米及黃豆製品中嘉磷塞、固殺草及其主要代謝物同步分析方法之建立及臭氧處理對其殘留之影響

本研究建立了以氣相層析儀附脈衝式火焰光度檢出器 (gas chromatograph-pulsed flame photometric detector, GC-PFPD)進行米及黃豆製品中嘉磷塞 [N-(phosphonomethyl)glycine]、固殺草 [DL-homoalanine-4-yl-(methyl) phosphinic acid]及其主要代謝物，AMPA (aminomethylphosphonic acid)及3-MPPA [3-(methylphosphinico)propionic acid]之同步定量方法及串聯質譜分析 (tandem mass spectrometry) 確認方法。檢體前處理結合AG1-X8陰離子交換層析、TMOA (trimethyl orthoacetate)衍生化反應及矽酸鎂固相萃取匣 (Florisil cartridge)淨化，在DB-608毛細管之分離下，可於14 min內完成GC之同步分析。於米、黃豆芽及黃豆檢體中外添加 (spike) 0.5及2.5 ppm殺草劑及其代謝物，3-MPPA、AMPA、嘉磷塞及固殺草之回收率分別為83~96、67~88、72~103及95~119 %，三重複之變異係數 (coefficient of variation, CV)均小於10 %。方法中3-MPPA、AMPA、嘉磷塞及固殺草之最低偵測極限 (limit of detection, LOD)分別為0.01、0.03、0.02及0.02 ppm。在氣相層析質譜分析部分，比較電子離子化 (electron ionization, EI)質譜法 (EI-MS)與化學離子化 (chemical ionization, CI)質譜法 (CI-MS)應用於此四種成分之分析，結果以CI-MS法較不易受基質干擾且感度較佳，於0.5~10 ng間線性關係良好。進一步建立了化學離子化串聯質譜分析法(CI-MS/MS)，以鑑別食品中3-MPPA、AMPA、嘉磷塞及固殺草衍生物之最適條件，分別以m/z 181、m/z 182、m/z 254及m/z 252為母離子，碰撞誘導解離 (collision-induced dissociation, CID)能量分別為0.5、0.3、0.3及0.5 volt，其鑑定性及偵測感度皆相當良好，最低偵測極限分別可達0.01、0.02、0.005及0.02 ppm。以嘉磷塞152 ppm溶液浸漬6 hr之黃豆為培芽材料，仍可在5日內培育 (20~23℃)出與空白組相當大小之黃豆芽。培芽期間，嘉磷塞總量呈下降趨勢，並由種子 (子葉)快速部分轉移至芽體 (下胚軸)部分，且以芽尖之濃度最高（0.52±0.02 ppm），並有少量AMPA (~0.1 ppm)生成。抽購市售黃豆、黃豆芽及豆漿計27件檢體，進行嘉磷塞、固殺草、AMPA及3-MPPA殘留量之分析，黃豆芽3件及豆漿2件檢體皆為未檢出；黃豆檢體部分，發現8件非基因改造黃豆皆未檢出嘉磷塞、AMPA、固殺草或3-MPPA，14件含基因改造黃豆之檢體則皆檢出嘉磷塞及AMPA，嘉磷塞檢出範圍為0.23~0.94 ppm，平均值0.59 ppm，AMPA檢出範圍為0.10~0.34 ppm，平均值0.24 ppm。以不同之臭氧劑量進行水溶液中嘉磷塞、固殺草、愛殺松 (ethion)及百滅寧 (permethrin)農藥之降解，並分別以標準品及農藥製劑進行其消褪曲線之測試，發現在製劑成分共存下，農藥之消褪速率較為緩慢。四種農藥之氧化降解速率，由快而慢依序為：百滅寧＞愛殺松＞嘉磷塞＞固殺草。起始濃度為50 mg/mL之農藥，在臭氧劑量為18 mg/min、處理60 min下，百滅寧、愛殺松及嘉磷塞之降解消褪率分別為98、61及32 %，固殺草則幾乎未消褪。在臭氧劑量為32 mg/min之臭氧化 (ozonation)期間，由氣相層析圖中發現，百滅寧及愛殺松皆先後出現了兩個降解產物，嘉磷塞及固殺草則分別出現了AMPA及3-MPPA，但含量都不高 (0.9~2.5 mg/mL)。

