Classification of illicit heroin by UPLC–Q-TOF analysis of acidic and neutral manufacturing impurities. (December 2015)

• Record Type: Journal Article
• Authors: Liu, Cuimei; Hua, Zhendong; Bai, Yanping
• Is Part Of: Forensic science international. Volume 257(2015)
• Page Start: 196
• Page End: 202
• Publication Date: 2015-12
• Subjects: Heroin -- Origin classification -- Acidic and neutral manufacturing impurities -- UPLC–Q-TOF -- PLS-DA; Medical jurisprudence -- Periodicals; Chemistry, Forensic -- Periodicals; Forensic Medicine -- Periodicals; Médecine légale -- Périodiques; Chimie légale -- Périodiques; Gerechtelijke geneeskunde; Gerechtelijke chemie; Gerechtelijke psychiatrie; Chemistry, Forensic; Medical jurisprudence; Electronic journals; Periodicals; Electronic journals; 614.1
• DOI: 10.1016/j.forsciint.2015.08.009
• Languages: English
• ISSNs: 0379-0738

Abstract

Highlights: Acidic and neutral impurities were analyzed by UPLC–Q-TOF for heroin profiling. 19 acidic and neutral impurities were found to be significant for classification. A PLS-DA classification model was successfully established and validated. Abstract: The illicit manufacture of heroin results in the formation of trace levels of acidic and neutral manufacturing impurities that provide valuable information about the manufacturing process used. In this work, a new ultra performance liquid chromatography–quadrupole-time of flight mass spectrometry (UPLC–Q-TOF) method; that features high resolution, mass accuracy and sensitivity for profiling neutral and acidic heroin manufacturing impurities was developed. After the UPLC–Q-TOF analysis, the retention times and m / z data pairs of acidic and neutral manufacturing impurities were detected, and 19 peaks were found to be evidently different between heroin samples from "Golden Triangle" and "Golden Crescent". Based on the data set of these 19 impurities in 150 authentic heroin samples, classification of heroin geographic origins was successfully achieved utilizing partial least squares discriminant analysis (PLS-DA). By analyzing another data set of 267 authentic heroin samples, the developed discrimiant model was validated and proved to be accurate and reliable.