Datasets:

---

LLMDH

/

OpenScience

like 0

Follow

LLM - Digital Humanities 16

ERROR: type should be string, got "https://doi.org/10.1007/s00216-020-02722-5\nAnalytical and Bioanalytical Chemistry (2020) 412:4941–4952 https://doi.org/10.1007/s00216-020-02722-5\nAnalytical and Bioanalytical Chemistry (2020) 412:4941–4952 RESEARCH PAPER A rapid method for quantification of persistent and mobile organic\nsubstances in water using supercritical fluid chromatography\ncoupled to high-resolution mass spectrometry Stefanie Schulze1 & Heidrun Paschke1 & Till Meier1 & Matthias Muschket1 & Thorsten Reemtsma1 Received: 11 March 2020 /Revised: 8 May 2020 /Accepted: 18 May 2020\n# The Author(s) 2020\n/ Published online: 10 June 2020 Abstract Persistent and mobile organic substances (PM substances) are a threat to the quality of our water resources. While\nscreening studies revealed widespread occurrence of many PM substances, rapid trace analytical methods for their\nquantification in large sample sets are missing. We developed a quick and generic analytical method for highly mobile\nanalytes in surface water, groundwater, and drinking water samples based on enrichment through azeotrope evaporation\n(4 mL water and 21 mL acetonitrile), supercritical fluid chromatography (SFC) coupled to high-resolution mass spec-\ntrometry (HRMS), and quantification using a compound-specific correction factor for apparent recovery. The method\nwas validated using 17 PM substances. Sample preparation recoveries were between 60 and 110% for the vast majority\nof PM substances. Strong matrix effects (most commonly suppressive) were observed, necessitating a correction for\napparent recoveries in quantification. Apparent recoveries were neither concentration dependent nor dependent on the\nwater matrix (surface or drinking water). Method detection and quantification limits were in the single- to double-digit\nng L−1 ranges, precision expressed as relative standard deviation of quadruplicate quantifications was on average < 10%,\nand trueness experiments showed quantitative results within ± 30% of the theoretical value in 77% of quantifications. Application of the method to surface water, groundwater, raw water, and finished drinking water revealed the presence\nof acesulfame and trifluoromethanesulfonic acid up to 70 and 19 μg L−1, respectively. Melamine, diphenylguanidine, p-\ndimethylbenzenesulfonic acid, and 4-hydroxy-1-(2-hydroxyethyl)-2,2,6,6-tetramethylpiperidine were found in high\nng L−1 concentrations. Keywords Persistent and mobile organic substances . PM substances . PMOCs . Supercritical fluid chromatography . Evaporation . Water samples Chemicals and reagents For method development and validation, 17 model PM sub-\nstances were selected (Fig. 1) from a list of PM substances\ndetected in environmental water samples in a qualitative\nscreening study [11]. The target PM substances were selected\nto span a broad range of (yet very low) log D values (estimated\nat pH 7 to −3.06 to 1.23, ChemAxon, JChem for Office,\nversion 19.26.0.571), molecular masses (84 to 361 g mol−1),\nand charge states, as detailed in Table S1 in the Electronic\nSupplementary Material (ESM). Stock standard solutions of\nthe analytes were prepared at 1 mg mL−1 in acetonitrile or\nwater (depending on solubility) and stored in darkness at −\n20 °C. From these stock solutions, mixture solutions of the 17\nPM substances at different concentration levels were pre-\npared. All used chemicals, solvents, and reagents were of an-\nalytical grade (ESM Table S2). Water samples Besides chromatography, the extraction and enrichment of\nPM substances from water samples also poses a challenge [5]. Enrichment is necessary, since direct injection of water sam-\nples into the analytical instrument [23] is often not sensitive\nenough for detection of trace levels in samples from back-\nground areas. Additionally, large-volume injection of water\n(> 10 μL) is not compatible with SFC. Solid-phase extraction\nis the most commonly used method for enrichment of contam-\ninants from water samples. However, retention of PM sub-\nstances on common SPE material is usually poor (again due\nto the high mobility) or very specific, such as, e.g., for nega-\ntively charged PM substances on an anion exchange resin\n[10]. Evaporation [7] or freeze-drying [9] are more generic\nmethods for analyte enrichment with the disadvantage that\nall non-volatile constituents in the sample are quantitatively\nenriched as well, which may lead to significant matrix effects. For method development and validation, surface water sam-\nples from the rivers Götsche and Mulde (near Halle and\nLeipzig, Germany) and drinking water samples from the tap\nin our laboratory were used. For method application, six water\nsamples were obtained from two different regions in Germany\n(South Hessia and Berlin) including surface water, groundwa-\nter, and water from drinking water treatment plants (for details\non the samples, see ESM Table S3). The samples from the\ndrinking water treatment plant Tegel were taken and analyzed\nwith permissions from the drinking water company (Berliner\nWasserbetriebe). Sampling took place between 2017 and\n2019. The samples were stored up to 2 weeks at + 4 °C until\nanalysis. Introduction Published in the topical collection Persistent and Mobile Organic\nCompounds – An Environmental Challenge with guest editors Torsten C. Schmidt, Thomas P. Knepper, and Thorsten Reemtsma. Persistent and mobile organic substances (PM substances, al-\nso referred to as PMOCs) are characterized by a high environ-\nmental stability and a very low potential to sorb to surfaces [1,\n2]. PM substances that are emitted into the environment [3]\nwould thus partition to and stay in the water phase and pene-\ntrate natural (bank filtration, subsurface passage) and technical\n(wastewater treatment plants, drinking water treatment) bar-\nriers in water cycles. Therefore, PM substances are of concern\nregarding the quality of our drinking water resources [4]. Electronic supplementary material The online version of this article\n(https://doi.org/10.1007/s00216-020-02722-5) contains supplementary\nmaterial, which is available to authorized users. * Urs Berger\nurs.berger@ufz.de 1\nDepartment of Analytical Chemistry, Helmholtz Centre for\nEnvironmental Research - UFZ, Permoserstrasse 15,\n04318 Leipzig, Germany The characteristic of high aquatic mobility makes PM sub-\nstances hard to analyze using common reversed phase liquid\nchromatography (RPLC) techniques [1, 5]. Since hydropho-\nbic interactions are the driving force of retention in RPLC, 2\nInstitute of Analytical Chemistry, University of Leipzig, Linnéstrasse\n3, 04103 Leipzig, Germany 4942 Schulze S. et al. was thus to develop and validate a trace analytical method\nbased on SFC coupled to high-resolution mass spectrometry\nas an alternative and potentially complimentary method for\nquantitative analysis of a variety of PM substances in environ-\nmental as well as in drinking water samples. highly polar and thus in-water-mobile compounds are not\nretained and elute in the void volume together with very polar\nmatrix constituents. In recent years, alternative separation\ntechniques were developed for retention and separation of\nhighly mobile substances [5], including ion chromatography\n[6], hydrophilic interaction liquid chromatography (HILIC)\n[7, 8], mixed-mode liquid chromatography (MMLC) [9, 10],\nand