水道水中農薬のGC/MSスクリーニング分析におけるクロマトグラム解析の誤差要因の分析

小林 憲弘; 土屋 裕子; 高木 総吉; 吉田 仁; 大窪 かおり; 北原 健一; 坂本 晃子; 木下 輝昭; 仲野 富美; 橋本 博之; 古川 浩司; 粕谷 智浩; 岩間 紀知; 平林 達也; 小嶋 隼; 林 幸範; 古口 健太郎; 五十嵐 良明

doi:10.5985/jec.33.26

報文

水道水中農薬のGC/MSスクリーニング分析におけるクロマトグラム解析の誤差要因の分析

小林憲弘, 土屋裕子, 高木総吉, 吉田仁, 大窪かおり, 北原健一, 坂本晃子, 木下輝昭, 仲野富美, 橋本博之, 古川浩司, 粕谷智浩, 岩間紀知, 平林達也, 小嶋隼, 林幸範, 古口健太郎, 五十嵐良明

著者情報

ジャーナルオープンアクセス

2023 年 33 巻 p. 26-40

DOI https://doi.org/10.5985/jec.33.26

詳細

抄録

In order to adopt "screening analysis" in drinking water quality testing, which performs qualitative and quantitative analysis based on information registered in a database without measuring standard materials at the time of each testing, a validation test was conducted to evaluate the difference on the chromatogram data analysis by analysts. In the validation test, chromatograms of 6 water samples were analyzed using the same database and software by 16 analysts in 13 laboratories. Peak identification was performed based on Retention time (RT), (qualifier and target ion) QT ratio, and degree of matching of mass spectra, and if necessary, peak shapes in water samples were compared with those in calibration curve samples. Large differences were observed in the detection of pesticides among the analysts. There were two main reasons for the disagreement. The first is that the judgement of peak detection at low concentrations near the quantification limit of each pesticide, particularly below the concentration at three times of quantification limit, differed depending on the analyst. The other reason was that some analysts made clear misidentifications and identified wrong peaks close to the predicted retention time, despite the QT ratio and mass spectrum didn't match that in the database. In addition, there were some cases where the quantified values were reported by misidentifying the peaks of degradation products and metabolites, even when the peak detection was correctly judged. In order to use screening analysis practically by drinking water quality testing laboratories, it is important to conduct a validation test like this study in advance to use screening analysis and to coordinate criteria for peak identification and quantification among data analysts.

責任著者(Corresponding author)

J-STAGEへの登録はこちら（無料）