To demonstrate the generation of mass voltammograms from processed feature lists, the oxidation behavior of the SCRA MDMB‐4en‐PINACA was studied using a mixed standard solution. The solution contained the analyte of interest, MDMB‐4en‐PINACA, alongside another SCRA, namely, PX‐2. Figure 3A depicts the resulting raw data in the form of a mass voltammogram generated by the newly developed software tool introduced above. Two intense signals can be observed that both decrease in intensity with increasing applied oxidation potential. Those signals can be assigned by their m/z to the protonated species of the analyte of interest, MDMB‐4en‐PINACA, as well as to the simulated interference, PX‐2. Additionally, several less intense signals can be observed that form at different potentials, therefore corresponding to different TPs. With both MDMB‐4en‐PINACA and PX‐2 being oxidized simultaneously in the electrochemical cell, it is not immediately possible to assign the TPs to their corresponding parent compound.

Here, the possibility to process the data by making use of the built‐in data workflows in mzmine comes in beneficially. As a reference, a second solution containing solely PX‐2 was oxidized under the same conditions as the mixed standard solution. By subtracting the signals from this reference solution from the data acquired for the mixed standard solution, only signals corresponding to MDMB‐4en‐PINACA remain. Due to the synergistic interplay between the newly developed software module and the established mzmine software platform, it is possible to generate a mass voltammogram from the resulting data after subtraction. Using this workflow, a mass voltammogram for MDMB‐4en‐PINACA can be generated from the mixed standard solution without the need for a pure analytical standard of MDMB‐4en‐PINACA. This approach can also be applied when a novel NPS is discovered as part of a mixture of other, already established NPS. Because there are typically reference standards available for the established NPS, the NPS mixture can be directly used for the simulation of the metabolism of the newly discovered NPS. A previous purification of the NPS mixture would not be necessary as long as analytical reference standards are available for all adulterants present in the mixture. Although, this approach could be limited by the complexity of the NPS mixture. Here, the accuracy of the background subtraction in cases of mixtures of multiple structurally similar compounds needs to be further addressed in the future. Simultaneous electrochemical transformation of structurally similar compounds could result in interfering TPs possibly impairing the background subtraction process.

Figure 3B depicts the resulting mass voltammogram where only signals corresponding to the analyte of interest, MDMB‐4en‐PINACA, and its TPs are observed. The signals corresponding to PX‐2 were removed. The two major TPs are both formed at a potential of around 1700 mV vs. Pd/H₂ pRE. According to their m/z of 374.2078 and 388.1873, respectively, they are referred to as TP374 and TP388. Detected accurate m/z, proposed ion formulae, exact m/z, and resulting mass deviations (Δm/z) are summarized in Table 1.

For TP374, a mass difference of 15.9949 Da to MDMB‐4en‐PINACA was observed. This, together with the predicted ion formula shown in Table 1, suggests that TP374 corresponds to a hydroxylated product of MDMB‐4en‐PINACA. Accordingly, for TP388, a mass difference of 29.9744 Da to MDMB‐4en‐PINACA was observed. The addition of two oxygen atoms as well as the loss of two hydrogen atoms result in a theoretical mass difference of 29.9742 Da. This suggests that TP388 corresponds to a doubly hydroxylated and dehydrogenated product of MDMB‐4en‐PINACA.

Figure 4 depicts the HRMS/MS spectrum for TP374 as well as the proposed fragmentation pattern. Fragment spectra were acquired alongside the mass voltammogram by DDA. The depicted spectrum was obtained by summation of HRMS/MS spectra with collision energies of 15 and 30 eV, respectively. The signal at m/z 356.1980 corresponds to the neutral loss of a water molecule supporting the previous assumption of TP374 being a hydroxylated product of MDMB‐4en‐PINACA. The signal at m/z 229.0979 can be explained through cleavage of the amide bond with the fragment ion still carrying the hydroxy group from the electrochemical hydroxylation. This suggests the hydroxy group being located in either the indazole core structure or the pentene side chain. Furthermore, the signal at m/z 230.1291 can be explained through an initial loss of the methyl formate moiety resulting in a fragment ion of m/z 314.1863. Again, the fragment ion still contains the hydroxy group supporting the previous thesis of localization. A following loss off the pentene side chain together with this hydroxy group results in the fragment ion of m/z 230.1288. Finally, the signal at m/z 290.1506 shows the loss of the pentene side chain together with the hydroxy group. Altogether, this confines the localization of the electrochemical hydroxylation to the pentene side chain.

