2.5.1. Biological samples.

Stool and urine samples were self-collected by participants following instructions and materials provided. For stool, a big garbage bag was used to completely cover the toilet, and a big piece of aluminum foil was placed on top of it in order to intercept the stool and avoid contact with the plastic garbage bag. A disposable wooden spoon was used to collect a stool sample (≈50 grams), that was placed in another aluminum foil, used to completely wrap it, that in turn was placed in a 120 mL container for storage and transportation. Urine samples were collected in 150 mL polyethylene terephthalate (PET) containers (bisphenol A free) with screw aluminum cap in parallel to the stool sample collection phase, including a sample late at night and one first in the morning, leading to a variable number of urine samples. Samples were labeled with the volunteer’s code, time and date, and stored in hermetic seal bags in the refrigerator until shipment to the research institute on the same day. Participants telephoned the study personnel after each stool sample collection in order to arrange a courier to collect and transport samples to the research institute, where samples were promptly stored at -20ºC. Samples were centrally shipped to the laboratory for analysis after fieldwork completion. Prior than that, urine samples where thawed to mix night and morning samples into a unique homogenized sample.

2.5.2. Water samples.

During tap water consumption phase, a daily water sample was collected by study personnel following previously validated methods [20]. To facilitate sampling, which involved filtering 26–40 liters over approximately 30 minutes, water was collected from the nearest public access point in the distribution system prior to the household. Comparability of MNP concentrations relative to the tap water at participants residence was confirmed by two sampling campaigns where water samples were simultaneously collected in both locations (Table A in S1 File). Prior to tap water sampling, all material was cleaned to minimize contamination. We used a mix of analysis grade ethanol from Merck (Darmstadt, Germany) and MilliQ water supplied by MilliQ system from Millipore (Bedford, MA, USA) in a proportion of 70:30 [20]. Water was filtered on site using a PTFE holder (Savillex, Eden Prairie, MN, USA) containing a 13 mm glass fiber filter with 1 μm pore size (Millipore, Burlington, MA, USA). The filtration system was designed to prevent air exposure, with water flowing directly from the cabinet to the filter, thereby minimizing the risk of external contamination. Additionally, no loss of microplastics occurred during the preconcentration stage. Immediately after filtration, filters were placed in glass petri dishes for further analysis. To prevent no microplastic contamination from the sampling apparatus, several 1L blanks of HPLC-grade water were processed using a peristaltic pump, particularly to verify the absence of contamination. A total of 24 drinking water samples from the water distribution network were collected from each participant, according to the predetermined scheme outlined in Table 1 (7 samples for participant 1 and 3, and 10 samples for participant 2). Two bottled water samples of the commercial brand provided to participants were analyzed for microplastic concentrations using the same procedure. This limited bottled water sampling effort was deemed sufficient as the quality of natural mineral water is generally stable, and prior evidence indicated that MNP contamination primarily originates from packaging rather than raw water sources [11].

2.6.1. Py-GC/MS analytical methodology.

Water samples were analyzed using a previously developed analytical methodology [21]. Briefly, samples were analyzed utilizing pyrolysis coupled to gas chromatography/mass spectrometry (Py-GC/MS) using a 6890N gas chromatograph connected to a 5977B MSD single quadrupole mass spectrometer (Agilent, CA, USA), equipped with a multi-shot pyrolyzer EGA/PY-3030D (Frontier Laboratories, Fukushima, Japan). Pyrolysis was conducted at a temperature of 600ºC in single shot mode. The interface temperature was set at 300ºC. The injection was performed with a split ratio of 10:1 at a temperature of 300ºC. Chromatographic separation was performed using a DB-5MS fused-silica capillary column (5% phenyl- 95% methylpolysiloxane) measuring 30 m x 0.25 mm I.D. x 0.25 µm film thickness (Agilent, CA, USA). The initial oven temperature was set at 40ºC (1 minute), then ramped to 320ºC at a rate of 20ºC/min and held for 5 minutes. The total runtime was 20 minutes. Helium (99.9999%) was used as the carrier gas, at a constant flow rate of 1 mL/min. The mass spectrometer operated in electron ionization mode at 70 eV. The temperatures of the MS Source, MS Quad, and transfer line were set at 230ºC, 150ºC, and 280ºC, respectively. The acquisition was conducted in full scan mode (50–550 m/z). Specific pyrolysis ions for each polymer type were used to identify and quantify MPs. Additionally, the obtained results were verified using the F-Search software (Frontier Laboratories).

