Deliverable D1.4 - GC- and LC-HRMS methods and data treatments workflows for suspect screening and retrospective identification of relevant CECs in the water cycle and for identifying transformation products of PFAS during remediation process

Project ID N°: 101036449 Call: H2020-LC-GD-2020-3 Topic: LC-GD-8-1-2020 - Innovative, systemic zero-pollution solutions to protect health, environment, and natural resources from persistent and mobile chemicals Preventing Recalcitrant Organic Mobile Industrial chemicalS for Circular Economy in the soil-sediment-water System Start date of the project: 1st November 2021 Duration: 36 months Main authors: Christian Dietrich (BAFG), Alicia Hartmann (BAFG), Alessandro Frugis (ACEA), Marco Mancini (ACEA), Valentina Gioia (ACEA), Marco Lazzazzara (ACEA), Dana Pierina Orlando-Veliz (IDAEA-CSIC), Manuel García Vara (IDAEA-CSIC), Rocío Inés Bonansea (IDAEACSIC), Marta Llorca (IDAEA-CSIC), Marinella Farré (IDAEA-CSIC), Miren López de Alda (IDAEACSIC), Frederik Zietschmann (BWB), Tobias Hensel (BWB) Lead Beneficiary: BAFG Type of delivery: R Dissemination Level: PU Filename and version: PROMISCES_D1-4_SuspectScreening (version 1) Website: https://promisces.eu/ Due date: 31 October 2024 D1.4 - GCand LC-HRMS methods and data treatments workflows for suspect screening and retrospective identification of relevant CECs in the water cycle and for identifying transformation products of PFAS during remediation process D1.4 – Suspect Screening Workflows 2 © European Union, 2024 No third-party textual or artistic material included on the publication without the copyright holder’s prior consent to further dissemination by other third parties. Reproduction is authorized provided the source is acknowledged. Disclaimer The information and views set out in this report are those of the author(s) and do not necessarily reflect the official opinion of the European Union. Neither the European Union institutions and bodies nor any person acting on their behalf may be held responsible for the use which may be made of the information contained therein. D1.4 – Suspect Screening Workflows 3 Document History This document has been through the following revisions: Authorisation Distribution This document has been distributed to: Version date Author/Reviewer Description 0.1 14/10/2024 C. Dietrich First draft for review 0.2 07/11/2024 C. Dietrich Second draft for review 0.3 14/11/2024 C. Dietrich Implementation of review comments 0.4 28/11/2024 F. Sermondadaz Version after QC 1.0 28/11/2024 C. Dietrich Final Version for distribution Authorisation Name Status Date Review L. Amalric Reviewer BRGM head of laboratory 08/11/2024 Validation A. Togola WP1 leader 25/11/2024 Quality Control F. Sermondadaz Administrative and financial manager 28/11/2024 Approval J. Lions Project coordinator 29/11/2024 Name Title Version issued Date of issue All WP1 partners WP1 Partners Version 1 29/11/2024 D1.4 – Suspect Screening Workflows 4 Executive Summary A large number of anthropogenic substances including PFAS and iPM(T)s are present in the water cycle. For most of these compounds, information about occurrence, sources and fate are scarce as targeted methods alone cannot cope with the large variety of these substances. This document reports on the results regarding development of suspect screening workflows for PFAS and other iPM(T)s realized within the framework of the task dealing with “Suspect screening and database-assisted analysis”. Five different workflows have been developed using different analytical setups for analyses of different matrices ranging from groundand surface water over wastewater and landfill leachates to lettuce and sediment and. Two workflows are based on liquid chromatography coupled to highresolution mass spectrometry and are specifically designed for the screening of PFAS compounds in landfill leachates and their transformation products in groundand surface water, wastewater treatment plant effluents, lettuce and sediment. Two other LC-HRMS-based workflows are dedicated to the identification of iPM(T)s in surfaceand wastewater. These workflows are complemented by a GC-MS-based method enabling a screening for more unpolar compounds not amenable to most LCMS methods and is dedicated to industrial wastewater. Subsequent data treatment is performed using commercial as well as open-source software. Matching of the spectral information of the analytical data was performed with both commercial and public compound databases (massbank.eu and mzcloud.org). An overview about the main aspects of the five workflows is given in Table 1. The developed workflows have been applied to various samples from the PROMISCES case studies CS#1, CS#2, CS#3, CS#4 and CS#7. D1.4 – Suspect Screening Workflows 5 Table 1: Comparative summary of the five methods. ACEA CSIC 1 BAFG CSIC 2 BWB Chromatography HPLC HPLC HPLC HPLC GC Analyser QToF QExactive QToF QToF GC-MS Analytes PFAS PFAS TPs iPM(T)s iPM(T)s iPM(T)s Matrix landfill leachates Surface water, groundwater, WWTP effluent, lettuce, sediments Surface water, WWTP effluent Waste water WWTP effluent, Industrial