supercritical fluid chromatography (SFC) [11, 12]. Experimental section SFC as a separation technique was first reported by Klesper\nand co-workers [13] in 1962. Since then, the number of re-\nports on applications of SFC is continuously increasing [14,\n15]. SFC is often described as an alternative to normal-phase\nchromatography and as the method of choice for\nenantioselective separations, particularly for non-volatile\ncompounds [16]. However, due to the possibilities of using\nreversed phase as well as normal-phase stationary phases and\nto mix a polar co-solvent into supercritical (non-polar) CO2 in\nthe mobile phase, SFC is a very versatile separation technique,\neven encompassing applications for highly polar and mobile\nanalytes [17]. Recent examples include environmental water\npollutants [18, 19], polar urinary metabolites [20], and polar\ncompounds in anti-doping control [21]. Desfontaine and co-\nworkers [22] compared matrix effects in SFC and RPLC\ncoupled to tandem mass spectrometry for analysis of doping\nagents and pharmaceuticals in urine. They found that SFC\ngenerally led to lower matrix effects than RPLC, especially\nwhen applying a simple dilute-and-shoot protocol [22]. Sample preparation In a recent study, we have applied innovative analytical\nmethods for a qualitative screening study of PM substances\nin environmental water samples [11]. Out of 57 target\nanalytes, 43 (75%) were detected in surface water and/or\ngroundwater samples. This high detection percentage under-\nlines the importance of being able to quantify PM substances\nin different types of environmental waters, including drinking\nwater. For ion chromatography, HILIC, and MMLC, quanti-\ntative methods have been explored for PM substances (see\nabove), but not for SFC so far. The aim of the present study The samples were filtered through a glass fiber filter (see ESM\nTable S2 for details on materials and instrumentation used in\nsample preparation). Azeotrope evaporation (AZEVAP) was\nused as enrichment procedure. An aliquot of 4 mL of the\nfiltered sample was mixed with 21 mL acetonitrile (ratio for\nthe minimum azeotrope mixture) in an evaporation glass vial\nwith a tip in the bottom. This mixture was evaporated to dry-\nness at 40 °C under a stream of argon, while the glass walls A rapid method for quantification of persistent and mobile organic substances in water using supercritical... 4943 ACE\nAAMPS\nHHTMP\nBTMA\nMPSA\nMEL\nDMPMA\nATA\nTFMSA\nDMSP\nCG\nTSA\nSAC\n3,4-DMBSA\n2,3-DMBSA\nDCHSS\nDPG\nFig. 1 The 17 selected PM substances with their acronyms used in the present study. For full names and CAS numbers, see ESM Table S1. Ionizable\nsubstances are shown in their charge state at a pH value of 7 ACE\nAAMPS MPSA\nMEL\nTFMSA MEL DMPMA\nTSA\n2,3-DMBSA ACE\nAAMPS\nBTMA\nMPSA\nMEL\nDMPMA\nATA\nTFMSA\nCG\nTSA\nSAC\n3,4-DMBSA\n2,3-DMBSA MPMA\n2,3-DMBSA ACE\nAAMPS\nBTMA\nATA\nTFMSA\n3,4-DMBSA MEL CG\nSAC BTMA 2,3-DMBSA SAC CG 3,4-DMBSA 3,4-DMBSA ATA DPG HHTMP\nDCHSS HHTMP Fig. 1 The 17 selected PM substances with their acronyms used in the present study. For full names and CAS numbers, see ESM Table S1. Ionizable\nsubstances are shown in their charge state at a pH value of 7 ubstances with their acronyms used in the present study. For full names and CAS numbers, see ESM Table S1. Ionizable\nr charge state at a pH value of 7 20,000. Mass calibration on a mass range of m/z 50 to 1200\nwas performed using a calibration solution to generate 17\nreference masses in positive and 16 in negative ionization\nmode. A root mean square residual mass error < 1 ppm was\nobtained. Sample preparation During measurements in both ionization modes, a\nlock-spray containing leucine enkephalin was continuously\ninfused. Two ions were selected for identification of the PM\nsubstances (except for MPSA and CG that produced only one\nion), and the most intense ion was used for quantification\n(ESM Table S4). A mass tolerance of 5 ppm was used when\nextracting high-resolution mass chromatograms of the\nanalytes. were repeatedly rinsed with acetonitrile to make sure that the\nresidue concentrated in the tip. The residue was reconstituted\nin 100 μL acetonitrile:Milli-Q water (90:10), resulting in a\nsample-to-extract enrichment factor of 40. In case of precipi-\ntation, the extract was filtered through a lint-free paper wipe\ncovering the tip of a Pasteur pipette while it was transferred\ninto an autoinjector vial. Procedural blanks and method detection and quantification\nlimits Procedural blank experiments were performed by applying\nthe full sample preparation procedure but without any water\nmatrix in the enrichment step (i.e., starting from 21 mL pure\nacetonitrile). Five replicates of procedural blanks were pre-\npared. The method detection limit (MDL) and method quan-\ntification limit (MQL) were determined by spiking surface\nwater and drinking water samples at two to three different\nconcentrations per analyte (ESM Table S6) and quantifying\nthem according to the described protocol. The signal-to-noise\nratios were recorded, and the quantified concentrations were\nextrapolated (from a signal with a signal-to-noise ratio close to\n10) to a signal-to-noise ratio of 3 (MDL) or 10 (MQL). In case\nof procedural blank contamination, the MDL and MQL were\ncalculated from the quantified signal areas in the procedural\nblank chromatograms based on mean signal area plus 3 times\n(MDL) or 10 times (MQL) standard deviation. Instrumental blanks, instrumental detection limits,\nand linearity ¼\nnetAreaPM substance spiked before enrichment\nAreaPM substance in pure solvent\n\u0001\n\u0003\n\u0002 100\nð3Þ ð3Þ Instrumental blank contamination was evaluated by sol-\nvent injections (acetonitrile:Milli-Q water 90:10) into the\nSFC-HRMS system. Instrumental detection limits (IDLs)\nand the linear range of detection were determined using a\ndilution series (n = 10) of the standard mixture (consisting\nof the 17 PM substances) covering a concentration range\nof 0.05–500 ng mL−1. The coefficient of determination\n(R2) for linear regression was calculated. IDLs were set\nfor each PM substance to the injected amount, leading to\na signal in the extracted high-resolution mass chromato-\ngram with a signal-to-noise ratio of at least 3. In case of\ninstrumental blank contamination, the IDL was calculated\nfrom the quantified signal areas in 10 solvent blank injec-\ntions based on mean plus 3 times standard deviation of\nthe signal areas in the 10 blanks. Further, matrix effects on the chromatography were\nassessed qualitatively by comparison of chromatograms (re-\ntention times and signal shape) from standards in pure solvent\nand from spiked extracts of environmental water samples. Method performance validation ME %\nð\nÞ ¼\nnetAreaPM substance spiked after enrichment\nAreaPM substance in pure solvent\n\u0001\n\u0003\n\u0002 100−100\nð2Þ We validated the method by determining instrumental blanks,\ninstrumental detection limits, linear range of detection, appar-\nent recoveries (i.e., sample preparation recoveries and matrix\neffects), procedural blanks, method detection and quantifica-\ntion limits, and accuracy (i.e., precision and trueness). \u0002 100−100\nð2Þ ð2Þ \u0002 100−100 Finally, the apparent recovery (combination of Recov and\nME) was calculated according to Eq. (3) Apparent recovery %\nð\nÞ Quantification concentrations differed between the PM substances based on\nthe differences in