Figure 5 depicts the HRMS/MS spectrum for TP388 as well as the proposed fragmentation pattern. Fragment spectra were acquired alongside the mass voltammogram by DDA. The depicted spectrum was obtained by summation of HRMS/MS spectra with collision energies of 15 and 30 eV, respectively. The signal at m/z 243.0769 corresponds to the loss of the amide side chain which is unchanged in comparison to the MDMB‐4en‐PINACA precursor structure. This suggests that the electrochemical modifications are either located in the indazole core structure or the pentene side chain. The signal at m/z 260.1038 further confines the localization of the electrochemical modifications. This signal can be explained through an initial loss of the methyl formate moiety resulting in a fragment ion of m/z 328.1656. A following loss of the pentene side chain results in the fragment ion of m/z 260.1030. The electrochemical modifications still being present in this fragment ion suggest that they are localized in the core indazole structure. Earlier, the mass difference of 29.9742 Da between TP388 and MDMB‐4en‐PINACA suggested that TP388 is a twofold hydroxylated and dehydrogenated product of MDMB‐4en‐PINACA. Together with the localization of these modifications in the indazole core structure, a quinoid structure can be proposed for TP388. This assumption is further supported by the absence of signals in the HRMS/MS spectrum that correlate to the neutral loss of water. Neutral loss of water would be expected if free hydroxy groups were formed but not in the case of the formation of a quinoid system. However, the absolute configuration of the quinoid system cannot be determined based on these data and therefore remains unclear.

Watanabe et al. studied the metabolism of MDMB‐4en‐PINACA by incubating the drug with human hepatocytes as well as human liver microsomes and by analysis of authentic human in vivo blood and urine samples [30]. In total, they identified 32 different metabolites, two of which were found in the in vivo samples. Among these metabolites, two suggested urinary markers were a hydroxy metabolite as well as a metabolite formed by ester hydrolysis. Molecular structures for both metabolites are shown in Figure 6. On the one hand, the hydroxy metabolite found by Watanabe et al. likely corresponds to TP374 found in our study, demonstrating the ability of electrochemical metabolism simulation to form TPs which are also found as metabolites in conventional in vivo and in vitro studies. On the other hand, the metabolite formed by ester hydrolysis was not generated electrochemically. This is likely due to the ester hydrolysis not being a redox reaction, demonstrating the limitations of metabolism simulation by EC‐MS.

Furthermore, the electrochemically generated TP388 with a quinoid structure was not found by Watanabe et al. Because of the high redox activity of most quinones, a reaction of TP388 with endogenous molecules like peptides and proteins would be very likely in case of an in vivo formation of TP388. This could be a possible reason for the absence of TP388 as a metabolite in the work of Watanabe et al. In our electrochemical metabolism simulation, the well‐controlled chemical environment and absence of endogenous molecules prevent such reactions enabling the detection of highly reactive products like quinones. However, further conventional metabolism studies looking specifically for possible adducts of TP388 and endogenous molecules are needed to confirm the in vivo formation of TP388 as a metabolite. This is of interest because quinoid metabolites are known to induce severe oxidative stress in cells by oxidation of cellular macromolecules, thus causing acute cytotoxicity [31]. Therefore, the possible formation of TP388 as an in vivo metabolite could lead to the adverse and toxic side effects observed after the recreational use of MDMB‐4en‐PINACA [14].