wastewater Sample preparation Dilution and addition of internal standards Centrifugation and addition of mix internal standards Filtration Addition of internal standards, lyophilization, and sample extract reconstitution Addition of internal standards, SPEEnrichment Software SCIEX OS, SCIEX LibraryView Thermo Scientific Compound Discoverer 3.3 with public and homemade libraries Homemade R scripts Bruker MetaboScape® 2022b Agilent Masshunter, Excel-script Database SCIEX Fluorochemical HR-MS/MS library 2.0 MassBank (massbank.eu) and NIST mass spectral libraries In-House spectral database (also available on massbank.eu)) mzCloud (mzcloud.org) NIST Mass Spectral Library, inhouse database D1.4 – Suspect Screening Workflows 6 Table of contents 1 Introduction ................................................................................................................................. 10 2 Analytical methods and data treatment ...................................................................................... 11 2.1 Large suspect screening of PFAS in landfill leachates (ACEA) ............................................... 11 2.1.1 General ........................................................................................................................... 11 2.1.2 Chemicals and reagents ................................................................................................. 11 2.1.3 Sample preparation and pre-treatment ........................................................................ 12 2.1.4 Chromatography ............................................................................................................ 12 2.1.5 Mass spectrometry ........................................................................................................ 13 2.1.6 Data treatment .............................................................................................................. 15 2.1.7 Performance .................................................................................................................. 20 2.1.8 Application to real samples ........................................................................................... 20 2.2 Suspect screening of PFAS transformation products (CSIC 1) .............................................. 21 2.2.1 General ........................................................................................................................... 21 2.2.2 Chemicals and reagents ................................................................................................. 21 2.2.3 Sample preparation and pre-treatment ........................................................................ 22 2.2.4 Chromatography ............................................................................................................ 22 2.2.5 Mass spectrometry ........................................................................................................ 22 2.2.6 Data treatment .............................................................................................................. 22 2.2.7 Application to real samples ........................................................................................... 23 2.3 LC-MS Suspect screening of iPM(T)s in surface - and wastewater (BAFG) ........................... 24 2.3.1 General ........................................................................................................................... 24 2.3.2 Chemicals and reagents ................................................................................................. 24 2.3.3 Sample preparation and pre-treatment ........................................................................ 24 2.3.4 Chromatography ............................................................................................................ 24 2.3.5 Mass spectrometry ........................................................................................................ 25 2.3.6 Data treatment .............................................................................................................. 25 2.3.7 Performance .................................................................................................................. 26 2.3.8 Application to real samples ........................................................................................... 27 2.4 LC-MS Suspect screening of iPM(T)s in wastewater (CSIC 2) ................................................ 28 2.4.1 General ........................................................................................................................... 28 2.4.2 Chemicals and reagents ................................................................................................. 28 2.4.3 Sample preparation and pre-treatment ........................................................................ 28 2.4.4 Chromatography ............................................................................................................ 28 2.4.5 Mass spectrometry ........................................................................................................ 