D1.2 Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater

D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 1 D1.2 Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater June 2024 Thomas M. Karlsson (UCPH) Jan H. Christensen (UCPH) Data driven implementation of hybrid nature-based solutions for preventing and managing diffuse pollution from urban water runoff Ref. Ares(2025)4481611 - 04/06/2025 D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 2 D1.2 Standard Operating Protocols (SOP) for suspect screening and non target screening (NTS) analysis workflows for stormwater Work Package WP1 Deliverable lead University of Copenhagen (UCPH) Author(s) Thomas M. Karlsson (UCPH) Jan H. Christensen (UCPH) Contact [email protected] Grant Agreement number 101060638 Start date of the project / Duration 1 September 2022 / 42 months Type of deliverable (R, DEM, DEC, other) R Dissemination level (PU, SEN) PU Project website www.d4runoff.eu R=Document, report; DEM=Demonstrator, pilot, prototype; DEC=website, patent fillings, videos, etc.; OTHER=other PU=Public, SEN=Sensitive, limited under the conditions of the GA Document history Version Date Detailed updates 0.1 24.06.2024 First version by UCPH, send to VCS for internal review 2.0 04.06.2025 Revision by Thomas Karlsson (UCPH) D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 3 ACKNOWLEDGEMENTS This project has received funding from the European Union’s Horizon Europe research and innovation programme under grant agreement No 101060638. COPYRIGHT STATEMENT The work described in this document has been conducted within the D4RUNOFF project. This document reflects only the D4RUNOFF Consortium views, and the European Union is not responsible for any use that may be made of the information it contains. This document and its content are the property of the D4RUNOFF Consortium. All rights relevant to this document are determined by the applicable laws. Access to this document does not grant any right or license on the document or its contents. This document or its contents are not to be used or treated in any manner inconsistent with the rights or interests of the D4RUNOFF Consortium or the Partners detriment and are not to be disclosed externally without prior written consent from the D4RUNOFF Partners. Each D4RUNOFF Partner may use this document in conformity with the D4RUNOFF Consortium Grant Agreement provisions. D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 4 Executive Summary This document presents the standard operating protocol (SOP) developed in WP1 for suspect and non-target screening (NTS) analysis of runoff samples using liquid chromatography-high resolution mass spectrometry (LC-HRMS). It is the result of further developments of the preliminary SOP presented in D1.1. The final SOP has been tested for analysis of runoff samples collected in WP1 and implemented for developing an inventory of contaminants of emerging concern (CECs) (T1.3) in urban runoff as well as to assess the fate of pollutants (T1.4). Furthermore, it will be used to analyze around 100 samples from across Europe in relation to the PARC project where UCPH is participating. The overall workflow is presented in Figure 1. It covers both sampling and sample handling, sample preparation using solid-phase extraction (SPE) to prepare samples for analysis with LC-HRMS. Finally, data processing is done to remove false positives. All of these steps will be covered in the following. Figure 1: Overview of suspect and non-target screening (NTS) workflow •1) Urban runoff samples collected over 2 hours from the beginning of rain event •2) Transported to lab within 72 hours for further processing Collection and handling of samples •1) Adjusting pH to 6.5 •2) Filtration of samples to separate liquid and solid fraction •3) Solid-phase extraction (SPE) for clean-up and preconcentration of liquid fraction •4) Spiking with internal standards and dilution to REF 10 or 50 (depending on sample) Sample preparation •1) Validation of system performance before running sequences •2) Triplicate injections of samples in positive and negative ionization mode •Blanks and quality control (QC) samples included to monitor performance Data acquisition on LC-HRMS •1) Feature detection and annotation in MS-Dial •2) Filtering of data to remove false positives •3) Manual inspection of annotations and visualization of results Data processing D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 5 Table of Contents 1 INTRODUCTION ............................................................................................................ 