IDLs. Additionally, both water matrices\nwere also enriched and analyzed without spiking. Areas of\nPM substances in the chromatograms of the non-spiked sam-\nples were subtracted from areas in the chromatograms of the\nrespective spiking experiments (→netArea). Quantification was performed using an external 5-point cali-\nbration curve in pure solvent and applying a compound-\nspecific correction factor for the apparent recovery (corre-\nsponding to a matrix- and method-matched calibration). The\ncalculation of the apparent recovery is explained in the section\n“Apparent recoveries (sample preparation recoveries and ma-\ntrix effects)”, and the applied values are listed in ESM\nTable S6. The correction factor was calculated from the mean\nof the apparent recoveries determined in different experiments\n(varying in spike concentrations and water matrices), as nei-\nther the spike concentration nor the water matrix (surface or\ndrinking water) had a significant influence on the apparent\nrecovery (see “Results and discussion” section below). The sample preparation recovery (Recov) was calculated\naccording to Eq. (1) Recov %\nð\nÞ ¼\nnetAreaPM substance spiked before enrichment\nnetAreaPM substance spiked after enrichment\n\u0001\n\u0003 ð1Þ \u0002 100 The matrix effect (ME) in ionization was calculated ac-\ncording to Eq. (2) Instrumental analysis SFC (Waters Acquity UPC2 system) was performed on a\nBEH column (for analyses in positive ion mode) or Torus\nDiol column (for analyses in negative ion mode) coupled to\nquadrupole time-of-flight high-resolution mass spectrometry\n(HRMS; Waters Synapt GS2) (ESM Table S2). Aliquots of\n10 μL of the sample extracts were injected. Separation was\nperformed at 55 °C at a flow rate of 1.5 mL min−1 using a\ncarbon dioxide-methanol/water gradient containing 0.2% am-\nmonium hydroxide in the methanol/water co-solvent (ESM\nFig. S1A). A methanol/water make-up flow containing 0.1%\nformic acid was used at 0.3 mL min−1 to transfer the column\neffluent into the mass spectrometer. The HRMS instrument\nwas operated in positive or negative electrospray ionization\n(ESI) and full scan mode (m/z 50 to 600) at a resolution of As a reference for SFC separation and retention, commonly\nused RPLC based on a C18 stationary phase (Waters Acquity\nUPLC HSS T3 column) was used (ESM Table S2). The\nRPLC was coupled to triple quadrupole tandem mass spec-\ntrometry (MS/MS). Aliquots of 10 μL of the sample extracts\nwere injected in water. Separation was performed at 60 °C at a\nflow rate of 500 μL min−1 using a water/methanol gradient\ncontaining 5 mM ammonium formate (ESM Fig. S1B). The\nmass spectrometer was operated in positive/negative-\nswitching ESI mode. Scheduled multiple reaction monitoring\n(MRM) mode was applied, acquiring two transitions for each\nanalyte (ESM Table S5). 4944 Schulze S. et al. Accuracy (precision and trueness) BEH 2-EP (all from Waters), which can be classified as nor-\nmal phase or hybrid stationary phases. The mobile phase\nconsisted of supercritical CO2 and a methanol/water co-\nsolvent (ESM Fig. S1A) with formic acid, ammonium hy-\ndroxide, ammonium formate, or ammonium acetate as modi-\nfier. Ammonium hydroxide and formic acid slightly improved\npeak shape and response for most PM substances and were\nsuperior to the other two tested modifiers. Ammonium hy-\ndroxide was chosen for the final method and added at an\noptimized ratio of 0.1% to the co-solvent. In terms of station-\nary phases, BEH (a hybrid stationary phase showing both\nreversed and normal phase characteristics) and Torus Diol (a\nnormal phase) showed the best performances. Regarding\nchromatography, all analytes were well retained (see ESM\nTable S4 for retention factors) and showed sharp peaks (see\nthe section “Retention time stability and influence of sample\nmatrix on the chromatography”) on both of these columns. However, while PM substances that were recorded in positive\nionization mode generally showed a slightly better response\nafter separation on the BEH column, Torus Diol led to slightly\nmore sensitive detection for analytes in negative ionization\nmode. Both columns were thus used in the final method, one\nfor each polarity of mass spectrometric detection. For a higher\nsample throughput with polarity-switching MS, any of the two\ncolumns could be used without a substantial loss of\nsensitivity. For evaluation of precision and trueness, the following set of\nexperiments (independent from the earlier experiments for de-\ntermination of sample preparation recoveries and matrix ef-\nfects) was performed. All PM substances were spiked (n = 4)\ninto surface water from the river Götsche (see ESM Table S7\nfor compound-specific spiking levels). From these experiments\n(including non-spiked river Götsche water), the apparent recov-\neries and correction factors were calculated as described before. The PM substances were also spiked into surface water from\nthe river Mulde and into drinking water (each n = 4, the spiking\nconcentrations for Mulde water were 0.3 times and those for\ndrinking water were 0.03 times the concentrations spiked to\nwater from river Götsche; see ESM Table S7). All samples\nwere analyzed and quantified using the correction factors deter-\nmined for river Götsche. The quantified concentrations were\ncorrected with levels determined in the corresponding non-\nspiked samples. Effect of sample diluent and injection volume The most commonly used enrichment method for organic\ntrace pollutants from water samples is SPE. SPE has also been\nused in two methods for PM substances published earlier [7,\n10]. However, SPE sorbents are usually designed to selective-\nly retain certain groups of chemicals (e.g., only anions or only\ncations). Therefore, Zahn and co-workers [7, 24] developed\ntheir own homemade SPE cartridges from three different sor-\nbents while Montes et al. [10] subjected each water sample to\ntwo different SPE procedures. We attempted to develop a\nquick and generic method for a broad range of PM substances,\nwhich can be used in larger screening or monitoring programs. Therefore, we used AZEVAP, which requires very little sam-\nple handling and is applicable to all analytes, but also leads to\nenrichment of all other non-volatile constituents in the sam-\nples. Any generically applicable enrichment method would\ninherently also enrich the majority of matrix compounds. The composition of the injection solvent as well as the injec-\ntion volume play an important role in SFC, affecting both\npeak shape and intensity [25, 26]. A good compromise be-\ntween compound solubility, SFC compatibility, sensitivity,\nand peak shape needs to be found in a multi-analyte method. In the present study, different solvent compositions of aceto-\nnitrile and Milli-Q water (from 0 to 100% acetonitrile) were\ntested. Peak shapes and signal intensities varied considerably\ndepending on diluent composition. Pure Milli-Q water provid-\ned broad peaks and low intensity for all of the tested\nchemicals, with the exception of TFMSA. TFMSA showed\na very sharp peak when injected in water (Fig. 2a), but a split\npeak in the presence of acetonitrile (Fig. 2b). An explanation\nfor this peculiar behavior of TFMSA was not found. The best\ncompromise considering all tested PM substances was\nacetonitrile:water 90:10 as injection solvent, providing better\npeak areas than pure acetonitrile and good peak shapes\n(Fig. 2a, with the exception