29 2.4.6 Data treatment .............................................................................................................. 29 2.4.7 Performance .................................................................................................................. 29 2.4.8 Application to real samples ........................................................................................... 30 2.5 GC-MS Suspect screening of iPM(T)s in surface water (BWB) .............................................. 31 2.5.1 General ........................................................................................................................... 31 2.5.2 Chemicals and reagents ................................................................................................. 31 2.5.3 Sample preparation and pre-treatment ........................................................................ 31 2.5.4 Chromatography ............................................................................................................ 32 2.5.5 Mass spectrometry ........................................................................................................ 32 2.5.6 Data treatment .............................................................................................................. 33 D1.4 – Suspect Screening Workflows 7 2.5.7 Performance .................................................................................................................. 34 2.5.8 Application to real samples ........................................................................................... 34 3 Conclusions .................................................................................................................................. 36 4 References ................................................................................................................................... 37 D1.4 – Suspect Screening Workflows 8 List of abbreviations bbCID Broadband Collision Induced Dissociation CE Collision Energy CS Case Study Da Dalton ESI Electrospray Ionisation GC Gas Chromatography HPLC High-Performance Liquid Chromatography HRMS High Resolution Mass Spectrometry IDA information dependent acquisition iPM(T)s Industrial, Persistent, Mobile and potentially Toxic substances IS Internal Standard LC Liquid Chromatography LOD Limit of Detection LOQ Limit of Quantification MeOH Methanol MS Mass Spectrometer MS/MS Triple Quadrupole Mass Spectrometer m/z mass-to-charge ratio NTS Non-Target Screening QToF Quadrupole – Time-of-Flight PFAS Perand polyfluoroalkyl substances PTFE Polytetrafluoroethylene ppm parts per million RPM Rounds Per Minute RT Retention Time S/N Signal-to-Noise Ratio SPE Solid Phase Extraction TIC Total Ion Chromatogram ToF Time-of-Flight UPLC Ultra-high Performance Liquid Chromatography WP Work Package WWTP Wastewater Treatment Plant XIC eXtracted Ion Chromatogram D1.4 – Suspect Screening Workflows 9 List of Figures Figure 1: Illustration of the workflow used for suspect screening of PFASs in the ACEA LC-HRMS method ................................................................................................................................................. 16 Figure 2: Screenshot of parameter setup in ACEA workflow .............................................................. 17 Figure 3: Example of suspect presence of PFHxS in an analyzed sample. ........................................... 18 Figure 4: Extraction of PFHxS-XIC from the TIC chromatogram .......................................................... 19 Figure 5: XIC peak peaking and MS spectrum extraction .................................................................... 19 Figure 6: MS spectrum of PFHxS with its exact mass confirming ion. ................................................. 20 Figure 7: Example of workflow used for suspect screening of PFAS in IDAEA-CSIC. ........................... 23 Figure 8: Data treatment workflow for database-assisted suspect screening of iPM(T)s using BAFG method ................................................................................................................................................. 25 Figure 9: Chromatograms of the four example samples. .................................................................... 35 List of Tables Table 1: Comparative summary of the five methods. ........................................................................... 5 Table 2: List of Internal Standards used in the ACEA workflow. ......................................................... 12 Table 3: Chromatographic parameters used in the ACEA LC-HRMS method. ..................................... 13 Table 5: MS-Instrument parameters for suspect screening of PFAS in the ACEA LC-HRMS method. 14 Table 6: Parameters for peak picking within the ACEA suspect screening workflow ......................... 15 Table 7: List of PFASs used in CSIC 1 suspect screening method ........................................................ 