| Title: | Machine learning directed discrimination of virgin and recycled poly(ethylene terephthalate) based on non-targeted analysis of volatile organic compounds |
| Author(s): | Li H; Wu X; Wu S; Chen L; Kou X; Zeng Y; Li D; Lin Q; Zhong H; Hao T; Dong B; Chen S; Zheng J; |
| Address: | "National Reference Laboratory for Food Contact Material (Guangdong), Guangzhou Customs Technology Center, Guangzhou 510075, China; School of Chemistry and Chemical Engineering, South China University of Technology, Guangzhou 510641, China. National Reference Laboratory for Food Contact Material (Guangdong), Guangzhou Customs Technology Center, Guangzhou 510075, China. Department of Medical Statistics, School of Public Health, Sun Yat-sen University, Guangzhou 510080, China. Key Laboratory of Product Packaging and Logistics, Packaging Engineering Institute, Jinan University, Zhuhai 519070, China. National Reference Laboratory for Food Contact Material (Guangdong), Guangzhou Customs Technology Center, Guangzhou 510075, China. Electronic address: marco.zhong@iqtc-fcm.com. National Reference Laboratory for Food Contact Material (Guangdong), Guangzhou Customs Technology Center, Guangzhou 510075, China. Electronic address: dongb@iqtc-fcm.com" |
| DOI: | 10.1016/j.jhazmat.2022.129116 |
| ISSN/ISBN: | 1873-3336 (Electronic) 0304-3894 (Linking) |
| Abstract: | "The use of non-decontaminated recycled poly(ethylene terephthalate) (PET) in food packages arouses consumer safety concerns, and thus is a major obstacle hindering PET bottle-to-bottle recycling in many developing regions. Herein, machine learning (ML) algorithms were employed for the discrimination of 127 batches of virgin PET and recycled PET (rPET) samples based on 1247 volatile organic compounds (VOCs) tentatively identified by headspace solid-phase microextraction comprehensive two-dimensional gas chromatography quadrupole-time-of-flight mass spectrometry. 100% prediction accuracy was achieved for PET discrimination using random forest (RF) and support vector machine (SVM) algorithms. The features of VOCs bearing high variable contributions to the RF prediction performance characterized by mean decrease Gini and variable importance were summarized as high occurrence rate, dominant appearance and distinct instrument response. Further, RF and SVM were employed for PET discrimination using the simplified input datasets composed of 62 VOCs with the highest contributions to the RF prediction performance derived by the AUCRF algorithm, by which over 99% prediction accuracy was achieved. Our results demonstrated ML algorithms were reliable and powerful to address PET adulteration and were beneficial to boost food-contact applications of rPET bottles" |
| Keywords: | Ethylenes Machine Learning Phthalic Acids Polyethylene Terephthalates/analysis/chemistry *Volatile Organic Compounds/analysis HS-SPME-GCxGC-QTOF-MS PET bottle-to-bottle recycling Random forest Support vector machine; |
| Notes: | "MedlineLi, Hanke Wu, Xuefeng Wu, Siliang Chen, Lichang Kou, Xiaoxue Zeng, Ying Li, Dan Lin, Qinbao Zhong, Huaining Hao, Tianying Dong, Ben Chen, Sheng Zheng, Jianguo eng Research Support, Non-U.S. Gov't Netherlands 2022/05/16 J Hazard Mater. 2022 Aug 15; 436:129116. doi: 10.1016/j.jhazmat.2022.129116. Epub 2022 May 11" |
