Identification of new psychoactive substances and their metabolites using non-targeted detection with high-resolution mass spectrometry through diagnosing fragment ions/neutral loss analysis

Target and Nontarget Screening to Support Capacity Scaling for Substance Use Assessment through a Statewide Wastewater Surveillance Network in New York

“…C 12 H 17 NO 2 ( m / z 208.1332 for [M + H]+; ΔMass = −0.32 ppm) was tentatively identified as 4-methoxy- N , N -dimethylcathinone (also known as 4-methoxy- N , N -DMC; Figure f) or its isomer based on the characteristic neutral losses of C 4 H 11 N (e.g., N , N -dimethylethanamine) and C 5 H 11 NO (e.g., 2-(dimethylamino)­propanal) from the precursor ion to form the diagnostic (4-formylphenyl)­(methylene)­oxonium ion C 8 H 7 O 2 + (i.e., m / z 135.0439; ΔMass = −1.11 ppm) found for methoxy cathinones and the 4-methoxybenzene-1-ylium ion C 7 H 7 O + (i.e., m / z 107.0493; ΔMass = 1.40 ppm), respectively. C 12 H 15 NO 3 ( m / z 222.1124 for [M + H]+; ΔMass = −0.41 ppm) was tentatively identified as dimethylone (also known as bk-MDDMA; Figure S12) or its isomer based on the characteristic neutral losses of C 4 H 11 N and C 5 H 11 NO from the precursor ion to form the diagnostic (benzo­[ d ]­[1,3]­dioxol-5-ylmethylidyne)­oxonium ion C 8 H 5 O 3 + (i.e., m / z 149.0230; ΔMass = −2.28 ppm) found for methylenedioxy cathinones and the benzo­[ d ]­[1,3]­dioxol-5-ylium ion C 7 H 5 O 2 + (i.e., m / z 121.0284; ΔMass = 0.08 ppm), respectively. C 11 H 15 NO 2 ( m / z 194.1174 for [M + H]+; ΔMass = −0.57 ppm) was tentatively identified as N -methyl homarylamine (Figure S13) or its isomer based on the characteristic neutral losses of C 2 H 7 N (e.g., dimethylamine) and C 2 H 7 N plus CO (i.e., C 3 H 7 NO) from the precursor ion that led to the formation of two diagnostic ions, 1-(benzo­[ d ]­[1,3]­dioxol-5-yl)­ethan-1-ylium ion C 9 H 9 O 2 + (i.e., m / z 149.0595; ΔMass = −1.68 ppm) and benzo­[ d ]­[1,3]­dioxol-5-ylmethylium ion C 8 H 7 O 2 + (i.e., m / z 135.0442; ΔMass = 0.81 ppm), as typically observed for methylenedioxy phenethylamines .…”

“…C 12 H 15 NO 3 ( m / z 222.1124 for [M + H]+; ΔMass = −0.41 ppm) was tentatively identified as dimethylone (also known as bk-MDDMA; Figure S12) or its isomer based on the characteristic neutral losses of C 4 H 11 N and C 5 H 11 NO from the precursor ion to form the diagnostic (benzo­[ d ]­[1,3]­dioxol-5-ylmethylidyne)­oxonium ion C 8 H 5 O 3 + (i.e., m / z 149.0230; ΔMass = −2.28 ppm) found for methylenedioxy cathinones and the benzo­[ d ]­[1,3]­dioxol-5-ylium ion C 7 H 5 O 2 + (i.e., m / z 121.0284; ΔMass = 0.08 ppm), respectively. C 11 H 15 NO 2 ( m / z 194.1174 for [M + H]+; ΔMass = −0.57 ppm) was tentatively identified as N -methyl homarylamine (Figure S13) or its isomer based on the characteristic neutral losses of C 2 H 7 N (e.g., dimethylamine) and C 2 H 7 N plus CO (i.e., C 3 H 7 NO) from the precursor ion that led to the formation of two diagnostic ions, 1-(benzo­[ d ]­[1,3]­dioxol-5-yl)­ethan-1-ylium ion C 9 H 9 O 2 + (i.e., m / z 149.0595; ΔMass = −1.68 ppm) and benzo­[ d ]­[1,3]­dioxol-5-ylmethylium ion C 8 H 7 O 2 + (i.e., m / z 135.0442; ΔMass = 0.81 ppm), as typically observed for methylenedioxy phenethylamines . C 20 H 27 FN 2 O 3 ( m / z 363.2078 for [M + H]+; ΔMass = 0.01 ppm) was tentatively identified as 5-fluoro AMB-PICA (also known as MMB-2201; Figure S14) or its isomer based on the detection of the diagnostic (1 H -indol-3-yl)­(oxo)­methylium ion C 9 H 6 NO + (i.e., m / z 144.0440; ΔMass = −2.71 ppm) found for indole-3-carboxamide-based synthetic cannabinoids and the neutral loss of C 6 H 13 NO 2 (e.g., methyl 2-amino-3-methylbutanoate) from the precursor ion to form the (1-(5-fluoropentyl)-1 H -indol-3-yl)­(oxo)­methylium ion C 14 H 15 FNO + (i.e., m / z 232.1130; ΔMass = −1.42 ppm).…”

“…To extend library search beyond mzCloud and MassBank , , the search mzVault node was set to import the High-Resolution Mass Spectral Libraries for Opioid Analysis curated by the CDC and the HighResNPS consensus library (October 2023 version) . Given that psychoactive substances typically contain characteristic structural cores, the compound class scoring node was implemented to enable the filtering of diagnostic fragment ions for fentanyl analogs, synthetic cannabinoids, synthetic cathinones, phenyl-substituted phenethylamines, arylcyclohexylamines, and indolealkylamines (e.g., lysergamides and tryptamines) . To complement class coverage scoring, the search neutral loss node was also configured to enable the simultaneous filtering of neutral losses commonly observed for these substance classes .…”

“… 25 Given that psychoactive substances typically contain characteristic structural cores, 26 the compound class scoring node was implemented to enable the filtering of diagnostic fragment ions for fentanyl analogs, synthetic cannabinoids, synthetic cathinones, phenyl-substituted phenethylamines, arylcyclohexylamines, and indolealkylamines (e.g., lysergamides and tryptamines). 27 To complement class coverage scoring, the search neutral loss node was also configured to enable the simultaneous filtering of neutral losses commonly observed for these substance classes. 27 With this workflow, mass spectral features that met all the following criteria were prioritized for inspection: a peak rating of >5, a mass accuracy tolerance of 5 ppm, a mzCloud and/or mzVault best match score of >60, and a retention time within the 95% confidence interval predicted by the Log P –retention time relationship established via the analysis of 432 compounds (e.g., pharmaceuticals, pesticides, personal care and household chemicals, industrial additives, and their transformation products) covering a range of polarities ( Figure S1 ).…”

“… 27 To complement class coverage scoring, the search neutral loss node was also configured to enable the simultaneous filtering of neutral losses commonly observed for these substance classes. 27 With this workflow, mass spectral features that met all the following criteria were prioritized for inspection: a peak rating of >5, a mass accuracy tolerance of 5 ppm, a mzCloud and/or mzVault best match score of >60, and a retention time within the 95% confidence interval predicted by the Log P –retention time relationship established via the analysis of 432 compounds (e.g., pharmaceuticals, pesticides, personal care and household chemicals, industrial additives, and their transformation products) covering a range of polarities ( Figure S1 ). Mass spectral features of interest were either confirmed by reference standards (i.e., confidence level 1), identified as probable structures (i.e., level 2), or assigned as tentative candidates (i.e., level 3) when applicable.…”