of TFMSA). The high proportion\nof 90% acetonitrile also allowed using the maximum possible\ninjection volume of 10 μL provided by the instrument without\ncompromising peak sharpness. With these settings, TFMSA\nshowed a split peak for both standards and samples (Fig. 2b)\nand was integrated as the sum of the two signal areas. Accuracy (precision and trueness) The relative standard deviations of the quanti-\nfication (n = 4) were used as a measure of method precision. To\nassess trueness, the averaged quantified concentrations were\ncompared to the theoretical (spiked) concentrations. Apparent recoveries (sample preparation recoveries\nand matrix effects) Sample preparation recovery and matrix effect experiments\nwere performed using surface water from the river Götsche\nand drinking water from the tap in the laboratory (ESM\nTable S3). All experiments were performed in triplicates and\nanalyzed by SFC-HRMS. Each PM substance was spiked at\ntwo to three different concentrations in both water matrices\n(ESM Table S6) before and after enrichment. Spike A rapid method for quantification of persistent and mobile organic substances in water using supercritical... 4945 Selection of stationary and mobile phase The two peaks for DMBSA are\nthe chromatographically resolved\n3,4-isomer and 2,3-isomer. For\ncomparison, panel c shows the\nRPLC-MS/MS-extracted MRM\nchromatograms of a standard\nmixture injected in Milli-Q water. Note the different retention time\nscales between the SFC and the\nRPLC chromatograms a normalized signal intensity Retention time stability and influence of sample matrix\non the chromatography\nAs can be seen from Fig. 2a and b, matrix effects on chro-\nmatographic retention resulting from the surface water matrix\nwere only observed for BTMA with a shift to a shorter reten-\ntion time. The relative standard deviation (n = 10) of the\nretention times for all PM substances analyzed over a couple\nof days was < 0.1% in both standard mixtures and sample\nextracts. An important requirement for reproducible retention\ntimes was, however, a freshly prepared co-solvent with mod-\nifier (at least every other day). The co-extracted matrix had an\neffect of slight peak broadening on some target analytes,\nwhich was most pronounced for MPSA and ATA (Fig. 2a, b). 3\n4\n5\n6\n7\nACE ( )\nretention time [min]\n3\n4\n5\n6\n7\nACE (-)\nAAMPS (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nMEL (+)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nDMSP (-)\nCG (+)\nTSA (-)\nSAC (-)\nDMBSA (-)\nDCHSS (-)\nDPG (+)\nretention time [min]\nnormalized signal intensity\n0\n1\n2\n3\n4\n5\n6\n7\nACE (-)\nAAMPS (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nMEL (+)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nDMSP (-)\nCG (+)\nTSA (-)\nSAC (-)\nDMBSA (-)\nDCHSS (-)\nDPG (+)\nretention time [min]\nnormalized signal intensity\nc\nb\np\nthe chromatographically resolved\n3,4-isomer and 2,3-isomer. For\ncomparison, panel c shows the\nRPLC-MS/MS-extracted MRM\nchromatograms of a standard\nmixture injected in Milli-Q water. Selection of stationary and mobile phase Note the different retention time\nscales between the SFC and the\nRPLC chromatograms retention time [min]\n3\n4\n5\n6\n7\nACE (-)\nAAMPS (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nMEL (+)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nDMSP (-)\nCG (+)\nTSA (-)\nSAC (-)\nDMBSA (-)\nDCHSS (-)\nDPG (+)\nnormalized signal intensity\nb b normalized signal intensity 0\n1\n2\n3\n4\n5\n6\n7\nACE (-)\nAAMPS (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nMEL (+)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nDMSP (-)\nCG (+)\nTSA (-)\nSAC (-)\nDMBSA (-)\nDCHSS (-)\nDPG (+)\nretention time [min]\nnormalized signal intensity\nc c Retention time stability and influence of sample matrix\non the chromatography\nAs can be seen from Fig. 2a and b, matrix effects on chro-\nmatographic retention resulting from the surface water matrix\nwere only observed for BTMA with a shift to a shorter reten-\ntion time. The relative standard deviation (n = 10) of the\nretention times for all PM substances analyzed over a couple\nof days was < 0.1% in both standard mixtures and sample\nextracts. An important requirement for reproducible retention\ntimes was, however, a freshly prepared co-solvent with mod-\nifier (at least every other day). The co-extracted matrix had an\neffect of slight peak broadening on some target analytes,\nwhich was most pronounced for MPSA and ATA (Fig. 2a, b). Retention time stability and influence of sample matrix\non the chromatography retention times for all PM substances analyzed over a couple\nof days was < 0.1% in both standard mixtures and sample\nextracts. An important requirement for reproducible retention\ntimes was, however, a freshly prepared co-solvent with mod-\nifier (at least every other day). The co-extracted matrix had an\neffect of slight peak broadening on some target analytes,\nwhich was most pronounced for MPSA and ATA (Fig. 2a, b). Selection of stationary and mobile phase In the development of the SFC method, four different station-\nary phases and four modifiers in the co-solvent of the mobile\nphase were tested in a 4 × 4 matrix. The tested columns (sta-\ntionary phases) included Torus Diol, Torus 2-PIC, BEH, and e of sample matrix\nmatrix effects on chro-\nthe surface water matrix\na shift to a shorter reten-\nviation (n = 10) of the\nretention times for all PM substances analyzed over a couple\nof days was < 0.1% in both standard mixtures and sample\nextracts. An important requirement for reproducible retention\ntimes was, however, a freshly prepared co-solvent with mod-\nifier (at least every other day). The co-extracted matrix had an\neffect of slight peak broadening on some target analytes,\nwhich was most pronounced for MPSA and ATA (Fig. 2a, b). 3\n4\n5\n6\n7\nACE (-)\nAAMPS (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nMEL (+)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nDMSP (-)\nCG (+)\nTSA (-)\nSAC (-)\nDMBSA (-)\nDCHSS (-)\nDPG (+)\nretention time [min]\nnormalized signal intensity\n3\n4\n5\n6\n7\nACE (-)\nAAMPS (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nMEL (+)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nDMSP (-)\nCG (+)\nTSA (-)\nSAC (-)\nDMBSA (-)\nDCHSS (-)\nDPG (+)\nretention time [min]\nnormalized signal intensity\n0\n1\n2\n3\n4\n5\n6\n7\nACE (-)\nAAMPS (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nMEL (+)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nDMSP (-)\nCG (+)\nTSA (-)\nSAC (-)\nDMBSA (-)\nDCHSS (-)\nDPG (+)\nretention time [min]\nnormalized signal intensity\na\nc\nb\nSchulze S. et al. Schulze S. et al. 4946 3\n4\n5\n6\n7\nACE (-)\nAAMPS (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nMEL (+)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nDMSP (-)\nCG (+)\nTSA (-)\nSAC (-)\nDMBSA (-)\nDCHSS (-)\nDPG (+)\nretention time [min]\nnormalized signal intensity\na Fig. 2 SFC-HRMS extracted\nmass chromatograms of a\nstandard mixture of the target PM\nsubstances sorted by log D a\ninjected in acetonitrile:water\n90:10 except for TFMSA (in pure\nwater) and b spiked to surface\nwater and, after extraction,\ninjected in acetonitrile:water\n90:10 for all compounds. (+)\nindicates BEH chromatography\nand detection in positive ion\nmode, and (−) indicates Torus\nDiol chromatography and\ndetection in negative ion mode. Comparison with RPLC recoveries, i.e., the