Method for comprehensive nontarget screening of per- and polyfluoroalkyl substances (PFAS)

Abstract

A method for comprehensive nontarget screening of per- and polyfluoroalkyl substances (PFAS) is provided. The method includes the following steps: subjecting a sample to be tested to pretreatment, chr

Description

[0001] The present disclosure relates to the technical field of environmental analytical chemistry, and in particular to a method for comprehensive nontarget screening of per- and polyfluoroalkyl substances (PFAS). BACKGROUND

[0002] Per- and polyfluoroalkyl substances (PFAS), due to their superior physical and chemical properties, have been widely used in various industrial and commercial products including surfactants, surface protectants, antifouling textiles, food packaging, aqueous film-forming foams, and cosmetics. Due to their unique physical and chemical properties, the legacy PFAS can persist in the environment indefinitely and then bio-accumulate in vivo through food chain, posing potential risks to health.

[0003] With increasing regulatory restrictions on legacy PFAS in numerous countries, novel PFAS are coming into use progressively. Conventional targeted screening methods can only monitor less than 100 PFAS with standards, leaving thousands of novel PFAS requiring nontarget screening approaches. However, current research methodologies predominantly rely on limited software-based nontarget analysis, with existing protocols remaining insufficiently comprehensive and largely confined to suspect screening strategies. SUMMARY

[0004] An objective of the present disclosure is to provide a method for comprehensive nontarget screening of PFAS. The method can be used to screen for PFAS more comprehensively, and minimizing omissions.

[0005] To achieve the above objective, the present disclosure provides the following technical solutions:

[0006] The present disclosure provides a method for comprehensive nontarget screening of PFAS, including the following steps:

[0007] subjecting a sample to be tested to pretreatment, chromatographic separation, and mass spectrometry detection in sequence to obtain mass spectrum data;

[0008] converting the mass spectrum data into a file in mzXML format, conducting peak extraction and tandem mass spectrometry (MS / MS) spectra extraction, and then conducting isotope detection and adduct detection;

[0009] constructing a local database according to data obtained from the peak extraction and the MS / MS spectra extraction;

[0010] providing a suspect list, and then conducting suspect screening in the suspect list according to a structured query language (SQL) script and the local database;

[0011] conducting fragment-based nontarget screening according to the SQL script and diagnostic fragments and neutral losses of the PFAS;

[0012] conducting homolog-based nontarget screening according to homologous series of the PFAS; and

[0013] conducting structural identification and true positive testing based on screening results obtained from the suspect screening, the fragment-based nontarget screening, and the homolog-based nontarget screening as well as detection results obtained from the isotope detection and the adduct detection, thereby achieving the comprehensive nontarget screening of the PFAS in the sample to be tested.

[0014] Preferably, the mass spectrum data are converted into the file in mzXML format using an MSConvert module in ProteoWizard; parameters include: Output format of mzXML, Filters of Peak Picking, and Algorithm of Vendor; peakPicking is added in a workflow area, and other filtering means are removed.

[0015] Preferably, the peak extraction is conducted using a centWave algorithm with XCMS package of R; parameters of the centWave algorithm include: peakwidth=c(5,20), noise=5000, ppm=5, and mzdiff=0.003; and a peak table in csv format is obtained after the peak extraction is completed; and

[0016] the MS / MS spectra extraction is conducted using a chromPeakSpectra algorithm with the XCMS package of R; parameters of the chromPeakSpectra algorithm include: expandRt=15 and expandMz=0.001; and a MS / MS spectra table in csv format is obtained after the MS / MS spectra extraction is completed.

[0017] Preferably, the constructing the local database includes: converting the peak table and the MS / MS spectra table into the local database through MySQL, where a structure of the local database is a relational database model.

[0018] Preferably, the isotope detection and the adduct detection include: detecting isotopes and adducts generated by a same compound with CAMERA package of R using default parameters.

[0019] Preferably, the suspect screening includes: in the SQL script, searching for an expected [M-H]' precursor of each compound using a Query_msl statement according to the suspect list, with a mass tolerance set to 1.5 mDa and 5 ppm; and searching for results having MS / MS spectra records within a mass tolerance of 1.5 mDa using a mass-to-charge ratio as an input in a Query_ms2 statement.