21 Table 8: Chromatographic gradient used for suspect screening of iPM(T)s in the BAFG LC-HRMS method. ................................................................................................................................................ 24 Table 9: Parameters for peak picking within the BAFG suspect screening workflow ......................... 26 Table 10: Chromatographic gradient used for suspect screening of iPM(T)s in the CSIC LC-HRMS method. ................................................................................................................................................ 29 Table 11: Chemicals and materials used for suspect screening of iPM(T)s in the BWB GC-MS method. .............................................................................................................................................................. 31 Table 12: Chromatographic parameters used for suspect screening of iPM(T)s in the BWB GC-MS method. ................................................................................................................................................ 32 Table 13: Temperature gradient used for suspect screening of iPM(T)s in the BWB GC-MS method. .............................................................................................................................................................. 32 Table 14: MS-Instrument parameters for suspect screening of iPM(T)s in the BWB GC-MS method. .............................................................................................................................................................. 32 Table 15: Categories of confidence for suspect screening of iPM(T)s in the BWB GC-MS method. ... 33 Table 16: Example data of four wastewater samples analyzed by the BWBs GC-MS suspect screening method. ................................................................................................................................................ 34 D1.4 – Suspect Screening Workflows 16 If the integrated peaks meet the criteria of “acceptable differences” then it will go to Explorer mode processing for final confirmation. In Explore mode, in the case of a “suspect” analysis, the chromatographic peak (XIC) corresponding to the exact mass of the compound previously found in Analytics mode is extracted from the TIC. At this point we will go to search, in the mass spectrum that generated the peak, for the precursor ion and any other characteristic fragments to have a confirmation of the analyte. The entire workflow is depicted in the Figure 1. Figure 1: Illustration of the workflow used for suspect screening of PFASs in the ACEA LC-HRMS method In a real case, the acquired data will undergo our suspect analysis. In the so called “Analytics mode” we’ve settled down some flagging rules useful to know if the acquired spectra meets the library data. The next screenshot replicates the conditions in Table 6. D1.4 – Suspect Screening Workflows 17 In the following example (see Figure 3) there is a partial match for the presence of PFHxS in our real sample (a leachate sample). Specifically, what is found is a positive match for mass error and library confidence with a marginal difference for the isotope ratio confidence. As we can see, in the reported spectra (namely XIC, MS and MSMS) we can visualize the chromatographic peak, the main ion and (in the right part) the fragmented ions matched with the library spectrum (reported as grey negative comparative spectrum). Figure 2: Screenshot of parameter setup in ACEA workflow D1.4 – Suspect Screening Workflows 18 Once we’ve noticed these matches, the samples which are suspected to contain compounds of interest will be investigated in the “Explorer mode” in order to obtain a confirmation of our hypothesis. For this purpose, we use the data collected with IDA experiment. Indeed, in IDA mode (information dependent acquisition), the spectrometer can skim between the possible candidates which will undergo the fragmentation by following the criteria listed in 2.1.5.2. Considering the setting in criteria b), we can support that in IDA experiment, skimming between the possible candidates which will undergo the fragmentation, we will record a cleaner spectrum. In the reported example (Figure 4) we extract from the TIC chromatogram the XIC of the hypothesized molecule (PFHxS). By zooming and selecting a portion of the peak in the XIC, we extract the MS spectrum of the suspected compound (Figure 5). As we have obtained the XIC of the suspected compound using the brute formula of the molecule, we will look for its calculated mass to confirm his presence (see the zoomed MS spectrum in Figure 6 and the circled confirming ion). Figure 3: Example of suspect presence of PFHxS in an analyzed sample. D1.4 – Suspect Screening Workflows 19 Figure 4: Extraction of PFHxS-XIC from the TIC chromatogram Figure 5: XIC peak peaking and MS spectrum extraction D1.4 – Suspect Screening Workflows 20 Figure 6: MS spectrum of PFHxS with its exact mass confirming ion. As seen above the “Analytics mode” is a real and proper suspect analysis (rather than a NTS) where the comparison is done with a fluorinated library. Nevertheless if the focus must be on other compounds classes, there is still the possibility to compare our collected data and spectra with other libraries in order to satisfy the subject of investigation. Instead, at the state of the art, the same cannot be said for the “Explorer mode”. Indeed in this case, even if we manually explore the spectra, we start from the assumption that the compound is present or may be present. If conversely we extract the XIC and MS spectra without any previous assumption, the research of compounds would request much more effort and time indicating that this procedure is not suitable and recommendable as routine analysis screening. 