Part 3: Solid phase extraction of Russian VX and its chemical attribution signatures in food matrices and their detection by GC-MS and LC-MS

• Published in: Talanta (Oxford) 2018-08, Vol.186 (C), p.607-614
• Main Authors: Williams, Audrey M., Vu, Alexander K., Mayer, Brian P., Hok, Saphon, Valdez, Carlos A., Alcaraz, Armando
• Format: Article
• Language: English
• Subjects: Animals; Chemical attribution signatures; Chemical warfare agent; Chemical Warfare Agents - chemistry; Chemical Warfare Agents - isolation & purification; Chromatography, Liquid; Drinking Water - chemistry; Food Analysis; Food Contamination - analysis; Forensic attribution; Gas Chromatography-Mass Spectrometry; Humans; Infant; Infant Food - analysis; INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY; Machine learning; Mass Spectrometry; Milk - chemistry; Organothiophosphorus Compounds - chemistry; Organothiophosphorus Compounds - isolation & purification; Russian VX; Solid Phase Extraction
• ISSN: 0039-9140; 1873-3573
• DOI: 10.1016/j.talanta.2018.03.044

Chemical attribution signatures indicative of O-isobutyl S-(2-diethylaminoethyl) methylphosphonothioate (Russian VX) synthetic routes were investigated in spiked food samples. Attribution signatures were identified using a multifaceted approach: Russian VX was synthesized using six synthetic routes and the chemical attribution signatures identified by GC-MS and LC-MS. Three synthetic routes were then down selected and spiked into complex matrices: bottled water, baby food, milk, liquid eggs, and hot dogs. Sampling and extraction methodologies were developed for these materials and used to isolate the attribution signatures and Russian VX from each matrix. Recoveries greater than 60% were achieved for most signatures in all matrices; some signatures provided recoveries greater than 100%, indicating some degradation during sample preparation. A chemometric model was then developed and validated with the concatenated data from GC-MS and LC-MS analyses of the signatures; the classification results of the model were > 75% for all samples. This work is part three of a three-part series in this issue of the United States-Sweden collaborative efforts towards the understanding of the chemical attribution signatures of Russian VX in crude materials and in food matrices. [Display omitted] •Sampling method for forensic attribution of Russian VX in food is presented.•Gradient boosted machines model variance in Russian VX extracted from food.•Synthetic pathway used to synthesize Russian VX can be identified using GBM model.