7 1.1 Purpose of the document ....................................................................................... 7 1.1.1 Scope of the document ...................................................................................... 7 1.1.1.1 Structure of the document .......................................................................... 7 2 SOP for suspect screening and NTS workflows for stormwater ............................... 8 2.1 Sampling ................................................................................................................ 8 2.2 Sample preparation ................................................................................................ 9 2.2.1 Filtration ............................................................................................................. 9 2.2.2 Solid-phase extraction (SPE) ............................................................................10 2.2.3 Determining optimal enrichment factor ..............................................................11 2.3 Data acquisition and processing workflow .............................................................13 2.3.1 Data acquisition ................................................................................................14 2.3.2 Feature detection and annotation in MS-Dial .....................................................15 2.3.3 Correction of drift and matrix effect ...................................................................18 2.3.4 Inspection of suggested annotations and assigning confidence scores .............19 3 Reporting of results ....................................................................................................20 4 Conclusion and next steps .........................................................................................22 5 References ...................................................................................................................23 6 Acronyms.....................................................................................................................24 7 Annex ...........................................................................................................................25 7.1 Annex A: Sampling SOP .......................................................................................25 7.2 Annex B: Cleaning procedures ..............................................................................29 List of Tables Table 1: Parameters used in MS-Dial. ..................................................................................16 D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 6 List of Figures Figure 1: Overview of suspect and non-target screening (NTS) workflow .............................. 4 Figure 2: Setup used for filtration. .......................................................................................... 9 Figure 3: Relative signal of 14 internal standards (compared to matrix blanks) in 21 runoff samples and pooled QC sample (both positive and negative ionization mode). ....................12 Figure 4: Average signal suppression in runoff samples vs DOC. ........................................12 Figure 5: Average relative signal of 13 internal standards in selected runoff samples analyzed at varying REF (positive ionization mode). ............................................................13 Figure 6: Data processing workflow for suspect and non-target screening. ..........................14 Figure 7: LC gradient with varying % A and B solvents at 0.3 mL/min flow rate. ...................15 Figure 8: Example of MS-Dial initially detected features in 21 runoff samples (positive mode). Each dot represents one feature with retention time and m/z value. .....................................16 Figure 9: Example of MS-Dial annotated compounds based on matching with suspect list, suggested identification of Terbucarb. ..................................................................................17 Figure 10: Example of suspect list entry (caffeine). ..............................................................18 Figure 11: Example of data used to evaluate the effect of drift correction. ............................18 Figure 12: PCA of features detected in 21 runoff samples with NTS workflow. .....................20 Figure 13: Correlation plot showing features detected in 21 runoff samples with NTS workflow. ..............................................................................................................................21 D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 7 1 INTRODUCTION 1.1 Purpose of the document The purpose of this document is to present the SOP for suspect and non-target screening of runoff samples with LC-HRMS developed and used in D4Runoff. The SOP is a further development of the workflow draft presented in D1.1. 