2.1.7 Performance Performance of the LC-HRMS measurement in terms of LOD (limit of detection) cannot be determined for this workflow, as the non-target method applied here works without calibration. Furthermore, it has to be noted that LODs in LC-MS measurements are highly substance-specific. Of those compounds identified by this workflow that have been additionally quantified via targeted methods (cf. Deliverable D1.2 and CS#4), concentrations less than 1 µg/L could be determined. The LODs estimates by the use of IS, were in the range 0.5-5 µg/L. 2.1.8 Application to real samples The ACEA workflow has been applied to more than 10 leachate samples from landfill leachate treatment plants (LTP) in CS#4. Out of 10 samples analysed, all PFAS previously identified by target analysis were correctly confirmed. Only PFBA was not automatically confirmed as the concentrations were close to the LOD. The range of PFAS identified is between 1-40 µg/L. No other PFAS besides those analysed as targets were identified. All internal standards were correctly identified Other applications, outside the scope of this project, were carried out on wastewater samples in routine monitoring at the inlet and outlet of sewage treatment plants. D1.4 – Suspect Screening Workflows 21 2.2 Suspect screening of PFAS transformation products (CSIC 1) 2.2.1 General The study of PFAS and related products by means of suspect screening has been based on previous group experience with other groups of compounds. The workflow allows us to tentatively identify up to level 2 of confidence PFAS & related compounds according to Schymanski scale (Schymanski 2014) and eventually level 1 for those PFAS for which CSIC has the standards. 2.2.2 Chemicals and reagents The chemicals used for the development of the workflow are shown in Table 7. Table 6: List of PFASs used in CSIC 1 suspect screening method Carboxylic acids Sulfonic acids & sulfonamides New PFASs PFBA PFBS 6:2 diPAP PFPeA PFPeS 8:2 diPAP PFHxA PFHxS ADONA PFHpA PFHpS EtFOSA PFOA PFOS EtFOSAA PFNA PFNS FOSAA PFDA PFDS MeFOSAA PFUnA PFDoS MeFOSA PFDoA FOSA HFPO-DA (Gen-X) PFTrDA Fluorotelomer sulfonic acids PFMOAA PFTeDA 4:2 FTSA PFMOPrA PFHxDA 6:2 FTSA PFMOBA PFODA 8:2 FTSA PFO2HxA 10:2 FTSA PFO3OA Other chemicals and reagents used for the analysis of PFAS include ultrapure water, methanol, ammonium acetate and ammonium hydroxide. Furthermore, the pre-concentration of waters is done by solid phase extraction with Oasis WAX 3cc cartridges. D1.4 – Suspect Screening Workflows 22 2.2.3 Sample preparation and pre-treatment Water samples are centrifuged before extraction at 2000 rpm at room temperature for 10 min. Afterwards, 200 mL of water is transferred to a PET (Polyethylene) container and spiked with a mixture of surrogate internal standards in methanol for a final concentration of 10 pg/mL in sample. Then, the sample is processed following a method adapted from another one previously developed by IDAEA-CSIC (Barbosa 2023). Briefly, solid phase extraction (SPE) cartridges are successively conditioned with 2 mL of methanol and 2 mL of ultrapure water (gravity conditions). Sample loading of 200 mL of surface water, and 100 mL of wastewater (WW) is done under vacuum conditions using PEEK capillary tubes, at a flow rate of 1 mL/min. The cartridges are then dried under vacuum for 15 min and PFAS eluted with 8 mL of methanol (0.1% NH4OH) in Polypropylene (PP) tubes and evaporated under a gentle stream of nitrogen near to dryness. The final extracts are transferred to LC-vials with 250 µL inserts, dried, and reconstituted in 100 µL of ultrapure water/methanol (90:10). A SPE blank sample is always carried out in parallel to real samples in order to monitor any crosscontamination. 2.2.4 Chromatography The chromatographic separation is achieved using an Acquity LC (Waters, Milford, MA, USA) system, equipped with a C18 analytical column Hypersil GOLD PFP LC (50x3 µm) (Thermo Fisher Scientific, San Jose, CA). The mobile phase used consists of (A) aqueous ammonium acetate 20 mM and (B) methanol ammonium acetate 20 mM. Very briefly, at the starting point, the elution gradient is 20% B and within 5 min rise to 80% B, then in 5 min increases to 90% B and is maintained for 2 min more. Finally, initial conditions are achieved within 1 min and maintained for 1 min more. Therefore, the total run time is 12 min for each injection using a flow rate of 0.2 mL/min. The optimal injection volume is 10 μL. 2.2.5 Mass spectrometry The chromatographic system is coupled to a Q-Exactive hybrid quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific) equipped with an electrospray ionization (ESI) source, working in negative ionization conditions. Data is acquired in full scan (FS) (90-1500 Da) with a resolution of 70,000 FWHM and data-dependent scan (ddS) of the most intense ions at resolution of 15,000 FWHM. The entire system is controlled by Xcalibur 4.1 software. 