Quantification was performed through internal standard calibration of microplastic standards. In addition, 4 procedural blanks were performed along the sample batch and successfully analyzed. Only the samples with an area greater than three times the blank were considered as positives. Moreover, to minimize interferences, all the used materials (glass fiber filters, pyrolysis cups, and tweezers) were burned at 450ºC for 3 hours.

2.6.2. Validation parameters for the MNPs from water samples.

Quality controls and recoveries applied in this study are those published in the previous study by Dalmau-Soler, 2024 [21]. Methods limit of detection (MLOD) and quantification (MLOQ) were calculated by using the response of each MNPs in spiked 1L of HPLC-grade water (between 7 and 30 µg) based on a signal-to-noise (S/N) ratio of 3 and 10 respectively.

2.7.1. Extraction procedures.

Plastic particles were extracted according to the Fenton digestion [22, 23]. Very briefly, 100 mL of 30% H₂O₂ and 40 mL of an iron catalyst solution were mixed with 0.5 g of lyophilized sample in 400 mL glass vials. The mixture was incubated at 40°C for 5 hours for digestion. Subsequently, the sample was centrifuged at 3,500 rpm for 15 minutes. The centrifuged sample was filtered using 0.7 µm glass fiber filters. 50 mL of 65% HNO₃ was added to the filters for digestion at 50°C for 30 minutes, followed by 10 minutes at 70°C. After digestion, the mixture was eluted from the filters, and the filters were washed with 50 mL of HPLC-grade water twice. A final wash with 50 mL of ethanol was applied. The filters were dried overnight in an oven at 40°C, then extracted with 10 mL of toluene in an ultrasonic bath for 15 minutes. This extraction step was repeated twice, and the resulting toluene extracts were combined and evaporated under a nitrogen stream until approximately 1 mL remained. The concentrated extract was transferred to a liquid chromatography (LC) vial and evaporated to a final volume of 1 mL.

0.5 mL of urine samples were extracted with 0.5 mL of toluene in an ultrasonic bath for 10 minutes. The supernatant was collected and transferred to an LC vial. This extraction process was repeated twice, and the resulting supernatants were combined and evaporated under a nitrogen stream until the sample volume was reduced to 0.5 mL. For the analysis of plastic additives, the same procedure was performed, but methanol was used as the extraction solvent instead of toluene.

2.7.2. Determination of MNPs and plastic additives.

MNPs quantitative analyses in stool and urine samples was carried out by size exclusion chromatography and with high-resolution mass spectrometry (LC(SEC)-HRMS) using the previously validated method for complex samples using a QExactive-hybrid quadrupole−Orbitrap mass analyzer [10, 24, 25]. The separation was done in an Acquity HPLC system from Waters Corp. (Milford, MA, United States) equipped with an advanced polymer column (Acquity APC XT45 1.7 µm particle size), using toluene as an isocratic mobile phase and at a flow rate of 0.5 mL/min for 10 µL of sample injected. The chromatographic system was coupled to an HRMS Q-Exactive hybrid quadrupole-Orbitrap mass spectrometer from Thermo Fisher Scientific (San José, CA, United States), equipped with an atmospheric pressure photo ionization (APPI) source which was operating in negative and positive ionization conditions. Acquisition data was in full scan mode from 500 to 3,000 m/z and at a resolution of 30,000 full widths at half maximum (FWHM). Data acquisition and processing was carried out by Xcalibur v4.2 software (Thermo Fisher Scientific). The suspect screening approach for MNP polymers identification was according to the procedure previously developed and applied by the IDAEA-CSIC research center [10, 11]. In summary, the identification (level 2) of polymers was by means of confirmation by Kendrick Mass Defect (KMD) analysis followed by confirmation (level 1). For the final confirmation and quantification of the polymers an external calibration curve of polymer standards was performed based on equivalent concentrations.