combination of sample preparation recov-\neries and matrix effects, were investigated in order to evaluate\nif an external calibration curve of standards in pure solvent\ncould be applied. For AAMPS and DMSP, the spiking con-\ncentrations were too low to reliably determine apparent recov-\neries. These two PM substances were excluded from further\nquantitative work, but still these were analyzed qualitatively. For all other analytes, apparent recoveries varied consider-\nably, as representatively illustrated for one drinking water\nsample and one surface water sample in Fig. 3a. Thus, the\nquantification procedure had to include a correction for appar-\nent recoveries. However, the apparent recoveries did not differ\nsignificantly between the drinking water and the surface water\n(Fig. 3a), nor were they concentration dependent. This obser-\nvation was later confirmed during the accuracy testing using a\ndifferent set of surface water and drinking water samples (see\nbelow). Therefore, an average compound-specific correction\nfactor was calculated from both sample types at all tested\nconcentrations and used in the quantification (ESM\nTable S6). The apparent recoveries for ACE and SAC obtain-\ned in our study were comparable with previously reported\nrecoveries from water samples, e.g., by Tran et al. [27] or\nMontes et al. [10]. Tran and co-workers [27] used SPE enrich-\nment with subsequent LC-MS/MS analysis. Montes at al. [9,\n10] further published an apparent recovery for DPG (80%)\nfrom water samples after enrichment using a cation exchanger\nand analysis by MMLC-HRMS or LC-MS/MS, which is\nhigher than our apparent recovery for DPG (61%) (Fig. 3a,\nESM Table S6). For comparison of chromatographic method performance, a\nRPLC-MS/MS method using the MRM mode was also devel-\noped for the 17 target PM substances (see ESM, Tables S2 and\nS5 and Fig. S1). Chromatography was optimized to obtain the\nbest possible retention on the C18 column for a maximum of\nanalytes. For this purpose, three different columns (two polar-\nmodified C18 materials and a porous graphitic carbon column)\nwere tested with formic acid, ammonium formate, or diethyl\namine in the mobile phase. Anyhow, in the optimized system\n(described in the “Experimental section”), 8 of the PM sub-\nstances eluted very close to or in the void volume, preventing\ncomplete separation of the DMBSA isomers (Fig. 2c). Of the\n9 retained substances, 6 showed very poor peak shapes. Instrumental blanks, instrumental detection limits,\nand linearity Sample preparation recoveries and matrix effects were inves-\ntigated independently to better understand the variability in\napparent recoveries. Sample preparation recoveries were be-\ntween 60 and 110% for the vast majority of PM substances for\nboth water types (Fig. 3b) and could thus not (fully) explain\nthe partially very low apparent recoveries (see, e.g., ACE or\nDCHSS in Fig. 3a and b). HHTMP, MEL, TFMSA, and DPG were repeatedly detected\nin instrumental blanks. In the present study, no effort was\nmade to elucidate or eliminate the sources of these back-\nground contaminations. However, instrumental blanks were\nconsidered in the determination of IDLs, as detailed in the\n“Experimental section.” SFC-HRMS-based IDLs are summa-\nrized in Table 1 for all studied PM substances and range be-\ntween 0.1 and 5 pg for SFC with BEH and between 0.02 and\n10 pg for SFC with Torus Diol. Good linearity of the instru-\nmental method was observed for all analytes over at least 3\norders of magnitude (with the exception of the guanidines CG\nand DPG with smaller linear ranges) with correlation coeffi-\ncients (R2) higher than 0.99 (Table 1) and residuals < 25%. Retention time stability and influence of sample matrix\non the chromatography As can be seen from Fig. 2a and b, matrix effects on chro-\nmatographic retention resulting from the surface water matrix\nwere only observed for BTMA with a shift to a shorter reten-\ntion time. The relative standard deviation (n = 10) of the A rapid method for quantification of persistent and mobile organic substances in water using supercritical... 4947 Comparison with RPLC Only\nDMSP, DCHSS, and DPG showed good retention and sharp\npeaks in RPLC, the latter two being the PM substances with\nthe highest log D values among the tested analytes. In com-\nparison, the SFC method is clearly superior to the RPLC\nmethod in terms of peak shapes and retention, which consid-\nerably facilitates signal detection and integration (compare\nFig. 2a and c). Furthermore, SFC was able to separate the\ntwo isomeric compounds (3,4-DMBSA and 2,3-DMBSA)\n(Fig. 2a, b). While RPLC showed a slight tendency towards\nhigher retention for analytes with higher log D values (Fig. 2c,\nESM Table S5), no association between log D value and re-\ntention factor could be observed in SFC. Matrix effects Matrix effects were an important reason for non-quantitative\napparent recoveries in the present study, presumably mainly\ninfluencing the ESI process. Matrix effects are depicted in\nFig. 3c as relative deviation of the signal area in the chromato-\ngram of a spiked sample extract compared to a standard in\npure solvent. Despite good retention of all analytes in SFC\n(Fig. 2a, b), strong suppression of the chromatographic signal\nby matrix was observed for 4 analytes (ACE, MPSA, 3,4-\nDMBSA, DCHSS) in both water matrices (Fig. 3c). Consistent signal enhancement with up to + 41% was only\nobserved for 3 PM substances (MEL, TSA, 2,3-DMBSA). Apparent recoveries For the vast majority of investigated PM substances, isotope-\nlabeled analogues that could be employed as internal stan-\ndards are not commercially available. Therefore, an external\nquantification method needed to be developed. Apparent 4948 Schulze S. et al. Table 1\nInstrumental detection limits (IDLs), linear ranges with correlation coefficients (R2), as well as method detection and quantification limits\n(MDLs/MQLs) for the analysis of 17 PM substances by SFC-HRMS\nAnalyte\nIDLBEH (pg)\nLinear range,\nng mL−1 (R2)\nIDLTorus Diol (pg)\nLinear range,\nng mL−1 (R2)\nMDL/MQL\nAZEVAP–BEH\n(ng L−1)\nAZEVAP–Torus Diol\n(ng L−1)\nACE (−)\n0.6\n–\n0.5\n0.05–500 (0.998)\n–\n14/33\nAAMPS (−)\n0.2\n–\n0.3\n0.05–75 (0.998)\n–\n10/30\nHHTMP (+)\n0.2\n0.05–15 (0.998)\n0.09\n–\n5/14\n–\nBTMA (+)\n0.3\n0.05–150 (0.999)\n0.3\n–\n3/10\n–\nMPSA (−)\n3\n–\n0.4\n0.5–150 (0.997)\n–\n50/90\nMEL (+)\n0.3\n0.05–75 (0.998)\n4\n–\n4/10\n–\nDMPMA (+)\n0.3\n0.05–75 (0.999)\n0.6\n–\n3/10\n–\nATA (+)\n0.2\n0.5–150 (0.995)\n4\n–\n2/5\n–\nTFMSA (−)\n0.1\n–\n0.05\n0.05–150 (0.998)\n–\n4/5\nDMSP (−)\n0.1\n–\n0.1\n0.05–150 (0.999)\n–\n10/30\nCG (+)\n5\n5–75 (0.990)\n10\n–\n30/61\n–\nTSA (−)\n0.5\n–\n0.2\n0.05–300 (0.997)\n–\n14/31\nSAC (−)\n0.8\n–\n0.8\n0.5–500 (0.993)\n–\n15/30\n3,4-DMBSA (−)\n1\n–\n0.5\n0.05–150 (0.998)\n–\n5/8\n2,3-DMBSA (−)\n0.1\n–\n0.02\n0.05–500 (0.998)\n–\n26/42\nDCHSS (−)\n0.3\n–\n0.08\n0.05–75 (0.996)\n–\n10/30\nDPG (+)\n0.3\n0.05–15 (0.999)\n0.3\n–\n33/71\n– Table 1\nInstrumental detection limits (IDLs), linear ranges with correlation coefficients (R2), as well as method detection and quantification limits\n(MDLs/MQLs) for the analysis of 17 PM substances by SFC-HRMS limits (IDLs), linear ranges with correlation coefficients (R2), as well as method detection and quantification limits\nf 17 PM substances by SFC-HRMS Matrix effects were largely comparable for the two different\ntypes of water. In an earlier