To assess a workflow combining electron ionization (EI) and chemical ionization (CI) with gas chromatography coupled to high-resolution mass spectrometry (GC-HRMS) for toxicological screening.

The identification of new psychoactive substances (NPS) in biological samples remains an analytical challenge especially when context and seized products are not available. On one hand, the analytical platform GC-EI-MS allows to query a large number of database including numerous NPS compounds. On the other hand, this platform still suffer from sensitivity in scan mode, mass measurement accuracy and the extensive in-source fragmentation leads in most cases to the absence of the molecular ion. To circumvent these drawbacks, we propose to implement a two steps pipeline for the untargeted toxicological screening of NPS using a GC-HRMS platform with two complementary ion sources. The developed workflow was applied to a forensic case of a man found at home deadly.

Sample preparation was based on a liquid-liquid extraction performed on collected biological samples which included bile, lung and liver exudates as well as urine. The biological extracts were then analyzed on GC-HRMS Orbitrap® platform using sequentially EI and CI as ion sources. Source parameters were optimized for chemical ionization using gas calibrant. A data processing workflow were implemented to automatically processed GC-EI-HRMS data, including a deconvolution step, using MZmine 3.9.0 and Compound Discoverer 3.0 which allowed to perform database querying in the NIST environment. For GC-CI-HRMS, additional data processing steps using MZmine 3.9.0 were also developed.

The automatic data processing of GC-EI-HRMS along with database querying allowed to propose the identification of six NPS sparsely encountered namely fluoroethylamphetamine as well as five designer benzodiazepines, desalkylgidazepam, bromazolam, pyrazolam, fluclotizolam and deschloroetizolam, all detected in the whole investigated biological samples. Regarding the identification steps, taking as exemple fluoroethylamphetamine, the EI-HRMS data displayed several product ions diagnostic of parts of the molecule but the molecular ion and thus the isotopic pattern cannot be readily detected. As a second step, analyzing the extracts using GC-CI-HRMS in positive ion mode with the same analytical parameters enabled the rapid detection of fluoroethylamphetamine as an [M+H]⁺ ion at the same retention time as observed when performing analysis with EI. This complementary two-steps analysis have been performed for each proposed NPS. In addition to the identification of NPS, opioid substitution medication, including methadone and its metabolite EDDP, were also detected, as well as antiretrovirals, including raltegravir, abacavir, and lamivudine.

Using our two-step analysis based on GC-HRMS with complementary ion sources and automatic data processing, it allowed to detect in a single case six unusual NPS belonging to amphetamine and designer benzodiazepine family. While the first step using EI is useful to screen and automatically propose identification of NPS, the second step using CI allows to confirm the exact mass of the molecular ion as well as the isotopic pattern. It thus allowed a confidence identification of NPS, which are not commonly encountered such as molecules detected in this case. Regarding the forensic context, it allowed to confirm the consumption by the deadly man of NPS. To the best of our knowledge, this is the first time that such a cocktail of compounds belonging to amphetamine and benzodiazepine has been detected. This study is the first to propose a combined workflow for toxicological screening of NPS using GC-HRMS with complementary ion sources and open new insights in both clinical and forensic context.