Target compounds were 25C–NBOMe, 25D–NBOMe, 25E–NBOMe, amphetamine, dimethoxymethamphetamine (DMMA), methyl diethanolamine (MDEA), 3,4–methylenedioxymethamphetamine (MDMA), methamphetamine, N–Methyl–2–aminoindane (N–methyl–2–AI), p–methoxyamphetamine (PMA), p–methoxy–N–methylamphetamine (PMMA), cocaine, benzoylecgonine (BE), co-caethylene, norcocaine, ketamine, norketamine, lysergic acid diethylamide (LSD), 2–oxo–3–hydroxy–LSD, mescaline, cathinone, 3,4–methylenedioxypyrovalerone (MDPV), 4–fluoromethcathinone (flephedrone), buphedrone, butylone, ethylone, mephedrone, methcathinone, methedrone, methylone, pentedrone, butyrfentanyl, fentanyl, furanylfentanyl, valerylfentanyl, methadone, and EDDP. Of the 37 target compounds, 32 NPS were purchased from Cerilliant (Texas, USA), 5 NPS (25D–NBOMe, 25E–NBOMe, DMMA, PMMA, and N–methyl–2–AI) received from the Ministry of Food and Drug Safety (MFDS) of Korea. The internal standards included 25C–NBOMe–d₃, amphetamine–d₆, BE–d₈, butylone–d₃, cathinone–d₅, cocaethylene–d₈, cocaine–d₃, EDDP–d₃, ephedrine–d₃, ethylone–d₅, fentanyl–d₅, ketamine–d₄, LSD–d₃, MDMA–d₅, mephedrone–d₃, methadone–d₃, methamphetamine–d₈, norcocaine–d₃, and norketamine–d₄. All the internal standards were purchased from Cerilliant (Texas, USA). High–performance liquid chromatography (HPLC) grade solvents (methanol (MeOH), acetonitrile, ethyl acetate, water) were purchased from Fisher Scientific (CA, USA). Formic acid (99.9%) and ammonia hydroxide (NH₄OH; 28–30%) were obtained from Wako (Osaka, Japan) and Junsei (Tokyo, Japan). Oasis MCX (200 mg, 6 cm³) and Oasis HLB (200 mg, 6 cm³) cartridges were purchased from Waters (MA, USA). Multi–layer cartridge was prepared by mixing three mixtures of Isolute ENV+ (150 mg; Biotage, UK), Strata X–AW, and Strata X–CW (each 100 mg; Phenomenex, USA) as the bottom layer and Oasis HLB (200 mg, Waters, USA) as the top layer.

The influent wastewater samples were collected from six WWTPs at a rate of 100 mL/h over a 24–h period using a composite sampler (WS700; Global Water, CA, USA) from April 3 to 4, 2023. The collected samples were stored in brown glass bottles, transported to the laboratory in an ice box, and kept in a refrigerator at temperatures below −20°C.

Before extraction, the wastewater samples were filtered with a glass microfiber filter (GF/F) (47 mm; Whatman, Maidstone, UK) and fortified with 19 internal standards (80 ng/L). To optimize the pretreatment method, extraction methods using various cartridges and elution solvents were tested, drawing on the previous studies. The cartridge test was performed using MCX, HLB, and multi–layer cartridges, following the procedures described in earlier research [9,8,14]. Due to the diverse properties of the target compounds, the elution solvent test was conducted using three different solvents: MeOH (for universal application), 5% NH₄OH in MeOH (for acidic compounds), and 2% formic acid in MeOH (for basic compounds). The tested experimental methods are summarized in Fig. 1.

The target compounds were analyzed using an Agilent 1200 HPLC system coupled with a 6460 triple quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA, USA). Chromatographic separation was achieved by using a ZORBAX Eclipse XDB–C18 column (4.6 mm i.d × 150 mm, 3.5 μm particle size; Agilent Technologies) with a guard column (ZORBAX Eclipse XDB–C18; 4.6 mm i.d × 5 mm, 1.8 μm particle size; Agilent Technologies). The mobile phases used were 0.1% formic acid in acetonitrile:water (95:5, v/v) (A) and 0.1% formic acid in acetonitrile:water (1:99, v/v) (B). The system settings included a flow rate of 0.5 mL/min, a column oven temperature of 35°C, and an injection volume of 10 μL. For effective separation, gradient elution was performed, and the gradient of (B) was as follows: 0–5 min, 95%, 5–10 min, 70%, 10–20 min, 30%, 20–23 min, 95%, 23–30 min, 95%. The pressure was stabilized by giving 3 minutes post time until the next sample analysis. The total run time was 33 minutes. Electrospray ionization (ESI) was conducted in the positive mode, with the ESI parameters set as follows: capillary voltage at 4,000 V; gas flow at 8 L/min; sheath gas flow at 12 L/min; and nebulizer pressure at 40 psi.