1.1.1 Scope of the document The scope of the document is to provide a detailed SOP for the analytical workflow including sampling, sample preparation, data acquisition on LC-HRMS, and data processing (see Figure 1 for an overview of the workflow). While alternative methods have been tested during the method development (e.g. vacuum-assisted evaporative concentration (VEC) as an alternative to solid-phase extraction (SPE), and SFC-HRMS (supercritical fluid chromatography) as an alternative to LC-HRMS), the main focus is on the core solid-phase extraction (SPE) LC-HRMS workflow which is a benchmark methodology for water analysis that covers a wide range of relevant CECs. 1.1.1.1 Structure of the document The document covers the analytical workflow from sampling to final results in systematic order, i.e. from first to last steps. D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 8 2 SOP for suspect screening and NTS workflows for stormwater 2.1 Sampling A detailed description of the sampling protocol is provided in D1.3, and the full sampling protocol is attached in Annex A. In short, urban runoff samples are collected during the first 2 hours of rain events to get a representative sample. Sub-samples can either be combined to generate one averaged runoff sample (used to develop the inventory of runoff pollutants for D1.3) or analyzed individually as a time-series to investigate changes in runoff composition during the rain event. Both types of samples are treated similarly. After collection, samples are transported to the laboratory for sample preparation and further processing. To minimize changes in the sample composition before analysis, samples are kept cold (around 5 degrees Celsius) and dark, and transported to the laboratory as fast as possible in special transport boxes specifically designed for the purpose (made by GEUS). For samples collected in Denmark, the transportation time below 24 hours is generally achievable, whereas samples collected outside Denmark are processed within 72 hours, as shipping of samples takes longer time. To avoid contamination of samples, sampling equipment undergoes extensive cleaning to avoid contamination of samples (cleaning procedure is described in Annex B). Furthermore, a field blank sample is prepared for approximately every 10th sample, consisting of highpurity water with 300 mg/L CaCl2 added to mimic the ionic strength of runoff. Field blanks are for filtering features detected in suspect screening and NTS (see section 2.3). D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 9 2.2 Sample preparation The following section presents the final SOP for sample preparation for LC-HRMS analysis, which has been further developed from D1.1. Procedures used to clean equipment is presented in Annex B. 2.2.1 Filtration Figure 2: Setup used for filtration. Equipment ◼ Filtering (vacuum) flask (2L) – use cleaning procedure (5) ◼ Büchner glass funnel (90 mm in diameter) – use cleaning procedure (5) ◼ Whatman GF/F 0.7 μm glass microfiber filter – use cleaning procedure (6) ◼ Whatman GF/A 1.6 μm glass microfiber filter – use cleaning procedure (6) ◼ Rubber stopper ◼ Vacuum pump ◼ Hose ◼ Milli-Q water ◼ 20% formic acid (HPLC-grade) in LC-MS grade water ◼ 2% formic acid (HPLC-grade) in LC-MS grade water ◼ 20% ammonium hydroxide in LC-MS grade water ◼ 2% ammonium hydroxide in LC-MS grade water ◼ Acid-washed 1L bottles – use cleaning procedure (2) ◼ Heat-treated Pasteur pipettes – use cleaning procedure (4) First, pH is adjusted to 6.5 with 20% formic acid and ammonium hydroxide (pH in samples is generally >7, so typically only formic acid is needed). Use 2% solutions for blank samples. For filtration, a vacuum pump and Whatman glass microfiber filters with diameter 1,6μm and 0.7μm will be used. To avoid contamination, glass microfiber filters are cleaned with dichloromethane according to procedure (6). D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 16 Table 1: Parameters used in MS-Dial. Peak detection Minimum peak height Mass slice width 1000 amplitude 0.1 Da Deconvolution parameters Sigma window value MS/MS abundance cut off 0.5 100 amplitude Identification Accurate mass tolerance MS1 Accurate mass tolerance MS2 