shows the representative LC-MS/MS chromatograms of BPA and TBBPA in the blank tea sample, spiked tea sample, matrix-matched standard and real tea sample.
image file: c7ay02145c-f3.tif
Fig. 3 Typical LC-MS/MS chromatograms of BPA (a1) and TBBPA (b1) in blank tea sample, spiked tea sample at 100 μg kg−1 (a2 and b2), matrix-matched standard at 25 μg L−1 (a3 and b3), real tea sample containing 372.6 μg kg−1 BPA (a4) and TBBPA 11.3 μg kg−1 (b4).

Table 5 shows the positive rates and concentrations of BPA and TBBPA detected in the tea samples. It was found that the positive rate of BPA detected was 66–100%. In general, an increase of the positive rate of BPA detected in the tea samples can be confirmed to be in the following order: green tea (66%) < oolong tea (88%) < non-smoked black tea (93%) < smoked black tea (100%). The positive rate of BPA detected in the smoked black samples (100%) was obviously higher than those in other types of tea samples. The maximum concentrations of BPA in green tea, oolong tea, non-smoked black tea and smoked black tea were 45.6, 86.0, 646.0 and 580.0 μg kg−1, respectively (Table 5). These results emphasized that the mean concentrations of BPA in the smoked black tea sample (157.0 μg kg−1) were much higher than that in the other types of tea samples (29.0–91.7 μg kg−1). A particularity for the manufacture of smoked black tea is that, after fermentation, the tea leaves are smoked by burning pine, spruce, or bamboo. Studies have indicated that BPA can be formed during various combustion processes.22 Thus, combustion of the pine, spruce, or bamboo could be a significant source of BPA in tea leaves during the tea manufacturing process. Although higher mean concentrations of BPA were found in smoked tea samples, the highest concentration (646.0 μg kg−1) of BPA was found in the non-smoked black tea sample (Table 5). Orthodox black tea is dried by combustion gases produced from burning of wood, oil, or coal. Therefore, tea leaves may have the absorption ability of BPA from combustion, resulting in high BPA contents.

Table 5 Concentrations of BPA and TBBPA in 29 green teas, 34 oolong teas, 54 non-smoked black teas and 19 smoked black teas
Tea sample BPA TBBPA
Positive rate (%) Mean residue (μg kg−1) Max residue (μg kg−1) Positive rate (%) Mean residue (μg kg−1) Max residue (μg kg−1)
Green tea 66 29.0 45.6 93 3.5 11.3
Oolong tea 88 44.6 86.0 35 4.4 6.5
Non-smoked black tea 93 91.7 646.0 96 1.5 2.7
Smoked black tea 100 157.0 580.0 100 1.6 4.5


As shown in Table 5, the positive rates of TBBPA detected in the non-smoked black tea (96%) and smoked black tea (100%) samples were higher than those in the oolong tea (35%) and green tea (93%) samples. Previous studies indicated that brominated BPA could be produced from various combustion processes.22 The manufacturing process of orthodox black tea and smoked black tea may also lead to TBBPA contamination in the tea products. By contrast, the mean concentrations of TBBPA (Table 5) follows the order: non-smoked black tea (1.5 μg kg−1) < smoked black tea (1.6 μg kg−1) < green tea (3.5 μg kg−1) < oolong tea (4.4 μg kg−1). Because the loss of water increased the dry matter concentration in tea leaves (up to 4.0–5.0 times) during manufacture, contaminant residues in tea leaves increased by 1.7–1.8 times after the manufacture of tea.20 Tea plant leaves could strongly accumulate pollutants from the environment, which implied that contaminant residues in tea leaves might be the main source of TBBPA in tea products.