The developed method was validated with a focusing on specificity, linearity, accuracy, precision, method detection limit (MDL), limit of quantification (LOQ), and robustness. Quantification of all target compounds was conducted using the isotope dilution method. Specificity was evaluated by presenting individual chromatograms for the analytes in the wastewater. Linearity was assessed through a correlation coefficient (r²) using at least five concentrations of calibration standards. MDL and LOQ were determined based on signal–to–noise ratios of 3:1 and 10:1, respectively. Intra–day precision was determined through repeated experiments performed on the same day, while inter–day precision was assessed based on the results of the intra–day precision tests repeated over 2 days. Robustness was evaluated by testing the stability of the final solution. This was done by examining the rate of recovery change by the hour (Initial, 12 h, 24 h, and 36 h) and by the day (Initial, 7 days, and 30 days). The vials for the hourly test were left at room temperature, while the vials for the daily test were stored in a dark room at −20°C.

The instrument conditions for the analysis were formulated based on parameters from previous studies [8,9]. A standard 1 ng/mL solution was prepared for each compound to set the analytical conditions. The MS/MS instrument was operated in the multiple reaction monitoring (MRM) mode, and precursor ions were identified through MS 1 scan mode. Product ions were selected for their abundance after optimizing the collision energy and were used as quantitative and qualitative ions. The MRM conditions were cross–verified with mzCloud, a mass spectrum library provided by Thermo Fisher Scientific. For quantification purposes, isotope–labeled internal standards were used for each substance. If corresponding internal standards were unavailable, substances with retention times close to the target compounds were used as matches. The fully optimized MRM condition can be found in the supplementary materials (Tables S1 and S2).

Given the diversity in polarity and acid–base properties among various compound classes, comprehensive sample extraction necessitates extensive testing of SPE cartridges and elution solvents [16]. As detailed in section 2.3, tests were carried out three times with 20 ng/L spiked wastewater samples to determine optimal conditions for SPE. The target average recovery rates were set within the 70–120% range.

Method validation was conducted using the optimized pretreatment method described earlier. Details of method validation results are shown in Table 1–3. Specificity was confirmed through the separation of individual compound peaks in the wastewater sample (Fig. S1–S6 in the supplementary materials). Representative chromatograms ensured the separation of each compound, and no interference peaks were observed. Linearity was assessed by the correlation coefficients (r²) of calibration curves (levels: 0.5, 1, 2, 5, 10, 20, 50, and 100 ng/mL; weight: 1/x). The r² values for all 37 compounds ranged from 0.9969–0.9999, showing satisfactory linearity (>0.99) for all compounds. Accuracy and precision were tested using spiked samples prepared at three concentrations (low: 20 ng/L; mid: 80 ng/L; high: 200 ng/L), repeated three times. The average recovery rates were 78.1–112.2%, 72.2–114.1%, and 82.4–114.7% at low, mid, and high concentrations, respectively. These results satisfied the target accuracy range (70–120%) and show improvement over previous studies accuracy (49–127%) of NPS [18]. Intra–day precision ranged from 0.45–5.05%, 1.39–10.80%, and 0.74–11.63% at low, mid, and high concentrations, while Inter–day precision ranged from 0.96–16.46%, 2.04–17.44%, and 1.52–17.99%, at low, mid, and high concentrations, respectively. All 37 compounds obtained satisfactory precision in the three concentrations (<20%). The MDLs and LOQs for 37 compounds were 0.01–2.91 ng/L and 0.03–9.71 ng/L, respectively. Valerylfentanyl showed the lowest MDL and LOQ, while cathinone had the highest values. Similar LODs and LOQs of NPS were 0.02–4.50 and 0.06–7.80 ng/L, respectively found in other NPS studies [18]. Robustness was evaluated using three samples with low spiking concentration (20 ng/L). The average recovery rate was 75.1–108.6%, 76.1–112.3%, 74.1–111.5%, and 72.4–107.8% at the 0 h, 12 h, 24 h, and 36 h, respectively, showing satisfactory results of all compounds for 36 h at room temperature. The average recovery rates were 76.8–111.4%, 73.5–118.0%, and 52.4–160.8% at the 0, 7, and 30 days, respectively. All 37 compounds satisfied the recovery rates until 7 days when stored at –20°C: however, 25E–NBOMe, mescaline, flephedrone, methedrone, methylone, and furanylfentanyl did not meet target range (70–120%) at 30 days. The robustness test indicated that the final solution was stable for 36 hours at room temperature and up to 7 days when stored in the dark at −20°C.

The analysis of NPS in wastewater collected from six WWTPs was conducted using the analytical method developed in this study. Among the 37 target chemicals, five substances (amphetamine, MDMA, methamphetamine, cathinone, and methcathinone) were detected in the wastewater samples (Table 4). The detected concentrations of amphetamine ranged from 1.90 to 5.17 ng/L with a detection frequency of 80%, MDMA was detected only at one sample (3.23 ng/L). In comparison, the concentrations of methamphetamine ranged from 10.92 to 44.60 ng/L with a detection frequency of 100%. Compared to previous studies in Italy and Spain, the levels of amphetamine and MDMA were relatively low in this study, whereas methamphetamine showed similar or slightly high concentrations [19,20]. The concentrations of amphetamine and methamphetamine were similar to those of a previous study in Korea, but MDMA was not detected in that study [8]. Cathinone was detected in four sites, with concentrations ranging from below the <LOQ to 13.33 ng/L. Methcathinone levels ranged from 5.89 to 13.99 ng/L in five sites. The concentrations of methcathinone were similar to or lower than those reported in the USA [21]. Additionally, methcathinone was detected in Korea WWTPs between 2020 and 2021 [5].