[0020] Preferably, the fragment-based nontarget screening includes: conducting searching using a Query fragment statement in the SQL script, with a mass tolerance set to 1.5 mDa, where the diagnostic fragments include CnF2n+f, CnF2n-f, CnF2n+i0‘, CnF2n-i0‘, SO2F', and SO3F, and the neutral losses include HF and CHFO2.

[0021] Preferably, the homolog-based nontarget screening includes: conducting filtering with a mass defect greater than 0.85 or a mass defect less than 0.15, calculating the homologous series of the PFAS through a Python script, and then screening a homologous PFAS.

[0022] Preferably, the structural identification includes: conducting literature or database matching identification on reported PFAS; determining an elemental composition of potential PFAS precursors and fragments on unreported positive substances using an online calculation tool ChemCalc, where parameters include: C<50, H<50, F<50, O<10, S<5, Cl<3, Br<3, P<3, and N<10; and a mass tolerance is set to 1.5 mDa.

[0023] Preferably, software-assisted verification is further conducted after the true positive testing is completed; and software for the software-assisted verification is one or more selected from the group consisting of Compound Discoverer, MS-DIAL, and CAMERA package of R.

[0024] Beneficial Effects: the present disclosure provides a method for comprehensive nontarget screening of PFAS. Specifically, the method can screen PFAS more comprehensively by combining three screening approaches: suspect screening, fragment-based nontarget screening, and homolog-based nontarget screening, thereby minimizing omissions. Meanwhile, the method can be used for rapid retrieval with the help of local database and SQL script, which can improve the screening speed and analysis flexibility, and has significant advantages in nontarget screening of large-scale data.

[0025] Further, fully integrated nontarget screening software in the prior art not only results in inaccurate screening results, but also prevents timely and targeted adjustments to the parameters involved in the screening. The parameters involved in the screening using the method of the present disclosure can be adjusted at any time.

[0026] Furthermore, based on the three screening approaches of suspect screening, fragment-based nontarget screening, and homolog-based nontarget screening, the method of the present disclosure also uses commercial software to allow auxiliary identification, which is beneficial to reduce the omission of screening results and improve the accuracy of screening results. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] FIG. 1 shows a flow chart of the method for comprehensive nontarget screening of PFAS in the present disclosure;

[0028] FIG. 2 shows a structural diagram of the local database in the present disclosure;

[0029] FIG. 3 shows a structural elucidation of H-O-U-PFCA;

[0030] FIG. 4 shows a chromatogram of H-O-U-PFCA;

[0031] FIG. 5 shows a structural elucidation of H-PFCAs;

[0032] FIG. 6 shows a chromatogram of H-PFCAs;

[0033] FIG. 7 shows a structural elucidation of F5S-PFSA;

[0034] FIG. 8 shows a chromatogram of F5S-PFSA;

[0035] FIG. 9 shows a structural elucidation of O-UPFSiA;

[0036] FIG. 10 shows a chromatogram of O-UPFSiA;

[0037] FIG. 11 shows a structural elucidation of 5:2 FTCA;

[0038] FIG. 12 shows a chromatogram of 5:2 FTCA;

[0039] FIG. 13 shows a structural elucidation of 2H-PFMEOH;

[0040] FIG. 14 shows a chromatogram of 2H-PFMEOH;

[0041] FIG. 15 shows a structural elucidation of 3:2 FTOH;

[0042] FIG. 16 shows a chromatogram of 3:2 FTOH;

[0043] FIG. 17 shows a structural elucidation of H-PFMEOH; and

[0044] FIG. 18 shows a chromatogram of H-PFMEOH. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] The present disclosure provides a method for comprehensive nontarget screening of PFAS, including the following steps:

[0046] subjecting a sample to be tested to pretreatment, chromatographic separation, and mass spectrometry detection in sequence to obtain mass spectrum data;

[0047] converting the mass spectrum data into a file in mzXML format, conducting peak extraction and MS / MS spectra extraction, and then conducting isotope detection and adduct detection;

[0048] constructing a local database according to data obtained from the peak extraction and the MS / MS spectra extraction;

[0049] providing a suspect list, and then conducting suspect screening in the suspect list according to a structured query language (SQL) script and the local database;

[0050] conducting fragment-based nontarget screening according to the SQL script and diagnostic fragments and neutral losses of the PFAS;

[0051] conducting homolog-based nontarget screening according to homologous series of the PFAS; and

[0052] conducting structural identification and true positive testing based on screening results obtained from the suspect screening, the fragment-based nontarget screening, and the homolog-based nontarget screening as well as detection results obtained from the isotope detection and the adduct detection, thereby achieving the comprehensive nontarget screening of the PFAS in the sample to be tested.