關鍵字

米；黃豆；臭氧化；固殺草；殘留分析；氣相層析；嘉磷塞

並列摘要

Procedures were developed for simultaneous quantification of glyphosate [N-(phosphonomethyl)glycine] and glufosinate [DL-homoalanine-4-yl-(methyl) phosphinic acid], and their major metabolites, AMPA (aminomethylphosphonic acid) and 3-MPPA [3-(methylphosphinico)propionic acid], in rice and soybean products by gas chromatograph (GC) equipped with a pulsed flame photometric detector (PFPD). Sample pretreatments, including anion-exchange chromatography, TMOA (trimethyl orthoacetate) derivatization and Florisil cleanup, were favorable for the GC-PFPD analysis. Four types of derivatives spiked in rice and soybean products matrices were eluted, reaching a base-line separation, in a sequence of 3-MPPA, AMPA, glyphosate, and glufosinate within 14 min using a DB-608 capillary column. Recoveries of 3-MPPA, AMPA, glyphosate and glufosinate (0.5, 2.5 ppm) spiked in sample matrices were determined to be 83 to 96, 67 to 88, 72 to 103, and 95 to 119 %, respectively, while the coefficient of variation (CV) was determined to be less than 10 % in three repeated determinations. The limit of detection (LOD) for 3-MPPA, AMPA, glyphosate, and glufosinate in sample matrices was 0.01, 0.03, 0.02, and 0.02 ppm, respectively. Comparison of EI-MS (electron ionization-mass spectrometry) and CI-MS (chemical ionization-mass spectrometry) results for TMOA derivatized 3-MPPA, AMPA, glyphosate, and glufosinate revealed that better responce and less matrix interference were obtained by CI-MS method. Analytical conditions for CI-MS/MS detection were optimized, and m/z 181, m/z 182, m/z 254, and m/z 252 were chosen as parent ions and 0.5, 0.3, 0.3, and 0.5 volt were chosen as the CID (collision-induced dissociation) excitation amplitudes for 3- MPPA, AMPA, glyphoste, and glufosinate derivative, respectively. The optimized CI-MS/MS method was highly sensitive for the confirmation of these herbicides and their major metabolites in foods. LOD for 3-MPPA, AMPA, glyphosate and glufosinate in sample matrices was 0.01, 0.02, 0.005, and 0.02 ppm, respectively. Soybeans incubated in 152 ppm glyphosate solution for 6 hr before cultivation grew up to sprouts with normal size in 5-day cultivation (20~23℃). Quantification of herbicides revealed that glyphosate was transported fast from seeds (cotyledons) to sprouts (hypocotyls) and then concentrated in sprout tips with a low amount of AMPA formation during sprouting. Twenty-seven commercial samples, including 22 soybean, 3 soybean sprout, and 2 soybean milk samples, from local markets were analyzed for 3-MPPA, AMPA, glyphosate, and glufosinate residues. None of the above four analytes was detected in soybean sprout, soybean milk, and Non-GM (genetically modified) soybean samples. However, all 14 GM soybean samples were found to contain glyphosate, ranging from 0.23 to 0.94 ppm (mean value of 0.59 ppm), and AMPA residues, ranging from 0.10 to 0.34 ppm (mean value of 0.24 ppm). Kinetics of ozone-induced degradation of glyphosate, glufosinate, ethion, and permethrin in respective aqueous solution were determined for both reagent grade and commercial products. It was found that pesticides degraded faster in the absence than in the presence of formulating agents. At an initial concentration of 50 mg/mL, permethrin, ethion, and glyphosate reduced 98, 61, and 32 %, respectively, while glufosinate remained almost intact, after 60-min ozonation at a dosage of 18 mg/min. Two intermediates were found in GC chromatograms for both ethion and permethrin, whereas AMPA and 3-MPPA were detected from their parent glyphosate and glufosinate, respectively, during ozonation at a dosage of 32 mg/min.