0.01 Da 0.05 Da Alignment MS1 tolerance N% detection in at least one group Sample max / blank average 0.015 Da 100 % 5 fold change Adducts (positive) [M+H]+, [M+NH4]+, [M+K]+, [M+CAN+H]+, [M+H-H2O]+ Adducts (negative) [M-H]-, [M-H2O-H]-, [M+Na-2H]-, [M+FA-H]-, [2M-H]- Only features that occur in all three replicate injections and with signal intensity at least 5 times blank average are kept. Examples of the output from MS-Dial for NTS and suspect screening are shown in Figures 8 for NTS and 9 for suspect screening. Figure 8: Example of MS-Dial initially detected features in 21 runoff samples (positive mode). Each dot represents one feature with retention time and m/z value. D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 17 Figure 9: Example of MS-Dial annotated compounds based on matching with suspect list, suggested identification of Terbucarb. For suspect screening, features are selected which have a suggested annotation by MS-Dial based on comparison with suspect list. This list containing >1600 compounds collected from previous studies on pollutants in runoff and related sources (Gasperi et al., 2022; Masoner et al., 2019; Page et al., 2014; Peter et al., 2022; Zgheib et al., 2011) as well as compounds identified in runoff screenings as part of T1.1 and T1.3. For each compound, the list contains unique identifier (InChi Key and/or SMILES), accurate mass, and fragmentation mass spectra. An example is shown in Figure 10. Mass spectra are either retrieved from the Mass Bank of North America (MoNA) database or generated in-silico if experimental data is not available. D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 18 Figure 10: Example of suspect list entry (caffeine). 2.3.3 Correction of drift and matrix effect The data is manually inspected to evaluate the need for drift correction. If systematic changes in response of individual features and known compounds is found in the QC samples (typically seen as a decrease in signal over time), drift correction is performed based on the measured signals in QC samples by fitting a second order polynomial to the peak intensity of all features in the QC samples over time and correcting features with the derived function (Tisler et al., 2022). Since matrix composition has a strong effect on the ionization efficiency (see section 2.2.3), matrix effect correction is required to compare across samples. This is done based on the total ion chromatogram (Tisler et al., 2021). Validation of drift corrections is done with Super QC samples to check that data has been improved. An example of data used for drift correction validation is shown in Figure 11. Figure 11: Example of data used to evaluate the effect of drift correction. 0.9 0.95 1 1.05 1.1 020 40 60 80 100 Injection number Without drift correction 0.9 0.95 1 1.05 1.1 020 40 60 80 100 Injection number With drift correction D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 19 2.3.4 Inspection of suggested annotations and assigning confidence scores Annotations given by MS-Dial are manually inspected to ensure proper alignment and matching of spectra. This is done by comparing deconvoluted MS/MS spectra with experimental data from MoNA as well as other sources and additional information about the compound (precious occurrences, use, etc.) Identification points are awarded according to: ➢ Chemical formula assigned with accurate mass within 5 ppm (10 points). o 5 points is possible in 10 ppm range ➢ Matching fragments with library (30 points). o For compounds without library experimental library spectra, up to 15 points possible from matching with in-silico fragmentation. ➢ Retention time prediction within 2 min window (20 points). o 10 points possible within 4 min window. ➢ Plausibility of occurrence (10 points). o Based on literature and compound use. ➢ Matching analytical standard (30 points) Up to 100 points is possible through matching with analytical standard, corresponding to level 1 identification (Schymanski et al., 2014). Only features with score above 10 (accurate mass within 5 ppm deviation) will be reported. D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 20 3 Reporting of results Results are reported as lists of tentatively identified compounds (suspect screening) or features (NTS) with relevant information about m/z, mass spectra, and signal intensities. Quantitative non-target analysis can also be done to estimate concentrations of unknown compounds through modelling of response factors (Tisler et al., 2024). Plots are made in R to visualize and interpret results. For