以超音波萃取搭配即時直接分析串聯式質譜術半定量食品包材中鄰苯二甲酸酯類與雙酚類

Abstract

鄰苯二甲酸酯 (phthalate esters, PAEs) 可軟化塑膠，被廣泛應用於製造聚氯乙烯 (polyvinyl chloride, PVC) 產品以及食品包材。雙酚類 (bisphenols, BPs) 則被用作熱感應紙的顯色劑與罐頭的塗層材料。鄰苯二甲酸酯類與雙酚類為已知的內分泌干擾物質 (endocrine disrupting chemicals)，對內分泌系統、生殖系統與孩童發展存在顯著的健康效應。有鑒於相關人體危害，許多國家已設立鄰苯二甲酸酯類與雙酚類於食品包材中的規範標準，造就此兩類物質的替代物使用量與日俱增。人體暴露於鄰苯二甲酸酯類與雙酚類的主要途徑為食入，而食品包材滲漏到食物中是主要的污染源之一。因此，本研究針對食品包材中的鄰苯二甲酸酯類與雙酚類進行了半定量分析。本研究之待測物為 12 種鄰苯二甲酸酯類與 6 種雙酚類，使用即時直接分析游離源 (direct analysis in real time, DART) 搭配 Waters Xevo TQ-XS 串聯式質譜儀進行半定量分析。本研究針對 DART 參數，包含游離源至質譜儀端入口的距離、錐電壓與游離源溫度進行了優化，12 種鄰苯二甲酸酯類與雙酚A二環氧甘油醚 (BADGE) 於正電模式下進行測定，參數為游離源至質譜儀端入口 30 毫米，錐電壓 400 V，游離源溫度 400℃。5 種雙酚類於負電模式下進行測定，參數為游離源至質譜儀端入口 33 毫米，錐電壓 300 V，游離源溫度 300℃。後續因高背景訊號以及基質抑制問題將鄰苯二甲酸二乙酯 (DEP)、鄰苯二甲酸二丁酯 (DBP)、鄰苯二甲酸二辛酯 (DNOP) 以及同分異構物鄰苯二甲酸二異丁酯 (DiBP) 與鄰苯二甲酸二 (二-乙基己基) 酯 (DEHP) 從分析中移除。25 cm2 之保鮮膜、食物保鮮袋與奶瓶樣本以剪刀裁剪成碎片，加入 10 mL 丙酮/甲醇 (1:1, v/v) 混合溶劑與 40 ?L 同位素標定內標準品 (isotope-labeled internal standards) 於 50℃ 以超音波萃取45 分鐘。取 1 mL樣本萃取液經 0.2-?m nylon過濾盤即可上機分析，分析時間為 2 分鐘。方法確效部分，待測物於聚苯乙烯 (polystyrene, PS) 生鮮托盤中受到的基質效應為 9–77%，萃取效率為 61–124%。待測物於 PVC 保鮮膜中受到的基質效應為 15–63% (不包含 DINP 0%、DEHT 149%、DIDP 250% 以及 DINCH 1152%)，萃取效率為 29–120% (不包含 DINP 0% 以及 DIDP 320%)。待測物於聚乙烯 (polyethylene, PE) 保鮮膜中受到的基質效應為 4–151%，萃取效率為 53–198% (不包含 BPA 231% 以及 BPS 337%)，基質效應與萃取效率普遍受到基質以及內生性待測物影響。同日與異日之相對標準偏差大部分在 20% 以內，然而半數待測物的準確度落在 80–120% 之外，除內生性待測物高於 120% 外，其餘待測物準確度低於 80%。待測物於 PS 生鮮托盤的偵測極限範圍與定量極限範圍分別為 0.01–4.4 ng/cm2 (不包含 DINP 13.0 ng/cm2) 與0.02–14.7 ng/cm2；於 PVC 保鮮膜的偵測極限範圍與定量極限範圍分別為 0.001–3.1 ng/cm2 (不包含 DEHT 64.3 ng/cm2) 與 0.004–10.4 ng/cm2 (不包含 DEHT 214 ng/cm2)；於 PE 保鮮膜的偵測極限範圍與定量極限範圍分別為 0.13–1.5 ng/cm2 (不包含 DEHT 7.3 ng/cm2 以及 DINP 7.6 ng/cm2) 與 0.42–5.0 ng/cm2 (不包含 DEHT 24.4 ng/cm2 以及 DINP 25.3 ng/cm2)。本研究開發之方法應用於實際採樣的九個食品包材樣本中，共測得三種鄰苯二甲酸酯類與三種雙酚類。雙 (2-乙基己基) 對苯二甲酸酯 (DEHT) 於 PVC 保鮮膜測得濃度為 177040 ng/cm2。環己烷-1,2-二羧酸二異壬酯 (DINCH) 與鄰苯二甲酸二異癸酯 (DIDP) 於 PVC 保鮮膜測得的濃度分別為 34.6 與 3.8 ng/cm2；於聚丙烯 (polypropylene, PP) 奶瓶測得的濃度分別為 0.64 與 0.56 ng/cm2；於聚苯? (polyphenylene sulfone resins ,PPSU) 奶瓶測得的濃度分別為 0.20 與 0.44 ng/cm2。雙酚 F 於 PPSU 奶瓶測得濃度為 2.8 ng/cm2。雙酚 S 於 PVC 保鮮膜與 PPSU 奶瓶測得濃度分別為 65.7 與 1.8 ng/cm2。雙酚 A 二環氧甘油醚於 PVC 保鮮膜與 PP 奶瓶測得濃度分別為 0.40 與 0.44 ng/cm2。本研究所測得之待測物換算至重量濃度單位多為 ppm 等級，其中雙 (2-乙基己基) 對苯二甲酸酯 (DEHT) 於 PVC 保鮮膜之濃度達到 13.5%。因此，應針對鄰苯二甲酸酯類與雙酚類之替代物進行更全面的健康效應評估，並進一步審視現行食品器具容器包裝衛生標準修改之適宜性，考慮將實際測得濃度較高之替代物納入規範標準中。