study by Montes et al. [10] based\non weak anion exchange or weak cation exchange enrichment\nof PM substances from water samples and MMLC-MS/MS\nanalysis, very strong matrix suppression was also observed. However, Montes and co-workers [9, 10] also frequently ob-\nserved matrix enhancement with up to + 150%. Thus, SPE\ndoes not necessarily produce “cleaner” extracts that lead to\nless matrix effects than the generic AZEVAP method. Apparent recoveries This\nfinding is corroborated by the results by Köke and co-workers\n[24], who compared matrix effects between PM substances\nspiked to water extracts enriched by mixed-mode SPE and\nthe same substances spiked to water extracts enriched by\nevaporation, in both cases analyzed by HILIC-MS/MS. The\nresults showed that for 9 of 26 investigated substances (35%),\nmatrix effects exceeded + 50% or −50% in the SPE extracts,\nwhile this fraction was only 2/26 (8%) in evaporation extracts. Svan et al. [28] compared matrix effects on a range of phar-\nmaceuticals between SFC-HRMS and RPLC-HRMS for a\nvariety of matrices, including wastewater influent and efflu-\nent. They concluded that in both techniques, strong matrix\neffects occurred. In RPLC-HRMS, signal enhancements were\ncommonly observed, while SFC-HRMS more often led to\nsignal suppression. Signal suppression for the vast majority\nof analytes was also observed in our study (Fig. 3c). samples is inherently extremely challenging, since PM sub-\nstances possess very similar physical-chemical properties as oth-\ner organic constituents in water (i.e., dissolved organic matter). In\nour quantification method, we correct for matrix effects (as well\nas for sample preparation recoveries) by applying a compound-\nspecific correction factor based on the observed apparent recov-\neries (ESM Table S6). This approach corresponds to using a\nmatrix- and method-matched calibration and can easily be ap-\nplied to large sample sets. Another (potentially more accurate but\nalso much more laborious) approach would be the standard ad-\ndition method, which results in compound-, method-, and even\nsample-specific corrections of apparent recoveries. Alternatively,\nstandard addition over final extracts can be performed to correct\nfor matrix effects, but not for sample preparation recoveries [10]. To simplify quantification methods (and to improve their preci-\nsion, trueness, and comparability), synthesis of stable isotope-\nlabeled internal standards for the most important PM substances\nshould be envisaged. Procedural blanks and method detection and quantification\nlimits Seven of the target compounds were detected in procedural\nblank samples. These were the PM substances already present\nin instrumental blanks (HHTMP, MEL, TFMSA, and DPG)\nas well as TSA, 3,4-DMBSA, and 2,3-DMBSA. Such proce-\ndural blank contamination was also observed in our earlier Mitigation of matrix effects (i.e., separation of analytes from\nmatrix constituents) in analysis of PM substances in water A rapid method for quantification of persistent and mobile organic substances in water using supercritical... A rapid method for quantification of persistent and mobile organic substances in water using supercritical... 4949 A rapid method for quantification of persistent and mobile organic substanc Fig. 3 a Apparent recoveries, b sample preparation recoveries, and c matrix\neffects of the target PM substances spiked at different concentrations (see\nESM Table S6) into drinking water (DW) and surface water (SW). Error\nbars indicate standard deviations (n ≥3). (+) or (−) indicates if the analyte\nwas detected in positive or negative ion mode, respectively and as vulcanization agents in polymerization processes. It is\nnot unlikely that trace level contamination with such\nchemicals occurs from labware, like SPE cartridges, pipette\ntips, sealing, and tubing, or from solvents and reagents applied\nin the analytical method, though we did not attempt to eluci-\ndate the specific sources of contamination for the different\nanalytes. Comparable MDLs and MQLs were found for both water\ntypes (DW and SW), and the values are thus presented in\nTable 1 independent of the water matrix. MDLs and MQLs\nare typically in the low ng L−1 range, with MDLs ranging\nfrom 2 to 50 ng L−1 and MQLs from 5 to 90 ng L−1, depending\non the compound. Our MDLs and MQLs are generally in the\nsame range as values reported for PM substances analyzed by\nSPE enrichment and MMLC-MS/MS [10]. For the sweeteners\nacesulfame and saccharin, previous studies reported MQLs of\n0.1 ng L−1 [29] or 25 ng L−1 [30], which are lower or similar\ncompared to our study (see Table 1). However, it must be\ntaken into account that the literature data was based on tandem\nMS quantification in selected reaction monitoring mode,\nwhich is more sensitive than detection by full-scan HRMS. Accuracy (precision and trueness) Spiking levels for river\nMulde and drinking water were\n0.3 and 0.03 times the levels of\nriver Götsche. See ESM Table S7\nfor actual spiking concentrations\nand for numerical results ACE (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nTSA (-)\nSAC (-)\n3,4-DMBSA (-)\n2,3-DMBSA (-)\nDCHSS (-)\nDPG (+)\n0\n25\n50\n75\n100\n125\n150\n175\n Mulde surface water\n drinking water\nrelative concentration [%] Fig. 4 Accuracy experiments. Precision (error bars indicate the\nstandard deviation of\nquantification, n = 4) and trueness\n(mean value of n = 4 relatively to\ntheoretical (spiked) concentra-\ntions indicated as 100%) for ana-\nlyte quantification in the spiked\nriver Mulde and drinking water\nsamples (after subtraction of\nlevels present in the non-spiked\nsamples). For calculation of the\ncorrection factors, the spiked river\nGötsche water samples were\nused. Spiking levels for river\nMulde and drinking water were\n0.3 and 0.03 times the levels of\nriver Götsche. See ESM Table S7\nfor actual spiking concentrations\nand for numerical results water, groundwater, unventilated raw water, and finished\ndrinking water; for sample details, see ESM Table S3). Quantification was performed based on correction factors giv-\nen in ESM Table S6, which were determined together with the\nsix samples, and the results are presented in Fig. 5 and ESM\nTable S8. Of the 15 target compounds, 9 were detected in at\nleast one of the samples from Berlin and Hessia and 4 PM\nsubstances were found in at least four samples. Six com-\npounds were even found in the final drinking water. This is\nunderlining the importance of accurate quantification methods\nfor PM substances as a crucial prerequisite for future risk\nassessment. For most of the compounds, the results indicated\nlow concentrations in the water samples, typically below\n100 ng L−1. However, ACE and TFMSA both exceeded\n10 μg L−1 in a groundwater sample from Hessia. These find-\nings are in agreement with the few reports from the literature\non PM substances [7, 31, 32]. HHTMP, MEL, and 3,4-\nDMBSA (also reported by Betowski et al. [33]) and DPG\n(also reported by Tang et al. [34]) exceeded 100 ng L−1 in\nsingle samples but did not reach μg L−1 concentrations in\nour limited sample set. drinking water. These values are in the same range as the preci-\nsion reported by Montes et al. Accuracy (precision and trueness) The results of the precision and trueness evaluation are given in\nFig. 4 and ESM Table S7. The spiking concentration for river\nMulde and drinking water samples was chosen to