Conclusions

In the present work, a rapid and sensitive method for the simultaneous determination of BPA and TBBPA in tea using LC-MS/MS was developed and validated. Acetonitrile with 0.7% acetic acid (pH = 4) was used to overcome incomplete extraction of the target compounds from tea samples. The adsorbents of d-SPE were optimized, and the use of C18, MWCNTs and SCX as d-SPE sorbents was applied in the purification process. The developed method was validated in terms of linearity, precision and accuracy of BPA and TBBPA in different tea matrices. Satisfactory precision and accuracy were obtained from the spiked samples. In conclusion, the proposed method possesses great potential for simultaneous determination of BPA and TBBPA in real tea samples.

Conflicts of interest

There are no conflicts of interest to declare for the authors of this study.

Acknowledgements

This study was funded by the earmarked fund for Innovative Research Team in Chinese Academy of Agricultural Sciences (CAAS-ASTIP-2017-TRICAAS), the National Natural Funding (31671941), Zhejiang Provincial Natural Science Foundation of China (LY15C200019), the Modern Agro-Industry Technology Research System (CARS-23), and the National Center for Engineering and Technology Project (2014FU125Q10).

Notes and references

  1. Norman-network, http://www.norman-network.net/?q=node/19, accessed June, 2017.
  2. D. D. Seachrist, K. W. Bonk, S. Ho, G. S. Prins, A. M. Soto and R. A. Keri, Reprod. Toxicol., 2016, 59, 167–182 CrossRef CAS PubMed.
  3. Y. Niu, J. Zhang, H. Duan, Y. Wu and B. Shao, Food Chem., 2015, 167, 320–325 CrossRef CAS PubMed.
  4. K. Liu, J. Li, S. Yan, W. Zhang, Y. Li and D. Han, Chemosphere, 2016, 148, 8–20 CrossRef CAS PubMed.
  5. I. Rykowska and W. Wasiak, Acta Chromatogr., 2006, 16, 7–27 CAS.
  6. A. Careghini, A. F. Mastorgio, S. Saponaro and E. Sezenna, Environ. Sci. Pollut. Res., 2015, 22, 5711–5741 CrossRef CAS PubMed.
  7. D. Y. Lai, S. Kacew and W. Dekant, Food Chem. Toxicol., 2015, 80, 206–214 CrossRef CAS PubMed.
  8. J. R. Rochester, Reprod. Toxicol., 2013, 42, 132–155 CrossRef CAS PubMed.
  9. J. S. Lakind and D. Q. Naiman, J. Exposure Sci. Environ. Epidemiol., 2011, 21, 272–279 CrossRef CAS PubMed.
  10. Environment Canada, Screening assessment report TBBPA. En14–110/2013E-PDF, Government of Canada, Ottawa, Ontario, Canada, http://www.ec.gc.ca/ese-ees/BEE093E4-8387-4790-A9CD-C753B3E5BFAD/FSAR_TBBPA_EN.pdf, accessed, June, 2017 Search PubMed.
  11. S. C. Cunha and J. O. Fernandes, Food Control, 2013, 33, 549–555 CrossRef CAS.
  12. X. Cao, J. Corriveau and S. Popovic, J. Agric. Food Chem., 2009, 57, 1307–1311 CrossRef CAS PubMed.
  13. S. Errico, M. Bianco, L. Mita, M. Migliaccio, S. Rossi, C. Nicolucci, C. Menale, M. Portaccio, P. Gallo, D. G. Mita and N. Diano, Food Chem., 2014, 160, 157–164 CrossRef CAS PubMed.
  14. N. C. Maragou, E. N. Lampi, N. S. Thomaidis and M. A. Koupparis, J. Chromatogr. A, 2006, 1129, 165–173 CrossRef CAS PubMed.
