Since its introduction, gas chromatography (GC) coupled to vacuum ultraviolet spectrophotometry (VUV) has been shown to complement mass spectrometry (MS) for materials such as petrochemicals, explosives, pesticides, and drugs. In forensic chemistry, opioids are commonly encountered but rarely are the samples pure. This work focuses on GC-VUV analysis applied to naturally occurring (e.g., morphine), semi-synthetic (e.g., heroin), and synthetic (fentanyl) opioids as well as common adulterants and diluents (e.g., lidocaine and quinine). The specificity of the VUV spectra were examined visually as well as via descriptive statistical methods (e.g., correlation coefficients and sums of square residuals). Multivariate pattern recognition techniques (principal component analysis and discriminant analysis (DA)) were used to prove the opioid spectra can be reliably differentiated. The accuracy of the DA model was 100% for a test set of VUV spectra. Finally, three "street" heroin samples were analyzed to show "real-world" performance for forensic analyses. These samples contained adulterants such as caffeine, as well as by-products of heroin manufacture.
Keywords: Drugs; GC-VUV; Gas chromatography; Opioids; Vacuum ultraviolet.