[0053] FIG. 1 shows a flow chart of the method for comprehensive nontarget screening of PFAS in the present disclosure, and the method in the present disclosure is described in detail below.

[0054] In the present disclosure, a sample to be tested is subjected to pretreatment, chromatographic separation, and mass spectrometry detection in sequence to obtain mass spectrum data. There is no particular limitation on a type of the sample to be tested, which may specifically be a water sample (such as a groundwater sample) or a serum sample. Preferably, corresponding pretreatment method and chromatographic separation method are selected according to a specific type of the sample to be tested. In an example, the sample to be tested as the water sample or the serum sample is taken as an example for description.

[0055] In the present disclosure, when the sample to be tested is the water sample, the pretreatment preferably includes the following steps: filtering the water sample, adjusting a pH value of an obtained filtrate to reach acidity and then adding a mixed internal standard solution to obtain a water sample containing an internal standard; and subjecting the water sample containing the internal standard to solid-phase extraction to obtain a solution to be tested.

[0056] In the present disclosure, a filter membrane used for the filtration is preferably a GF / F glass fiber filter membrane, and the filter membranes have pore sizes of preferably 1.2 pm and 0.7 pm. Specifically, the water sample is filtered in a vacuum filtration device using the GF / F glass fiber filter membranes with pore sizes of 1.2 pm and 0.7 pm in sequence. Preferably, the pH value of the filtrate is adjusted with glacial acetic acid, and the acidity specifically refers to a pH value of 4. The mixed internal standard solution has a concentration of preferably 1 mg / L; the mixed internal standard solution and the filtrate with a pH value of 4 are at a volume ratio of preferably 2 L to 50 pL; an internal standard in the mixed internal standard solution is preferably a PFAS standard with a C isotope (specifically including 13Ce-perfluorohexanoic acid, 13C8-perfluorooctanoic acid, 13C8-perfluorooctane sulfonic acid, 13C9-perfluorononanoic acid, 13C9-perfluorodecanoic acid, 13C2-8:2 fluorotelomer carboxylic acid, 13C2-8:2 fluorotelomer unsaturated acid, 13C2-6:2 fluorotelomer sulfonic acid, 13C2-8:2 polyfluorooctane monophosphate, 13C2-6:2 polyfluorooctane phosphodiester, ds-N-methyl-perfluorooctane sulfonamide, and dv-N-methyl-perfluorooctane sulfonamidoethanol), playing a role in quality control.

[0057] In the present disclosure, a solid-phase extraction column for the solid-phase extraction is preferably an Oasis HLB cartridge. Preferably, the cartridge is installed on a solid-phase extraction device; and after activation, the water sample containing the internal standard is passed through the cartridge, and then impurity removal and elution are conducted in sequence; an eluate is collected, blown to near dryness with nitrogen, and then reconstituted with methanol, filtered, and a resulting filtrate is collected as the solution to be tested (i.e., a PFAS-containing concentrate). Impurity removal reagents for the activation are preferably a methanol solution containing 1 wt% of ammonium hydroxide, methanol, and water in sequence, and the water is preferably ultrapure water. The water sample containing the internal standard is preferably passed through the cartridge under negative pressure, at a flow rate of preferably 15 mL / min. The impurity removal is preferably conducted under negative pressure, and the impurity removal reagent used is preferably water, more preferably ultrapure water, and a volume ratio of the water sample containing the internal standard to the impurity removal reagent is preferably 2 L to 10 mL; after the impurity removal reagent is eluted, it is maintained at negative pressure for 30 min to drain the cartridge until the filler is sandy, followed by conducting the elution. An eluent used for the elution is preferably a methanol solution containing 1 wt% ammonium hydroxide, and a volume ratio of the water sample containing the internal standard to the eluent is preferably 2 L: 10 mL; the eluent preferably flows naturally through the cartridge to allow the elution. The nitrogen blowing is conducted at preferably 35°C; methanol for reconstitution is at preferably 1 mL, and the reconstitution is preferably conducted under vortexing, and the vortexing is conducted for preferably 20 s to 30 s, more preferably 25 s; after the reconstitution is completed, filtration is preferably conducted with a 0.22 pm nylon filter (organic phase compatible), and a filtrate is collected as the solution to be tested, and the solution to be tested is transferred to a PP material chromatographic vial, sealed with a sealing film and then stored at -20°C.