reflect actual\nlevels occurring in surface or tap water, while for river Götsche,\nhigher concentrations were chosen to reliably determine the cor-\nrection factors for apparent recoveries (ESM Table S7). The\ncorrection factors determined earlier (ESM Table S6) were not\nused in this experiment, as there were several months between\nthe first apparent recovery experiments (ESM Table S6) and the\naccuracy experiments (ESM Table S7) and compound-specific\ncorrection factors can change over time, potentially due to fluc-\ntuations in instrumental (ionization) performance. We therefore\nrecommend to calculate apparent recoveries and correction fac-\ntors for every sampling campaign. For MEL and CG, accuracy\nexperiments could not be performed, due to the high concentra-\ntion of MEL already present in the non-spiked river Götsche and\ndue to the very small linear range of detection for CG (see\nTable 1). For all other target PM substances, precision, expressed\nas relative standard deviation of 4 replicate quantifications of\nspiked water samples (ESM Table S7), was in the range 4–\n14% (average ± standard deviation 8.5 ± 3.6%) for river Mulde\nand 2–32% (average ± standard deviation 9.8 ± 8.2%) for qualitative screening study [11]. These substances are high-\nproduction volume industrial chemicals (all > 100 t year−1)\nmainly used as plasticizers, as processing aids in polymers, 4950 Schulze S. et al. ACE (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nTSA (-)\nSAC (-)\n3,4-DMBSA (-)\n2,3-DMBSA (-)\nDCHSS (-)\nDPG (+)\n0\n25\n50\n75\n100\n125\n150\n175\n Mulde surface water\n drinking water\nrelative concentration [%] ACE (-)\nHHTMP (+)\nBTMA (+)\nMPSA (-)\nDMPMA (+)\nATA (+)\nTFMSA (-)\nTSA (-)\nSAC (-)\n3,4-DMBSA (-)\n2,3-DMBSA (-)\nDCHSS (-)\nDPG (+)\n0\n25\n50\n75\n100\n125\n150\n175\n Mulde surface water\n drinking water\nrelative concentration [%]\nFig. 4 Accuracy experiments. Precision (error bars indicate the\nstandard deviation of\nquantification, n = 4) and trueness\n(mean value of n = 4 relatively to\ntheoretical (spiked) concentra-\ntions indicated as 100%) for ana-\nlyte quantification in the spiked\nriver Mulde and drinking water\nsamples (after subtraction of\nlevels present in the non-spiked\nsamples). For calculation of the\ncorrection factors, the spiked river\nGötsche water samples were\nused. Accuracy (precision and trueness) [10] for a method based on mixed-\nmode SPE enrichment and mixed-mode liquid chromatography\ncoupled to tandem MS. Trueness, expressed as percentage devi-\nation of the quantified concentration (applying the correction\nfactor from ESM Table S7) from the theoretical (spiked) concen-\ntration, was in the range −25 to + 64% (average ± standard\ndeviation 8.8 ± 22.6%) for river Mulde and −58 to + 73% (av-\nerage ± standard deviation −3.2 ± 39.4%) for drinking water. The lower accuracy (lower precision and lower trueness) for\nthe drinking water compared to river Mulde samples is most\nlikely due to the 10 times lower concentrations spiked to drinking\nwater compared to river Mulde (see the section “Accuracy (pre-\ncision and trueness)” in the “Experimental section”). The excel-\nlent average values for trueness of + 8.8 and −3.2% show that\nthere is no systematic bias in the quantification method. Trueness\nreported for the method by Montes et al. [10] for individual PM\nsubstances was slightly better than that in our study. However,\nMontes and co-workers [9, 10] used a different set of test com-\npounds and a sample-specific “standard addition over the extracts\nmethodology” for quantification, which requires an individual\ncalibration curve for each sample. Our method does not rely on\nsample-specific standard addition and is thus applicable to larger\nscreening or monitoring studies. References 1. Reemtsma T, Berger U, Arp HPH, Gallard H, Knepper TP,\nNeumann M, et al. Mind the gap: persistent and mobile organic\ncompounds – water contaminants that slip through. Environ Sci\nTechnol. 2016;50:10308–15. Fig. 5 Concentration data of PM substances in environmental and\ndrinking water samples. Estimated concentrations between the MDL\nand the MQL are included in parentheses. Abbreviations: B, Berlin; H,\nHessia; SW, surface water; RW, raw water; DW, drinking water; GW,\ngroundwater. For details on samples, see Table S3. Note the logarithmic\nconcentration scale 2. Arp HPH, Brown TN, Berger U, Hale SE. Ranking REACH reg-\nistered neutral, ionizable and ionic organic chemicals based on their\naquatic persistency and mobility. Environ Sci: Processes Impacts. 2017;19:939–55. 3. Schulze S, Sättler D, Neumann M, Arp HPH, Reemtsma T, Berger\nU. Using REACH registration data to rank the environmental emis-\nsion potential of persistent and mobile organic chemicals. Sci Total\nEnviron. 2018;625:1122–8. analysis, and external quantification using a correction factor\nfor apparent recoveries. Trueness of quantification revealed\nresults within ± 30% of the theoretical value in 77% of quan-\ntifications. To further increase accuracy, we recommend syn-\nthesis of isotope-labeled internal standards for the most im-\nportant PM substances. This is the first method specifically\ndesigned for PM substances that does not include sample-\nspecific calibration curves (standard addition quantification). The method is thus well suitable for large screening and mon-\nitoring programs. The method is generic and can easily be\nexpanded to include further target PM substances or used in\nnon-target or suspect screening of highly polar contaminants\nin water samples. 4. Neumann M, Schliebner I. Protecting the sources of our drinking\nwater: the criteria for identifying persistent, mobile and toxic (PMT)\nsubstances and very persistent and very mobile (vPvM) substances\nunder EU Regulation REACH (EC) No 1907/2006. German\nEnvironment Agency (UBA Texte 127/2019). Dessau-Rosslau\n(87 pages, ISSN 1862-4804); 2019. 5. Zahn D, Neuwald IJ, Knepper TP. Analysis of mobile chemicals in\nthe aquatic environment – current capabilities, limitations and fu-\nture perspectives. Anal Bioanal Chem. 2020. https://doi.org/10. 1007/s00216-020-02520-z. 6. Bruzzoniti MC, de Carlo RM, Sarzanini C. Determination of sul-\nfonic acids and alkylsulfates by ion chromatography in water. Talanta. 2008;75:734–9. 7. Zahn D, Fromel T, Knepper TP. Halogenated methanesulfonic\nacids: a new class of organic micropollutants in the water cycle. Water Res. 2016;101:292–9. Conflict of interest\nThe authors declare that they have no conflicts of\ninterest. Open Access This article is licensed under a Creative Commons\nAttribution 4.0 International License, which permits use, sharing,\nadaptation, distribution and reproduction in any medium or format, as\nlong as you give appropriate credit to the original author(s) and the\nsource, provide a link to the Creative Commons licence, and indicate if\nchanges were made. The images or other third party material in this article\nare included in the article's Creative Commons licence, unless indicated\notherwise in a credit line to the material. If material is not included in the\narticle's Creative Commons licence and your intended use is not\npermitted by statutory regulation or exceeds the permitted use, you will\nneed to obtain permission directly from the copyright holder. To view a\ncopy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Method application to environmental water samples A rapid trace analytical method for the simultaneous quantifi-\ncation of 15 target PM substances (log D values −3.06 to\n1.23) was developed for water samples. The method is based\non azeotrope evaporation of the samples, SFC-HRMS The method was applied to six samples relevant to drinking\nwater production from Berlin and Hessia, Germany (surface A rapid method for quantification of persistent and mobile organic substances in water using supercritical... 