Phthalate esters (PAEs) are utilized as plasticizers for the production of polyvinyl chloride (PVC) goods and food packaging. Bisphenols (BPs) are extensively employed in producing the coating materials on the surface of tickets, receipts, and food cans. Both of them are endocrine-disrupting chemicals, which might cause a wide range of adverse health effects in animals and humans. Many countries have restricted the use of PAEs and BPs. Consequently, the consumption of their emerging alternatives has increased in the past few years. Studies have shown that PAEs and BPs can leach from containers to foodstuff. Thus, humans might get exposed to PAEs and BPs through ingestion. This study developed and validated an analytical method for semi-quantitative determining 12 PAEs and six BPs in food packing using direct analysis in real time (DART) ionization coupled with a tandem mass spectrometer (MS/MS). The DART parameters, including the distance between the ion source and the cone of the tandem mass spectrometer, grid voltage, and ion source temperature, were optimized in this study. 12 PAEs and BADGE were tested in positive mode, and the optimized parameters were 30 mm, 400 V, and 400℃; those for five BPs tested in the negative mode were 33 mm, 300 V, and 300℃, respectively. Due to high background noise and ion suppression problem of diethyl phthalate (DEP), dibutyl phthalate (DBP), di-n-octyl phthalate (DNOP), and their isomers diisobutyl phthalate (DiBP) and di(2-ethylhexyl) phthalate (DEHP), these compounds was excluded from the quantification of this study. In the sample preparation steps, 25 cm2 samples were cut into pieces, 10 mL of acetone/methanol (1:1, v/v) and 40 ?L of isotope-labeled internal standards were added to the samples and extracted by ultrasonic extraction at 50℃ for 45 minutes. 1-mL extracts of all samples were filtered through a 0.2-?m nylon filter for further instrument analysis. The analysis time for each sample was 2 minutes. In terms of method validation, the matrix effect of the analytes in polystyrene (PS) food packaging, PVC preservative film, and polyethylene (PE) preservative film were 9–77%, 15–63% (except for DINP 0%, DEHT 149%, DIDP 250%, and DINCH 1152%), 4–151%, respectively. The extraction efficiency of the analytes in PS food packaging, PVC preservative film, and PE preservative film were 61–124%, 29–120% (except for DINP 0% and DIDP 320%), 53–198% (except for BPA 231% and BPS 337%), correspondingly. The matrix effect and extraction efficiency were seriously affected by the matrixes and high levels of endogenous compounds. The precision (%RSD) for most analytes in three different matrices were lower than 20%. The accuracy (%bias) for half of the analytes were not within the range of ± 20% in three matrixes, most of which were lower than -20%; for those endogenous compounds were higher than 20%. The limit of detections (LODs) of the analytes in PS food packaging were 0.33–4.4 ng/cm2 (except for DINP 13.0 ng/cm2), and the limit of quantifications (LOQs) were 0.02–14.7 ng/cm2 (except for DINP 43.2 ng/cm2). The LODs of the analytes in PVC preservative film were 0.001–3.1 ng/cm2 (except for DEHT 64.3 ng/cm2), and the LOQs were 0.004–10.4 ng/cm2 (except for DEHT 214 ng/cm2). The LODs of the analytes in PE preservative film were 0.13–1.5 ng/cm2 (except for DEHT 7.3 ng/cm2 and DINP 7.6 ng/cm2), and the LOQs were 0.42–5.0 ng/cm2 (except for DEHT 24.4 ng/cm2 and DINP 25.3 ng/cm2). The developed method was applied to nine food packaging samples. DEHT was in the PVC preservative film sample at 177040 ng/cm2. DINCH and DIDP were detected in the PVC preservative film sample at 34.6 and 3.8 ng/cm2; in polypropylene (PP) infant formula bottle was at 0.64 and 0.56 ng/cm2; in polyphenylene sulfone resins (PPSU) infant formula bottle was at 0.20 and 0.44 ng/cm2, correspondingly. BPF was found in the PPSU infant formula bottle at 2.8 ng/cm2. BPS was 65.7 ng/cm2 in the PVC preservative film and 1.8 ng/cm2 in the PPSU infant formula bottle. BADGE was detected in the PVC preservative film and the PP infant formula bottle at 0.40 and 0.44 ng/cm2, sequentially. When converting the unit from per area of samples (ng/cm2) to per weight of samples (?g/g), the concentrations of most of the detected analytes were at ppm levels, and that of DEHT in the PVC preservative film was even up to the percentage level. Therefore, evaluating the health effects of the PAEs and BPs alternatives was a warrant, which could help further assessment of the suitability of adding new alternatives to the regulations on plastic food packaging.