[0058] In the present disclosure, when the sample to be tested is the water sample, the solution to be tested is preferably analyzed by ultrahigh-performance liquid chromatography (UPLC)-high resolution mass spectrometer (HRMS) (i.e., the chromatographic separation and mass spectrometry detection) to obtain the mass spectrum data. The UPLC preferably includes the following parameters: UPLC instrument: Dionex UltiMate 3000, Thermo Fisher; chromatographic column: ZORBAX Extend plus Cl8 chromatographic column (RRHD 2.1 mmxlOO mm, 1.8 pm), Agilent; guard column: ZORBAX Extend C18 guard column (RRHD 2.1 mm><5 mm, 1.8 pm), Agilent; column temperature 40°C; mobile phase flow rate 0.25 mL / min; injection volume 3 pL; mobile phase includes mobile phase A and mobile phase B, where mobile phase A is 2 mM ammonium acetate aqueous solution (adjusting pH to 10.25 with ammonia water), and mobile phase B is a mixture of A-methylpiperidine (1-MP), acetonitrile, and methanol (A-methylpiperidine concentration 5 mM, and the volume ratio of acetonitrile to methanol at 1:1); a gradient elution program is adopted, as shown in Table 1 in Example 1, which will not be repeated here. The HRMS preferably includes the following parameters: HRMS instrument using Q-Exactive Focus, Thermo Fisher; ion source type using ESI (-); ionization mode using negative mode; electrospray voltage at -4,000 V; sheath gas flow rate at 10 a.u; capillary temperature at 320°C; nebulizer temperature at 350°C; S lens RF level at 50; full-scan mass range at (150-1,100) Da; full scan resolution at 70,000; MS / MS spectra resolution at 17,500; fragment normalized collision energy at (35±15) eV; dynamic exclusion time at 50 s; vertex excitation range at 30%.

[0059] In the present disclosure, when the sample to be tested is the serum sample, the pretreatment and chromatographic separation are preferably achieved by a UPLC online solid-phase extraction chromatography coupling system, and mass spectrometry detection is achieved by coupling with HRMS to obtain the mass spectrum data. That is, the serum sample is directly processed and detected by a UPLC online solid-phase extraction chromatography coupling system-HRMS, specifically including sample injection, online solid-phase extraction, elution, chromatographic signal detection, and mass spectrum signal detection. The UPLC online solid-phase extraction chromatography coupling system preferably includes the following parameters: UPLC online solid-phase extraction chromatography coupling system using CHRONECT Symbiosis Advance, Axel Semrau; chromatographic column using ZORBAX Eclipse Plus Cl8 chromatographic column (2.1x50 mm, 1.8 pm), Aglient; column temperature at 50°C; mobile phase flow rate at 0.25 mL / min; injection volume at 180 pL. The mobile phase includes mobile phase A and mobile phase B, where mobile phase A is 5 mM ammonium acetate aqueous solution, and mobile phase B is methanol, and a gradient elution program is adopted, as shown in Table 6 in Example 2, which will not be repeated here. The parameters of the HRMS instrument are preferably consistent with those of the HRMS instrument when the sample to be tested is the water sample, and will not be described in detail here.

[0060] In the present disclosure, the mass spectrum data are converted into a file in mzXML format, peak extraction and MS / MS spectra extraction are conducted, and then isotope detection and adduct detection are conducted. There is no particular limitation on an order of the peak extraction and the MS / MS spectra extraction, and the peak extraction and the MS / MS spectra extraction are preferably conducted in sequence. The isotope detection and the adduct detection are preferably conducted simultaneously. Each step is described in detail below.