4951 Fig. 5 Concentration data of PM substances in environmental and\ndrinking water samples. Estimated concentrations between the MDL\nand the MQL are included in parentheses. Abbreviations: B, Berlin; H,\nHessia; SW, surface water; RW, raw water; DW, drinking water; GW,\ngroundwater. For details on samples, see Table S3. Note the logarithmic\nconcentration scale Compliance with ethical standards Conflict of interest\nThe authors declare that they have no conflicts of\ninterest. Conflict of interest\nThe authors declare that they have no conflicts of\ninterest. References Simultaneous determination of PPCPs,\nEDCs, and artificial sweeteners in environmental water samples\nusing a single-step SPE coupled with HPLC-MS/MS and isotope\ndilution. Talanta. 2013;113:82–92. 15. Taylor LT. Supercritical fluid chromatography for the 21st century. J Supercrit Fluids. 2009;47:566–73. 28. Svan A, Hedeland M, Arvidsson T, Pettersson CE. The differences\nin matrix effect between supercritical fluid chromatography and\nreversed phase liquid chromatography coupled to ESI/MS. Anal\nChim Acta. 2018;1000:163–71. 16. Lesellier E. Retention mechanisms in super/subcritical fluid chro-\nmatography on packed columns. J Chromatogr A. 2009;1216:\n1881–90. 17. West C. How good is SFC for polar analytes? Chromatogr Today. 2013:22–9. 29. Birch GF, Drage DS, Thompson K, Eaglesham G, Mueller JF. Emerging contaminants (pharmaceuticals, personal care products,\na food additive and pesticides) in waters of Sydney estuary\nAustralia. Mar Pollut Bull. 2015;97:56–66. 18. Bieber S, Greco G, Grosse S, Letzel T. RPLC-HILIC and SFC with\nmass spectrometry: polarity-extended organic molecule screening\nin environmental (water) samples. Anal Chem. 2017;89:7907–14. 30. Perkola N, Sainio P. Quantification of four artificial sweeteners in\nFinnish surface waters with isotope-dilution mass spectrometry. Environ Pollut. 2014;184:391–6. 19. Salvatierra-Stamp V, Ceballos-Magaña SG, Gonzalez J, Ibarra-\nGalván V, Muñiz-Valencia R. Analytical method development for\nthe determination of emerging contaminants in water using\nsupercritical-fluid chromatography coupled with diode-array detec-\ntion. Anal Bioanal Chem. 2015;407:4219–26. 31. Buerge IJ, Buser HR, Kahle M, Müller MD, Poiger T. Ubiquitous\noccurrence of the artificial sweetener acesulfame in the aquatic\nenvironment. Environ Sci Technol. 2009;43:4381–5. 20. Sen A, Knappy C, Lewis MR, Plumb RS, Wilson ID, Nicholson\nJK, et al. Analysis of polar urinary metabolites for metabolic phe-\nnotyping using supercritical fluid chromatography and mass spec-\ntrometry. J Chromatogr A. 2016;1449:141–55. 32. Ruff M, Mueller MS, Loos M, Singer HP. Quantitative target and\nsystematic non-target analysis of polar organic micro-pollutants\nalong the river Rhine using high-resolution mass-spectrometry –\nidentification of unknown sources and compounds. Water Res. 2015;87:145–54. 21. Parr MK, Wuest B, Naegele E, Joseph JF, Wenzel M, Schmidt AH,\net al. SFC-MS/MS as an orthogonal technique for improved screen-\ning of polar analytes in anti-doping control. Anal Bioanal Chem. 2016;408:6789–97. 33. Betowski L, Kendall D, Donnelly C. Characterization of ground-\nwater samples from superfund sites by gas chromatography/mass\nspectrometry and liquid chromatography/mass spectrometry. Environ Sci Technol. 1996;30:3558–64. 22. Desfontaine V, Capetti F, Nicoli R, Kuuranne T, Veuthey JL,\nGuillarme D. References Acknowledgments The authors are thankful for the support in sample\nacquisition by Peter Seel (Hessian Agency for Nature Conservation,\nEnvironment and Geology, Wiesbaden, Germany) and Berliner\nWasserbetriebe (Berlin, Germany). Finally, we would like to thank\nCoretta Bauer (Helmholtz Centre for Environmental Research - UFZ,\nLeipzig, Germany) for the technical help in the lab. 8. Boulard L, Dierkes G, Ternes T. Utilization of large volume zwit-\nterionic hydrophilic interaction liquid chromatography for the anal-\nysis of polar pharmaceuticals in aqueous environmental samples:\nbenefits and limitations. J Chromatogr A. 2018;1535:27–43. 9. Montes R, Aguirre J, Vidal X, Rodil R, Cela R, Quintana JB. Screening for polar chemicals in water by trifunctional mixed-\nmode liquid chromatography-high resolution mass spectrometry. Environ Sci Technol. 2017;51:6250–9. Funding information Open Access funding provided by Projekt DEAL. This work has been funded by the German BMBF in the frame of the\ncollaborative international consortium WATERJPI2013 – PROMOTE of\nthe Water Challenges for a Changing World Joint Programming Initiative\n(Water JPI) Pilot Call (02WU1347A) and through the project PROTECT\n(02WRS1495A). 10. Montes R, Rodil R, Cela R, Quintana JB. Determination of persis-\ntent and mobile organic contaminants (PMOCs) in water by mixed-\nmode liquid chromatography–tandem mass spectrometry. Anal\nChem. 2019;91:5176–83. 4952 Schulze S. et al. 11. Schulze S, Zahn D, Montes R, Rodil R, Quintana JB, Knepper TP,\net al. Occurrence of emerging persistent and mobile organic con-\ntaminants in European water samples. Water Res. 2019;153:80–90. 24. Köke N, Zahn D, Knepper TP, Frömel T. Multi-layer solid-phase\nextraction and evaporation-enrichment methods for polar organic\nchemicals from aqueous matrices. Anal Bioanal Chem. 2018;410:\n2403–11. 12. Laboureur L, Guérineau V, Auxilien S, Yoshizawa S, Touboul D. Profiling of modified nucleosides from ribonucleic acid digestion\nby supercritical fluid chromatography coupled to high resolution\nmass spectrometry. J Chromatogr A. 2018;1537:118–27. 25. Khalikova MA, Lesellier E, Chapuzet E, Šatínský D, West C. Development and validation of ultra-high performance supercritical\nfluid chromatography method for quantitative determination of nine\nsunscreens in cosmetic samples. Anal Chim Acta. 2018;1034:184–94. 13. Klesper E, Corwin AH, Turner DA. High pressure gas chromatog-\nraphy above critical temperatures. J Organomet Chem. 1962;27:\n700–1. 26. Abrahamsson V, Sandahl M. Impact of injection solvents on super-\ncritical fluid chromatography. J Chromatogr A. 2013;1306:80–8. 14. Saito M. History of supercritical fluid chromatography: instrumen-\ntal development. J Biosci Bioeng. 2013;115:590–9. 27. Tran NH, Hu J, Ong SL. References Systematic evaluation of matrix effects in supercriti-\ncal fluid chromatography versus liquid chromatography coupled to\nmass spectrometry for biological samples. J Chromatogr B. 2018;1079:51–61. 34. Tang J, Tang L, Zhang C, Zeng G, Deng Y, Dong H, et al. Different\nsenescent HDPE pipe-risk: brief field investigation from source\nwater to tap water in China (Changsha City). Environ Sci Pollut\nRes Int. 2015;22:16210–4. 23. Reemtsma T, Alder L, Banasiak U. A multimethod for the deter-\nmination of 150 pesticide metabolites in surface water and ground-\nwater using direct injection liquid chromatography-mass spectrom-\netry. J Chromatogr A. 2013;1271:95–104. Publisher’s note Springer Nature remains neutral with regard to jurisdic-\ntional claims in published maps and institutional affiliations."