The differentiation of 2,5-dimethoxy-N-(N-methoxybenzyl)phenethylamine (NBOMe) isomers using GC retention indices and multivariate analysis of ion abundances in electron ionization mass spectra

Abstract Synthetic phenethylamine derivatives known as 2,5-dimethoxy-N-(N-methoxybenzyl)phenethylamines (NBOMes) are a common class of novel psychoactive substances (NPS) that are causing many accidental deaths across the United States. Many derivatives are now banned at the federal and state levels, but such control requires reliable identification of the different positional isomers. This manuscript helps establish retention indices and characteristic ion ratios that can be used to distinguish between the positional isomers of 25C-NBOMe and 25I-NBOMe. This manuscript also provides additional support for the ortho effect as a reliable, general, fragmentation mechanism to differentiate positional isomers of NBOMes in electron ionization (EI) mass spectra. The retention indices and fragment ion abundances of the positional isomers of 25C-NBOMe and 25I-NBOMe were measured on two instruments using three different GC columns and parameters. The measured retention indices for the six compounds on three different 5% diphenyl columns are as follows: ortho-25C-NBOMe = 2614 ± 15; meta-25C-NBOMe = 2666 ± 13; para-25C-NBOMe = 2692 ± 13; ortho-25I-NBOMe = 2821 ± 16; meta-25I-NBOMe = 2877 ± 15; and para-25I-NBOMe = 2904 ± 12, where the errors represent the 95% confidence interval of the measurements. Principal component analysis (PCA) and canonical discriminant analysis (CDA) were used, respectively, to assess the variance and classification of NBOMe isomers based on the 15 most abundant ions relative to the base peak. The CDA classification accuracy for the six NBOMe compounds was 99.5% when the data set included spectra from three instrumental setups and the widest range of concentrations. Isomer classification was greater than 99.9% within an instrument and excluding low abundance spectra. These results support the use of chemometric approaches for the classification of unknown compounds, even when non-ideal lower-abundance spectra are used for classification.