[0061] In the present disclosure, initial mass spectrum data are in a RAW format file, which is converted into a file in mzXML format for subsequent related processing. The mass spectrum data are preferably converted into the file in mzXML format using an MSConvert module in ProteoWizard; parameters include: Output format of mzXML, Filters of Peak Picking, and Algorithm of Vendor; peakPicking is added in a workflow area, and other filtering means are removed. The ProteoWizard is in a specific version of 3.0.19039.

[0062] The peak extraction is preferably conducted using a centWave algorithm with XCMS package of R; parameters of the centWave algorithm include: peakwidth=c(5,20), noise=5000, ppm=5, and mzdiff=0.003; and a peak table in csv format is obtained after the peak extraction is completed. The XCMS package of R is in a specific version of 3.4.2.

[0063] In the present disclosure, the MS / MS spectra extraction is preferably conducted using a chromPeakSpectra algorithm and an R script with XCMS package of R; parameters of the chromPeakSpectra algorithm include: expandRt=15 and expandMz=0.001; and a MS / MS spectra table in csv format is obtained after the MS / MS spectra extraction is completed. The XCMS package of R is in a specific version of 3.4.2; the chromPeakSpectra algorithm is in a specific version of 3.5.2.

[0064] The isotope detection and the adduct detection preferably include: detecting isotopes and adducts generated by a same compound with CAMERA package of R with default parameters. The CAMERA package of R is in a specific version of 1.38.1.

[0065] In the present disclosure, a local database is constructed according to data obtained from the peak extraction and the MS / MS spectra extraction. The constructing the local database preferably includes: converting the peak table and the MS / MS spectra table into the local database through MySQL, where a structure of the local database is a relational database model (i.e., storing data and relationships between data in a tabular form). The MySQL is in a specific version of Community Server 8.0.11. FIG. 2 shows a structural diagram of the local database in the present disclosure. A data structure for subsequent data calling and screening is constructed in the SQL, through which data can be called flexibly and accurately for screening. When the MS / MS spectra table is converted and imported into the local database, the MS / MS spectra table is specifically split into a MS / MS spectra maximum response table, a MS / MS spectra basic information table, and a fragment table and then imported into the local database (the above three tables are connected by a same column to facilitate subsequent screening). Specifically in FIG. 2, the sample information table includes sample serial number (SN) and sample information (such as location); the peak table includes the peak SN, mass-to-charge ratio, retention time, response intensity, and sample SN; the MS / MS spectra maximum response table includes MS / MS spectra SN and maximum response value; the MS / MS spectra basic information table includes MS / MS spectra SN, parent ion mass-to-charge ratio, sample SN, retention time, and parent ion response intensity; the fragment table includes response ID, MS / MS spectra SN, ion mass-to-charge ratio, and response intensity.

[0066] In the present disclosure, a suspect list is provided, and then suspect screening is conducted in the suspect list according to an SQL script and the local database. Specifically, the suspect list is formulated according to a PFAS database, and the PFAS database used to formulate the suspect list preferably includes one or more of: AB SCIEX (including 95 PFAS), Norman Suspect List Exchange (including 1,030 PFAS), OECD PFAS list (including 3,706 PFAS), PFAS Master List (including 9,525 PFAS), U.S. Environmental Protection Agency CompTox database, AB SCIEX's commercial PFAS database with 95 MS / MS spectra, Norman Suspect List Exchange's 1,030 PFAS, OECD PFAS list's 3,706 PFAS, and PFAS Master List (including 9,525 PFAS). The suspect screening preferably includes: in the SQL script, searching for an expected [M-H]' precursor of each compound using a Query ms 1 statement according to the suspect list, with a mass tolerance set to 1.5 mDa and 5 ppm; and searching for results having MS / MS spectra records within a mass tolerance of 1.5 mDa using a mass-to-charge ratio as an input in a Query_ms2 statement. Specifically, when parameters of a compound in the suspect list match those of a substance in the database within an allowable error range, the compound is determined to be a suspect.