example, principal component analysis (PCA) can be used to identify clustering of samples with similar chemical composition (Figure 12). Figure 12: PCA of features detected in 21 runoff samples with NTS workflow. As Figure 12 shows, runoff samples can be clearly differentiated with the NTS workflow. First, the triplicate injections group nicely together which suggests that the determined features describe the samples rather than analytical uncertainties. Also, the plot shows some characteristic trends across samples. Particularly, the SA5 injections (green dots, in the top of the plot) are very different from the rest of the samples, which makes sense as this sample is from a combined sewer overflow. It will therefore have a strong contribution of wastewater and related chemicals. The reason that it is so different from the other combined sewer overflow sample (OD7, red boxes in the plot) could be that SA5 is less “diluted” by runoff and therefore has a chemical fingerprint more heavily affected by wastewater. This is supported by results from target analysis that shows very high concentrations of typical wastewater D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 21 compounds in SA5, for example Paracetamol which was found at 149 µg/L in SA5 compared to 6 µg/L on OD7. Another characteristic group of samples are OD1, CO1, and CO2 (left side of the plot). These are the runoff samples that were collected after the longest antecedent dry period, and which also have the highest signal suppression (see section 2.2.3). This is a likely explanation for their relative similarity compared to the other samples. Correlation plots showing co-occurring features and compounds can also be used to reveal sources and characteristic compound groups (Figure 13). In the figure below, which shows the results from 21 runoff samples analyzed with the NTS workflow, there are several groups of co-occurring compounds that potentially originate from the same sources and can therefore be used for source appointment as well as to identify useful markers of specific types of pollutants. Figure 13: Correlation plot showing features detected in 21 runoff samples with NTS workflow. D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 22 4 Conclusion and next steps SOP for suspect screening and NTS has been developed and implemented for analysis of runoff samples. The methods have been developed to provide robust and reliable results for different types of runoff samples that require different analytical considerations, particularly when it comes to dilution to avoid signal suppression in LC-ESI-HRMS analysis. The next steps will be to use the workflow for analysis of runoff samples in T1.3 and fate study samples in T1.4, as well as testing and validating the methods in WP5, specifically T5.1. Outside D4RUNOFF, the SOP will also be used to analyze samples from across Europe within the PARC project in which UCPH contributes. Additional approaches to identification of unknown pollutants using DDA approaches will also be developed to further improve the ability to identify compounds based on high-quality HRMS spectra. D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 23 5 References Gasperi, J., Le Roux, J., Deshayes, S., Ayrault, S., Bordier, L., Boudahmane, L., Caupos, E., Caubrière, N., Flanagan, K., Guillon, M., Huynh, N., Labadie, P., Meffray, L., Neveu, P., Partibane, C., Paupardin, J., Saad, M., Varnede, L., & Gromaire, M. (2022). Micropollutants in Urban Runoff from Traffic Areas: Target and Non-Target Screening on Four Contrasted Sites. Water, 14(394). https://doi.org/10.3390/ w14030394 Masoner, J. R., Kolpin, D. W., Cozzarelli, I. M., Barber, L. B., Burden, D. S., Foreman, W. T., Forshay, K. J., Furlong, E. T., Groves, J. F., Hladik, M. L., Hopton, M. E., Jaeschke, J. B., Keefe, S. H., Krabbenhoft, D. P., Lowrance, R., Romanok, K. M., Rus, D. L., Selbig, W. R., Williams, B. H., & Bradley, P. M. (2019). Urban Stormwater: An Overlooked Pathway of Extensive Mixed Contaminants to Surface and Groundwaters in the United States. Environmental Science & Technology, 53, 10070–10081. https://doi.org/10.1021/acs.est.9b02867 Page, D., Miotli, K., Gonzalez, D., Barry, K., Dillon, P., & Gallen, C. (2014). Environmental monitoring of selected pesticides and organic chemicals in urban stormwater recycling systems using passive sampling techniques. Journal of Contaminant Hydrology, 158, 65–77. https://doi.org/10.1016/j.jconhyd.2014.01.004 Paton, E., & Haacke, N. (2021). Merging patterns and processes of diffuse pollution in