2.2.6 Data treatment The total ion chromatograms (TIC) obtained by FS acquisition are processed using the Xcalibur software (Thermo Fisher Scientific) for quantification purposes of the standards that are available in CSIC (level 1 of confidence). Suspect screening of the samples for tentative identification of new PFASs is carried out by processing the data by Compound Discoverer 3.3 SP2 (Thermo Fisher Scientific). A homemade list of PFASs was assembled by PFAS IDAEA-CSIC team, with data from the literature and other databases containing the exact mass of 1,280 compounds. Structural information of the listed compounds, monoisotopic mass and properties such as LogP and logD have been calculated using Chemicalize platform of D1.4 – Suspect Screening Workflows 23 ChemAxon (chemicalize.com) and included in the homemade database. Furthermore, the software has been provided with the information in different databases like the EFS HRAM Compounds database (Thermo Fisher Scientific), PFAS NIST database (data.nist.gov), ChemSpider (chemspider.com) for structural information, and MzCloud (mzcloud.org) as a mass spectra database. An example of workflow used can be seen in Figure 7. Figure 7: Example of workflow used for suspect screening of PFAS in IDAEA-CSIC. 2.2.7 Application to real samples The IDAEA-CSIC workflow has been applied to ground coming from WP3 and CS#7 and wastewaters from WP4 and CS#3 in Spain. Furthermore, the workflow will be applied to samples coming from the experiments with lettuce irrigated with treated wastewater in collaboration with other partners from IDAEA-CSIC. Up to now, no new PFAS compounds in addition to those covered by targeted methods (cf. Deliverable D1.1) have been found, underlining the relevance and completeness of the previously chosen target analytes. D1.4 – Suspect Screening Workflows 24 2.3 LC-MS Suspect screening of iPM(T)s in surface - and wastewater (BAFG) 2.3.1 General The goal of this method is to screen for a large amount of substances present in surface water or treated wastewater. Analysis is performed via LC-HRMS in a database-assisted suspect-screening approach. Compounds tentatively identified by this method can be easily included into an existing LC-MS/MS targeted analysis method (cf. Deliverable D 2.1). 2.3.2 Chemicals and reagents Eluents for chromatography: acetonitrile and formic acid (both LC-MS grade) were received from Merck (Darmstadt, Germany) and Sigma-Aldrich (Seelze, Germany), respectively. Ultrapure water was prepared with a Milli-Q water purification system (Merck Millipore). 2.3.3 Sample preparation and pre-treatment Samples from monitoring studies (case studies CS#1 and CS#2) were send to BAFG by project partners and were stored in the refrigerator at 5°C until analysis. Prior to analysis samples were filtered through 0.45 µL PTFE syringe filters. No further pre-treatment was performed. 2.3.4 Chromatography Chromatographic separation is achieved with a Zorbax Eclipse Plus C18 column (Agilent Technologies) using an Agilent 1260 infinity (Agilent Technologies). Aliquots of the sample under investigation (80 µL) are directly injected into the system without further pre-treatment. Ultrapure water with 0.1% formic acid (eluent A) and acetonitrile with 0.1% formic acid (eluent B) are used as eluent at a flow rate of 0.3 ml/min and a column oven temperature of 40°C with the gradient shown in Table 8: Table 7: Chromatographic gradient used for suspect screening of iPM(T)s in the BAFG LC-HRMS method. Time Eluent A Eluent B 0 min 98% 2% 1 min 98% 2% 2 min 80% 20% 16.5 min 0% 100% 22 min 0% 100% 22.1 min 98% 2% 25 min 98% 2% D1.4 – Suspect Screening Workflows 25 2.3.5 Mass spectrometry Mass spectrometric analysis is performed with a TripleToF 6600 hybrid quadrupole time-of-flight mass spectrometer (Q-ToF-MS/MS) (SCIEX, Darmstadt, Germany) equipped with an ESI source and operated in positive and negative ionization mode in two separate runs. Data acquisition was done by means of full scan experiments ranged from 100 to 1200 Da. Subsequently, MS2 spectra of the eight most intense peaks were recorded via information dependent acquisition (IDA) with a collision energy (CE) set with potential 40 V and a collision energy spread (CES) set with potential 15 V. Acquisition time of the full scan and the IDA experiments were 150 ms and 30 ms, respectively. 