[0067] In the present disclosure, fragment-based nontarget screening is conducted according to the SQL script and diagnostic fragments and neutral losses of the PFAS. The fragment-based nontarget screening preferably includes: conducting searching using a Query fragment statement in the SQL script, with a mass tolerance set to 1.5 mDa, where the diagnostic fragments include CnF2n+f, CnF2n-f, CnF2n+iO", CnF2n-iO", SO2F', and SO3F, and the neutral losses include HF and CHFO2. Specifically, when a compound has at least one of the diagnostic fragments, the compound is determined to be a candidate PFAS.

[0068] In the present disclosure, homolog-based nontarget screening is conducted according to homologous series of the PFAS. The homolog-based nontarget screening preferably includes: conducting filtering with a mass defect greater than 0.85 or a mass defect less than 0.15, calculating the homologous series of the PFAS through a Python script, and then screening a homologous PFAS.

[0069] In the present disclosure, there is no particular limitation on an order of the suspect screening, the fragment-based nontarget screening, and the homolog-based nontarget screening, and the suspect screening, the fragment-based nontarget screening, and the homolog-based nontarget screening are preferably conducted in sequence.

[0070] In the present disclosure, structural identification and true positive testing are conducted based on screening results obtained from the suspect screening, the fragment-based nontarget screening, and the homolog-based nontarget screening as well as detection results obtained from the isotope detection and the adduct detection, thereby achieving the comprehensive nontarget screening of the PFAS in the sample to be tested. The screening results specifically include data information of a candidate PFAS to be confirmed, including MSI, MS / MS spectra, and chromatographic information.

[0071] In the present disclosure, the structural identification preferably includes: conducting literature or database matching identification on reported PFAS; determining an elemental composition of potential PFAS precursors and fragments on unreported positive substances using an online calculation tool ChemCalc, where parameters include: C<50, H<50, F<50, O<10, S<5, Cl<3, Br<3, P<3, and N<10; and a mass tolerance is set to 1.5 mDa.

[0072] In the present disclosure, the candidate PFAS is preferably subjected to true positive testing; the true positive testing preferably includes: determining whether the compound is an impurity based on detection frequency, chromatographic peak shape, and response, and then preliminarily determining whether in-source cleavage occurs based on whether a co-elution effect occurs and the MS / MS spectra to determine a final screening result.

[0073] In the present disclosure, software-assisted verification is preferably further conducted after the true positive testing is completed; and software for the software-assisted verification is preferably at least one selected from the group consisting of Compound Discoverer, MS-DIAL, and CAMERA package of R. The Compound Discoverer preferably includes the following parameters: (global) mass-to-charge ratio error: 5 ppm, and an atomic range for compound molecular formula calculation being: C<90, H<190, Br<3, Cl<8, F<50, I<2, N<10, O<18, P<3, and S<5. The MS-DIAL preferably includes the following parameters: ionization mode: soft ionization; separation mode: chromatography; mass spectrometry mode: legacy or data-dependent type; primary ion data type: centroid; ion mode: negative mode; in the Analysis parameter setting dialog box, the peak alignment file (Alignment), possible adduct (Adduct), and primary identification reference file (MSP file) are set. The CAMERA package of R preferably includes the following parameters: ppm=10, mzabs=0.01, intval="intb", and maxcharge=3. Preferably, original data (i.e., the mass spectrum data obtained after mass spectrometry detection) is imported into the above-mentioned software, and the software can conduct molecular structure matching according to the set parameters. The results of the candidate PFAS after true positive testing in the software are searched in the software, and then compared with the results identified by the method of the present disclosure to further improve an accuracy of the screening results.

[0074] The technical solutions of the present disclosure will be clearly and completely described below with reference to the examples of the present disclosure. Apparently, the described examples are merely a part rather than all of the examples of the present disclosure. All other examples obtained by those skilled in the art based on the examples of the present disclosure without creative efforts shall fall within the protection scope of the present disclosure.

[0075] In the following examples, internal standards used are PFAS standards with C isotope (there are 12 in total, specifically including 13C6-perfluorohexanoic acid, 13C8-perfluorooctanoic acid, 13C8-perfluorooctane sulfonic acid, 13C9-perfluorononanoic acid, 13C9-perfluorodecanoic acid, 13C2-8:2 fluorotelomer carboxylic acid, 13C2-8:2 fluorotelomer unsaturated acid, 13C2-6:2 fluorotelomer sulfonic acid, 13C2-8:2 polyfluorooctane monophosphate, 13C2-6:2 polyfluorooctane phosphodiester, ds-N-methyl-perfluorooctane sulfonamide, and dv-N-methyl-perfluorooctane sulfonamidoethanol), playing a role in quality control.

