Spotlights: Untargeted Forensic Drug Detection, Burn Pit Smoke Inhalation, and Problems in the Friction Sensitivity Literature

Abstract

As new designer drugs emerge, forensic laboratories need broad screening tools that work well for all analytes in a highthroughput lab. This study investigated ultrahigh-performance liquid chromatography high-resolution mass spectrometry (UHPLC-HRMS) as a method for the untargeted screening of DEA-classified Tier 1 drugs of abuse and their metabolites. 1 The method used supported liquid extraction and data-independent acquisition.The HRMS method was evaluated using whole blood samples fortified with drugs of abuse at low and high concentrations. Performance was assessed according to the limits of detection, recovery, ion suppression or enhancement, and precision. 35 analytes were studied. The limits of detection (LODs) were higher than expected for six analytes: amphetamine, lorazepam, oxymorphone, and all three cannabinoids. Ionization at low concentrations ranged from 30.1−267.6%, and ionization at high concentrations ranged from 8.6−383.5%. Recovery for low concentrations ranged from 8.5−330.5%, with an average of 105.3%. Recovery at high concentrations ranged from 9.4− 127.5%, with an average of 82.7%. Carryover and interference were not detected.Additionally, the method was used to screen 166 blood samples from real DUI investigations and post-mortems. A range of substances were detected, with the most frequent being COOH-Δ9 THC (51 samples), fentanyl (40 samples), and methamphetamines (36 samples). The authors conclude that their method was fast, required few steps, and required only 100 μL of blood, although it was not suitable for some cannabinoids. this equation.A goal of this ongoing research was to build a benchmarked data set of energetic compounds for modeling. They suggest dividing compounds into four different hazardous classes, ranging from very highly sensitive species (e.g., most known primary explosives) to low sensitivity species. Existing data sets are remain much smaller than those used, for example, for machine learning.