2.3.6 Data treatment For acquisition of raw data the instrument control software (Analyst TF 1.7, SCIEX) was used. Subsequently, acquired raw data files were transferred to the open data format mzXML using ProteoWizard 3.0 (proteowizard.sourceforge.io) The overall data treatment workflow is depicted in Figure 8. Figure 8: Data treatment workflow for database-assisted suspect screening of iPM(T)s using BAFG method Peak picking is performed by an algorithm written in R, which is described in detail in (Dietrich 2022). The algorithm extracts chromatograms (XIC – eXtracted Ion Chromatogram) of all previously recorded full scan data within a certain mass range (so-called « bin ») using the following parameters (Table 9). As a result, a list of features is obtained for each sample (feature = signal with defined m/z and retention time). D1.4 – Suspect Screening Workflows 32 2.5.4 Chromatography The samples are analyzed using an Agilent GC-MS, with the instrument parameters summarized in Table 12. Table 11: Chromatographic parameters used for suspect screening of iPM(T)s in the BWB GC-MS method. Parameter Gas chromatograph Agilent 7890A Injection volume 1 µL Injector parameters 65 °C – 300 °C (ramp at 720 °C/min) Analytical column HP-5 MS UI, 30m x 0.250mm x 0.25 ID Carrier gas Helium If a high concentration of trace substances is expected, the extract can be diluted with ethyl acetate. The samples are injected fully automatically. The temperature program of the column oven is shown in Table 13, and after each measurement, the oven cools to ambient temperature for approximately 5 minutes before a new sample run starts. Table 12: Temperature gradient used for suspect screening of iPM(T)s in the BWB GC-MS method. Time Temperature 0 min 40 °C 2 min 40 °C 7 min 70 °C 30 min 300 °C 35 min 300 °C To determine the retention time index (RI) a standard solution of multiple alkanes (chain length: C7C33) is analyzed with each measurement sequence along with an ethyl acetate blank. 2.5.5 Mass spectrometry Mass spectrometry is performed using an Agilent 7000 Triple-Quad equipped with a (electron impact) EI source. The run parameters of the device are summarized in Table 14. Table 13: MS-Instrument parameters for suspect screening of iPM(T)s in the BWB GC-MS method. Parameter Mass spectrometer Agilent 7000 GC-MS-Triple-Quad Source temp. 300 °C Transferline temp. 300 °C Detector temp. 280 °C Mass range 50-500 m/z Ionisation mode Electron impact (EI) Ionisation energy 70 eV D1.4 – Suspect Screening Workflows 33 2.5.6 Data treatment The obtained spectra were evaluated through library comparison to identify the chemical compounds present in the samples. A semi-quantitative analysis can be performed by comparing the peak areas with the ones from the internal standard diphenyl D10 (0.5 µg/L, ~2*106 area units), based on its fragment ion at 164 m/z. For the measurement to be considered reliable, a peak area ratio of at least 1000:1 between the internal standard diphenyl D10 and the baseline is required. This criterion ensures that the signal is distinguishable from noise and that the results are of sufficient quality. Additionally, the measured spectra were subjected to a Retention Index (RI) analysis using Agilents MassHunter Unknowns Analysis (UA) software. The individual RI is determined automatically by the UA software after subtracting the blank measurements. This step reduces background noise and enhances the accuracy of the spectral identification. The software then compares the features with the stored libraries (NIST and in-house) and outputs a list with compound suggestions and various matching factors. Adapted from Schymanski et al. (Schymanski 2014), 6 categories of confidence are defined, see Table 15. Table 14: Categories of confidence for suspect screening of iPM(T)s in the BWB GC-MS method. Category Details Peak area >10,000 Delta RI <100 NIST library match factor >80 % Inhouse library match factor >80 % Reference standard 1 Confirmed structure ✔ ✔ ✔ ✔ ✔ 2 Probable structure (very high confidence) ✔ ✔ ✔ ✔ ✖ 3 Tentative structure (high confidence) ✔ ✔ ✔ ✖ ✖ 4 Structure suggestion, low confidence ✔ ✖ ✔ ✖ ✖ 5 No Structure suggestion ✔ ✖ ✖ ✖ ✖ 6 No Structure suggestion ✖ ✔/✖ ✔/✖ ✔/✖ ✔/✖ For a feature to be included in the compound list, a minimum peak area must be reached, the deviation of the defined RI must not exceed 100 and there must be a match with at least one library. The spectral comparison uses the NIST 2020 database, which contains over 31,000 compounds, along with 1.3 million spectra. The in-house BWB-GCMS library, which includes over 1,000 compounds, is updated regularly and is based on reference standards analyzed on the GC-MS. This complements the identification process, expanding the scope of potential matches and enhancing the robustness of the analysis. D1.4 – Suspect Screening Workflows 34 2.5.7 Performance A limit of quantification cannot be determined for this method, as no quantification is performed. As a quality assurance criterion, the peaks are compared with the internal standard, as described in Section 2.4.6. “Data treatment“. Quality assurance is ensured through a control standard with known concentration. 