[0076] Example 1

[0077] (1) A water sample to be tested was subjected to pretreatment to obtain a test solution; the pretreatment included: the water sample to be tested was filtered in a vacuum filtration device using GF / F glass fiber filter membranes with pore sizes of 1.2 pm and 0.7 pm, respectively, and a pH value of the filtrate was adjusted to 4 using glacial acetic acid to obtain an acidic water sample. 50 pL of an internal standard mixed solution with a concentration of 1 mg / L was added into 2 L of the acidic water sample to obtain a water sample containing an internal standard. An Oasis HLB solid-phase cartridge was installed on a solid-phase extraction device, and the solid-phase cartridge was activated with a methanol solution containing 1 wt% ammonium hydroxide, methanol, and ultrapure water in sequence; under a negative pressure (provided by a vacuum filtration device), 2 L of the water sample containing the internal standard was passed through the solid-phase cartridge at 15 mL / min, and the solid-phase cartridge was rinsed with 10 mL ultrapure water to remove impurities, and then continued to be maintained under the negative pressure for 30 min to drain the solid-phase cartridge until the filler was sandy; 10 mL of methanol solution containing 1 wt% ammonium hydroxide was naturally passed through the solid-phase cartridge to allow elution, and all eluates were collected using a 15 mL PP centrifuge tube; the eluates were blown to near dryness with a nitrogen blower at 35°C, reconstituted with 1 mL of methanol and vortexed for 25 s, filtered with a 0.22 pm nylon filter (organic phase compatible), and a filtrate obtained was the solution to be tested (i.e., a PFAS-containing concentrate); the solution to be tested was transferred to a PP material chromatographic vial, sealed with a sealing film, and then stored at -20°C until analysis.

[0078] (2) The solution to be tested obtained in step (1) was detected by UPLC-HRMS instrument to obtain mass spectrum data; where the UPLC preferably included the following parameters: UPLC instrument using Dionex UltiMate 3000, Thermo Fisher; chromatographic column using ZORBAX Extend plus Cl8 chromatographic column (RRHD 2.1 mm* 100 mm, 1.8 pm), Agilent; guard column using ZORBAX Extend Cl8 guard column (RRHD 2.1 mmx5 mm, 1.8 pm), Agilent; column temperature at 40°C; mobile phase flow rate at 0.25 mL / min; injection volume at 3 pL. The mobile phase included mobile phase A and mobile phase B, where mobile phase A was 2 mM ammonium acetate aqueous solution (adjusting pH to 10.25 with ammonia water), and mobile phase B was a mixture of A-m ethylpiperidine (1-MP), acetonitrile, and methanol (A-methylpiperidine concentration 5 mM, and the volume ratio of acetonitrile to methanol at 1:1); a gradient elution program was adopted, as shown in Table 1:

[0079] Table 1 Mobile phase gradient elution program

[0080] Time (min) Volume of mobile phase A (%) Volume of mobile phase B (%) 0.00 90 10 2.00 90 10 3.00 65 45 8.00 5 95 15.00 5 95 15.20 90 10 19.00 90 10

[0081] The HRMS preferably included the following parameters: HRMS instrument using Q Exactive Focus-Orbitrap, Thermo Fisher; ion source type using ESI; ionization mode using negative mode; electrospray voltage at -4,000 V; sheath gas flow rate at 10 a.u; capillary temperature at 320°C; nebulizer temperature at 350°C; S lens RF level at 50; full-scan scan mass range at (150-1,100) Da; full scan resolution at 70,000; MS / MS spectra resolution at 17,500; fragment normalized collision energy at (35±15) eV; dynamic exclusion time at 50 s; vertex excitation range at 30%.

[0082] (3) The mass spectrum data obtained in step (2) were processed using ProteoWizard and written R script, including:

[0083] (31) The mass spectrum data (RAW format file) were preferably converted into the file in mzXML format using an MSConvert module in ProteoWizard (version 3.0.19039); parameters included: Output format of mzXML, Filters of Peak Picking, and Algorithm of Vendor; peakPicking was added in a workflow area, and other filtering means were removed.