  15. S. C. Cunha, C. Oliveira and J. O. Fernandes, Anal. Bioanal. Chem., 2017, 409, 151–160 CrossRef CAS PubMed.
  16. S. Chu, G. D. Haffner and R. J. Letcher, J. Chromatogr. A, 2005, 1097, 25–32 CrossRef CAS PubMed.
  17. M. Da Silva Pinto, Food Res. Int., 2013, 53, 558–567 CrossRef CAS.
  18. A. M. Abd El-Aty, J. Choi, M. M. Rahman, S. Kim, A. Tosun and J. Shim, Food Addit. Contam., Part A, 2014, 31, 1794–1804 CrossRef CAS PubMed.
  19. L. K. Dodgen, J. Li, D. Parker and J. J. Gan, Environ. Pollut., 2013, 182, 150–156 CrossRef CAS PubMed.
  20. P. Liu, H. Chen, G. Gao, Z. Hao, C. Wang, G. Ma, Y. Chai, L. Zhang and X. Liu, J. Agric. Food Chem., 2016, 64, 8909–8917 CrossRef CAS PubMed.
  21. D. Lin, L. Zhu, W. He and Y. Tu, J. Agric. Food Chem., 2006, 54, 3658–3662 CrossRef CAS PubMed.
  22. S. Sidhu, B. Gullett, R. Striebich, J. Klosterman, J. Contreras and M. DeVito, Atmos. Environ., 2005, 39, 801–811 CrossRef CAS.
  23. M. Anastassiades and S. J. Lehotay, J. AOAC Int., 2003, 86, 412–431 CAS.
  24. O. P. Luzardo, N. Ruiz-Suárez, M. Almeida-González, L. A. Henríquez-Hernández, M. Zumbado and L. D. Boada, Anal. Bioanal. Chem., 2013, 405, 9523–9536 CrossRef CAS PubMed.
  25. A. G. Frenich, R. Romero-González, M. L. Gómez-Pérez and J. L. M. Vidal, J. Chromatogr. A, 2011, 1218, 4349–4356 CrossRef CAS PubMed.
  26. H. Chen, P. Yin, Q. Wang, Y. Jiang and X. Liu, Food Anal. Method., 2014, 7, 1577–1587 CrossRef.
  27. P. Yin, X. Liu, H. Chen, R. Pan and G. Ma, Food Addit. Contam., Part A, 2014, 31, 1406–1413 CrossRef CAS PubMed.
  28. H. Chen, G. Gao, P. Liu, R. Pan, X. Liu and C. Lu, Food Anal. Method., 2016, 9, 2374–2384 CrossRef.
  29. M. Á. González-Curbelo, B. Socas-Rodríguez, A. V. Herrera-Herrera, J. González-Sálamo, J. Hernández-Borges and M. Á. Rodríguez-Delgado, TrAC, Trends Anal. Chem., 2015, 71, 169–185 CrossRef.
  30. M. Sadeghi, Z. Nematifar, N. Fattahi, M. Pirsaheb and M. Shamsipur, Food Anal. Method., 2016, 9, 1814–1824 CrossRef.
  31. S. Biedermann-Brem, K. Grob and P. Fjeldal, Eur. Food Res. Technol., 2008, 227, 1053–1060 CrossRef CAS.
  32. A. Wilkowska and M. Biziuk, Food Chem., 2011, 125, 803–812 CrossRef CAS.
  33. Sante/11945/2015. Method validation and quality control procedures for pesticide residues analysis in food and feed.
  34. European Food Safety Authority (EFSA), EFSA J., 2015, 13, 3978 CrossRef.
  35. B. A. Rocha, B. R. B. Da Costa, N. C. P. de Albuquerque, A. R. M. de Oliveira, J. M. O. Souza, M. Al-Tameemi, A. D. Campiglia and F. Barbosa Jr, Talanta, 2016, 154, 511–519 CrossRef CAS PubMed.
  36. I. Jiménez-Díaz, A. Zafra-Gómez, O. Ballesteros, N. Navea, A. Navalón, M. F. Fernández, N. Olea and J. L. Vílchez, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2010, 878, 3363–3369 CrossRef PubMed.
  37. T. Hayama, H. Yoshida, S. Onimaru, S. Yonekura, H. Kuroki, K. Todoroki, H. Nohta and M. Yamaguchi, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2004, 809, 131–136 CrossRef CAS PubMed.

This journal is © The Royal Society of Chemistry 2017