流过式介质阻挡放电电离质谱法结合随机森林模型快速鉴别棉、涤纶纺织品  

Rapid Identification of Cotton and Polyester Textiles Using Flowthrough Dielectric Barrier Discharge Ionization Mass Spectrometry Combined with Random Forest Model

作　　者：尚宇瀚 葛运程 高海燕 孟宪双[1] 马强[1] SHANG Yu-han;GE Yun-cheng;GAO Hai-yan;MENG Xian-shuang;MA Qiang(Chinese Academy of Inspection and Quarantine,Beijing 100176,China;School of Chemical Engineering,Ocean and Life Sciences,Dalian University of Technology,Panjin 124221,China;Inner Mongolia Autonomous Region Administration of Market Supervision Evaluation and Inspection Center,Hohhot 010090,China)

机构地区：[1]中国检验检疫科学研究院,北京100176 [2]大连理工大学化工海洋与生命学院,辽宁盘锦124221 [3]内蒙古自治区市场监督管理审评查验中心,内蒙古呼和浩特010090

出　　处：《分析测试学报》2024年第6期883-890,共8页Journal of Instrumental Analysis

基　　金：国家重点研发计划(2022YFF0607204);2022年度中央引导地方科技发展资金项目(2022ZY0151)。

摘　　要：为实现对棉、涤纶纺织品的快速准确鉴别,开发了电烙铁灼烧-流过式介质阻挡放电电离质谱法。结果表明,电烙铁灼烧-流过式介质阻挡放电电离质谱法无需样品前处理,操作简便,单次数据采集用时仅为4~5 s。使用棉、涤纶与棉涤混纺纺织品样品的质谱指纹图谱数据,建立了基于随机森林算法的预测模型。优化后的随机森林模型五折交叉验证与袋外验证的误判率分别为2.50%和2.52%。基于袋外验证混淆矩阵,求得该模型的预测准确率为97.5%,对棉与涤纶样品的精确率、召回率和F1分数均高于95%,满足快速鉴别要求。该方法可为棉、涤纶纺织品真伪鉴别与品质评价提供技术参考。To achieve the rapid and precise identification of cotton and polyester textiles,an ana⁃lytical method utilizing an electric soldering iron for cauterization combined with soft ionization by chemical reaction in transfer-mass spectrometry(SICRIT-MS)was developed.The SICRIT ion source could ionize gaseous analytes by generating a plasma torch at the inlet of the mass spectrome⁃ter through dielectric barrier discharge ionization and maintaining the negative pressure inside the mass spectrometer.The results indicated that the SICRIT-MS method did not necessitate sample pretreatment,was easy to operate,and required only 4-5 s for a single data acquisition.Subse⁃quently,a prediction model based on the random forest algorithm was established using the ac⁃quired mass spectrometry(MS)fingerprint data.The random forest model was optimized using the GridSearchCV function from the Scikit-learn library and underwent validation both through five-fold cross-validation and out-of-bag validation,yielding misclassification rates of 2.50%and 2.52%,respectively.According to the out-of-bag validation confusion matrix,the model achieved a pre⁃diction accuracy of 97.5%,with precision,recall,and F1 score for cotton and polyester samples all exceeding 95%,thereby meeting the requirements for rapid identification.This method can serve as a technical reference for the authentication and quality evaluation of cotton and polyester textiles.

关 键 词：流过式介质阻挡放电电离质谱 随机森林模型 棉 涤纶 快速鉴别 

分 类 号：O657.63[理学—分析化学] TS107[理学—化学]