2.5.8 Application to real samples This method was applied to more than 72 wastewater samples and, for comparison, to several industrial indirect discharger samples from CS#1. On average, more than 80 peaks can be identified in wastewater samples. Example data provided in Table 16 and Figure 9; as established in additional measurements (not expensed within PROMISCES). Table 15: Example data of four wastewater samples analyzed by the BWBs GC-MS suspect screening method. Proposed substances CAS Sample 3495 4153 4493 4856 Category Area % of IS-Area diisobutyl phthalate 84-69-5 2 2415 12852 9716 1741 dibutyl phthalate 84-74-2 1 - - 2507 3240 acetic acid, butyl ester 123-86-4 3 - - 1981 - 7,9-di-tert-butyl-1-oxaspiro(4,5)deca-6,9diene-2,8-dione 82304-66-3 2 689 590 626 669 benzoic acid, 4-ethoxy-, ethyl ester 23676-09-7 2 330 158 434 406 diethyl phthalate 84-66-2 1 313 243 201 336 acetophenone 98-86-2 1 - 47 199 137 diethyl carbitol 112-36-7 3 268 - 193 - dibutyl adipate 105-99-7 2 55 209 173 310 di(propylene glycol) methyl ether 13588-28-8 3 457 23 154 - The 10 most abundant features of four different wastewater samples after data treatment are presented in Table 16. The categories given are explained in Table 15, the values for the area are relative to the area of the IS. For reference the chromatograms of those four samples are shown in Figure 9. D1.4 – Suspect Screening Workflows 35 Figure 9: Chromatograms of the four example samples. D1.4 – Suspect Screening Workflows 36 3 Conclusions This document reports on five different approaches for a database-assisted suspect screening of CECs in various matrices. Two methods using liquid chromatography coupled to high-resolution mass spectrometry are specifically designed for the screening of PFAS compounds and their transformation products, respectively. Two other LC-HRMS-based methods are dedicated to the identification of iPM(T)s in surfaceand wastewater. The latter two methods are complemented by a GC-MS-based method enabling a screening for more unpolar compounds not amenable to most LC-MS methods. Each method has been designed to be applied in a specific case study and thus for analysis in different matrices. Sample preparation therefore ranges from simple filtration or dilution over centrifugation to lyophilization. Except for the GC-MS method, which requires a solid phase extraction and subsequent elution with a solvent suitable for GC measurements, all other mentioned sample preparation techniques aim to preserve the original CEC composition of the sample as far as possible. While signal detection is performed with a high-resolution MS for the four LC-MS-based methods, the GC-MS method works with a triple quadrupole machine. Nevertheless, in all five workflows raw data is acquired via an untargeted approach without pre-defining compounds of interest. In all cases this raw data is subsequently subjected to a separate data treatment step including comparison of the acquired spectral data to a database. For this, commercial as well as open-source software is used in combination with commercial and public compound databases. Being based on an untargeted approach, the suspect screening workflows presented here commonly suffer from reduced sensitivity compared to targeted analysis. Furthermore, quantitative information is limited due to missing calibration. To address these issues, still targeted analysis is the method of choice. The big benefit of the presented suspect screening methods is the possibility to comparatively fast perform a screening for a large number of compounds without the need to develop a new target method. Furthermore, the sequential approach of data acquisition and separate data treatment allows for retrospective screening of substances even after many years. Information about new compounds identified via suspect screening can then be used to systematically adapt targeted methods for quantification of these substances. Application of the workflows dedicated to screening of PFAS and PFAS TPs in cases studies CS#3, CS#4 and CS#7 confirmed the relevance and completeness of the previously chosen target analytes as already reported in Deliverable D1.1 “ Methods for PFAS in waters and complex matrices “. Suspect screening of iPM(T)s revealed the occurrence of more than 100 compounds in samples from CS#1, CS#2 and CS#7. After further prioritization of these findings, a first selection of substances has already been included into the compound lists of the respective targeted methods (Deliverable D1.2 “ Targeted methods for relevant iPM(T) substances in waters ”). D1.4 – Suspect Screening Workflows 37 4 References Barbosa, M. O., Ratola, N., Homem, V., Pereira, M. F. R., Silva, A. M., Ribeiro, A. R., Llorca, M. & Farré, M. (2023). Per-and poly-fluoroalkyl substances in Portuguese rivers: spatial-temporal monitoring. Molecules, 28(3), 1209. Dietrich C, Wick A, Ternes TA (2022). Open-source feature detection for non-target LC–MS analytics. Rapid Commun Mass Spectrom., 36(2):e9206. doi: 10.1002/rcm.9206. Schymanski, E. L., Jeon, J., Gulde, R., Fenner, K., Ruff, M., Singer, H. P., & Hollender, J. (2014). Identifying small molecules via high resolution mass spectrometry: communicating confidence. Environ. Sci. Technol., 48, 4, 2097-2098. doi: 10.1021/es5002105.