Plastic additives were analyzed in the same instrumentation as described above but the chromatographic separation was achieved by means of Purospher STAR RP-18 endcapped column (5 µm, 2 x 125 mm) from Merck. The mobile phase consisted of acetonitrile 100% (A) and water (B) in negative conditions and water acidified at 0.1% (v/v) with formic acid (B) for positive ionization mode as described in previous work [10, 11]. Data acquisition in HRMS equipped with an electrospray ionization source was in full scan-data dependent MS² (FS-ddMS²) mode (from 100 to 1,500 m/z) for the 5 most intense ions. Data processing was performed using the Xcalibur Qual Browser software (Thermo Fisher Scientific, San Jose, CA, USA), and Compound Discoverer software version 3.3 SP1 (Thermo Fisher Scientific, San Jose, CA, USA), respectively. For the suspect screening approach, the information contained in an internal database and Chemspider for structural information and MzCloud, as mass spectra database was used. The tentative identification criteria were based on those described by previous studies [26, 27]. Other filters applied were mass tolerance of 5 ppm, the retention time (tolerance ± 2.5% min), isotopic fit (>90%), fragmentation and the previously mentioned parameters in addition to the mass peak resolution and response. The suspect screening was performed as follows; the FS-ddMS² spectra were processed with Compound Discoverer to obtain a list of candidates between the ones contained in the inclusion list. Plastic additives tentatively identified at level 2 were the suspected structures matching with the experimental fragmentation spectra comparing them with the internal DBs or theoretical patterns. Finally, confirmation (identification level 1) was achieved by standard comparison and then quantification was carried out by external standard curves.

2.7.3. Validation parameters for the MNP extraction from stool samples.

The extraction method is widely applied to eliminate organic matter in the study of plastic particles from faces, and the performance of the analytical method was already validated for different complex matrices. Here, we have tested the approach for the quantitative analysis of NPs polymers in faces. Due to the lack of reference materials for faces free of microplastics, a wastewater sludge was selected for validation purposes and it is an accepted simulant material in terms of composition and complexity [28, 29]. The ILOD was estimated from the injection of the lowest calibration curve point (20 pg). The sludge was previously extracted to prepare the plastic-free matrix. Two grams of dry sludge were extracted by ultrasonic-assisted extraction with 10 ml of toluene for 10 min, the supernatant was collected, and the process was repeated four more times (quintuplicate samples). The MLOD, MLOQ, recoveries, and matrix effects for faces were detected by spiking experiments in sludge. The residual sludge was dried overnight at 40 °C. After this time, the quintuplicates were homogenized, and 0.5 g of the matrix was spiked with a mixture of selected polymers PE, PBD, PI, PP and PS in a range between 5 and 50 µg/kg each one at three spiking concentrations (5, 10 and 50 µg/kg) from a standard prepared in toluene from 1,200 Da pure polymer, and β-caprolactam at the same concentration for the measure of polyamide (PA) polymer type. In parallel, two samples of 0.5 g without spike were used as control blanks to monitor any remaining polymer the sludge might have. Then, the samples were extracted according to the protocol described in section 2.6.1. The main quality parameters for pure polymer standards at 1,200 Da and instrumental level showed good limits of quantification, from 0.05 µg/kg fg for polypropylene (PP) to 0.25 fg for PA. The method limits of quantification, calculated experimentally as 9 times the S/N ratio from the spiked samples, ranged from 0.10 (for polystyrene or PS) to 0.77 (PBD) µg/Kg. Furthermore, the recovery rates calculated by the external calibration curve in toluene solvent were between 40% (PE) and 65% (PBD), with an intraday precision below 30% in all the cases.

2.7.4. Validation parameters for the MNP extraction from urine samples.

0.5 mL of simulated urine composed by ultra-pure water with 2% of NaCl and 3% of urea was spiked with a mixture of selected polymers PE, PBD, PI, PP and PS at 10 µg/L each one (in matrix) from a standard prepared in toluene from c.a. 1,200 Da pure polymer, and β-caprolactam at the same concentration for the measure of PA. This was done also in quintuplicate, and the samples were extracted according to the protocol described in the section 2.5.1. The main quality parameters showed good limits of quantification ranging from 0.05 µg/L (PS) to 0.5 µg/L (PBD). The recovery rates were calculated by external calibration curve, also in toluene, and ranged from 60% (PE) to 70% (PS) with an intraday precision below 16%.

2.7.5. Blanks analysis of stool and urine samples.

Several measures were implemented to avoid background contamination and all sample pre-treatment and sample manipulation were performed in a laminar airflow bench in a plastic-free laboratory using cotton lab coats. All the equipment used for sampling manipulation was glass whenever possible. Moreover, 2 procedural blanks consisting of ultra-pure HPLC-water were carried out with each extraction batch to assess the eventual presence of MNPs that may come from external sources, and caps from LC vials were changed for aluminum foil to prevent contamination during the LC-HRMS analyses. Results showed that no MNP were detected during the procedural blanks analysis. In addition, the possible cross contamination coming from the instrumental analysis was monitored by injecting solvent blanks of toluene (polymers) and water-methanol 9:1 (plastic additives) between each sample injection.