urban watersheds : A connectivity assessment. March. https://doi.org/10.1002/wat2.1525 Peter, K. T., Lundin, J. I., Wu, C., Feist, B. E., Tian, Z., Cameron, J. R., Scholz, N. L., & Kolodziej, E. P. (2022). Characterizing the Chemical Profile of Biological Decline in Stormwater-Impacted Urban Watersheds. Environmental Science & Technology, 56, 3159–3169. https://doi.org/10.1021/acs.est.1c08274 Schymanski, E. L., Jeon, J., Gulde, R., Fenner, K., Ru, M., Singer, H. P., & Hollender, J. (2014). Identifying Small Molecules via High Resolution Mass Spectrometry: Communicating Confidence. Environmental Science & Technology, 2097–2098. Tisler, S., Engler, N., Jørgensen, M. B., Kilpinen, K., Tomasi, G., & Christensen, J. H. (2022). From data to reliable conclusions: Identification and comparison of persistent micropollutants and transformation products in 37 wastewater samples by non-target screening prioritization. Water Research, 219. https://doi.org/10.1016/j.watres.2022.118599 Tisler, S., Kilpinen, K., Pattison, D. I., Tomasi, G., & Christensen, J. H. (2024). Quantitative Nontarget Analysis of CECs in Environmental Samples Can Be Improved by Considering All Mass Adducts. Analytical Chemistry, 96(1), 229–237. https://doi.org/10.1021/acs.analchem.3c03791 Tisler, S., Pattison, D. I., & Christensen, J. H. (2021). Correction of Matrix Effects for Reliable Non-target Screening LC-ESI-MS Analysis of Wastewater. Analytical Chemistry, 93(24), 8432–8441. https://doi.org/10.1021/acs.analchem.1c00357 Tsugawa, H., Cajka, T., Kind, T., Ma, Y., Higgins, B., Ikeda, K., Kanazawa, M., Vandergheynst, J., Fiehn, O., & Arita, M. (2015). MS-DIAL: Data-independent MS/MS deconvolution for comprehensive metabolome analysis. Nature Methods, 12(6), 523– 526. https://doi.org/10.1038/nmeth.3393 Zgheib, S., Moilleron, R., Saad, M., & Chebbo, G. (2011). Partition of pollution between dissolved and particulate phases : What about emerging substances in urban stormwater catchments? Water Research, 45, 913–923. https://doi.org/10.1016/j.watres.2010.09.032 D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 24 6 Acronyms SPE: Solid-phase extraction VEC: Vacuum-assisted evaporative concentration CEC: Contaminants of emerging concern LC-HRMS: Liquid chromatography high-resolution mass spectrometry GC×GC-HRMS: Comprehensive two-dimensional gas chromatography high-resolution mass spectrometry SFC: Supercritical fluid chromatography NTS: Non-target screening SST: System suitability test QC: Quality control REF: Relative enrichment factor DIA: Data-independent acquisition DDA: Data-dependent acquisition ESI: Electrospray ionization PCA: Principal component analysis D1.2: Standard Operating Protocols (SOP) for suspect screening and non-target screening (NTS) analysis workflows for stormwater 25 7 Annex 7.1 Annex A: Sampling SOP DEFINITION OF RAIN EVENT AND PLANNING OF SAMPLING DATE Sampling should preferably be done when is has been dry for at least 3 days and the weather forecast predicts an intense rain event where the start of a rain event is clearly defined. This can be when a front zone passes, which is predictable from forecasts, or thunderstorms and cloudbursts, that are less predictable but also with a clearly defined start. Samples from both types of events can be used. We prefer high-intensity rain events, but any event with >5 mm in two hours is useful. The samples should include the first run-off to have the initial peak concentrations washed off surfaces followed by more dilute concentrations. Therefore, we want samples for up to 120 minutes after the water starts running, but the first four samples (0, 15, 30 and 45 min.) are sufficient if the rain stops. STATION CHARACTERIZATION FORM Fill out relevant boxes, some are mostly relevant for small-scale stations. Date of sampling: Name of the person who carried out the sampling and institution: Short description of site and mode of sampling: For instance: Inlet to retention pond at a residential area, sampled from inlet pipe Rain bed at city street, sampled from inlet Rain bed in residential area, sampled from inlet pipe Street runoff from city street, sampled at a storm drain Stormwater overflow basin at industrial site, sampled from inlet pipe Large retention basin with mixed stormwater, sampled from inlet pipe Overflow/bypass from municipal WWTP, sampled from bypass pipe Manual sampling Autosampler Country: City/town: Position: GPS coordinates from Google Maps on a mobile phone: Estimated time since previous rain event (>1 mm per day) if known: days, otherwise estimate 7-14 days