[0084] (32) The peak extraction was conducted using a centWave algorithm (version 3.5.2) with XCMS package of R (version 3.4.2); parameters of the centWave algorithm included: peakwidth=c(5,20), noise=5000, ppm=5, mzdiff=0.003; and a peak table in csv format was obtained after the peak extraction was completed.

[0085] (33) The MS / MS spectra extraction was conducted using a chromPeakSpectra algorithm and an R script with XCMS package of R (version 3.4.2); parameters of the chromPeakSpectra algorithm included: expandRt=15 and expandMz=0.001; and a MS / MS spectra table in csv format was obtained after the MS / MS spectra data were converted by R script.

[0086] (34) The isotope detection and the adduct detection included: isotopes and adducts generated by a same compound were detected with CAMERA package of R (version 1.38.1) with default parameters.

[0087] (35) The constructing the local database included: the peak table and the MS / MS spectra table were converted into the local database through MySQL (Community Server 8.0.11), where a structure of the local database was a relational database model.

[0088] (4) Nontarget screening of PFAS was conducted based on the data processed in step (3), including:

[0089] (41) Preparation of a suspect list: the suspect list was made in accordance with the databases AB SCIEX (95 PFAS listed), Norman Suspect List Exchange (1,030 PFAS listed), OECD PFAS list (3,706 PFAS listed), and PFAS Master List (9,525PFAS listed).

[0090] (42) The suspect screening included: in the SQL script, an expected [M-H]' precursor of each compound was searched for using a Querymsl statement according to the suspect list, with a mass tolerance set to 1.5 mDa and 5 ppm; and results having MS / MS spectra records within a mass tolerance of 1.5 mDa were searched for using a mass-to-charge ratio as an input in a Query_ms2 statement.

[0091] (43) The fragment-based nontarget screening included: searching was conducted using a Query fragment statement in the SQL script, with a mass tolerance set to 1.5 mDa, where the diagnostic fragments included CnF2n+f, CnF2n-f, CnF2n+i0‘, CnF2n-iO‘, SO2F', and SOsF', and the neutral losses included HF and CHFO2.

[0092] (44) The homolog-based nontarget screening included: filtering was conducted with a mass defect>0.85 or a mass defect<0.15, the homologous series of the PFAS were calculated through a Python script, and then a homologous PFAS was screened.

[0093] (45) The structural identification included: literature or database matching identification was conducted on reported PFAS; an elemental composition of potential PFAS precursors and fragments on unreported positive substances was determined using an online calculation tool ChemCalc, where parameters included: C<50, H<50, F<50, O<10, S<5, Cl<3, Br<3, P<3, and N<10; and a mass tolerance was set to 1.5 mDa.

[0094] (46) The candidate PFAS was subjected to true positive testing, including: whether the compound was an impurity was determined based on detection frequency, chromatographic peak shape, and response, and then whether in-source cleavage occurred was preliminarily determined based on whether a co-elution effect occurred and the MS / MS spectra to determine a final screening result.

[0095] (47) Compound Discoverer and MS-DIAL were used for software-assisted verification: the Compound Discoverer included the following parameters: (global) mass-to-charge ratio error 5 ppm, and an atomic range for compound molecular formula calculation being: C<90, H<190, Br<3, Cl<8, F<50, I<2, N<10, O<18, P<3, and S<5; and MS-DIAL preferably included the following parameters: ionization mode: soft ionization; separation mode: chromatography; mass spectrometry mode: legacy or data-dependent type; primary ion data type: centroid; ion mode: negative mode; in the Analysis parameter setting dialog box, the peak alignment file (Alignment), possible adduct (Adduct), and primary identification reference file (MSP file) were set.

[0096] The method for comprehensive nontarget screening of PFAS in this example was used to screen more than 100 groundwater samples across the whole country. The results showed that 51 PFAS of 20 categories were identified in the more than 100 groundwater samples across the whole country. Taking H-O-U-PFCA (hydrogen-substituted unsaturated perfluoroether carboxylic acid, C8), H-PFCAs (trihydrogen-substituted perfluorocarboxylic acid), F5S-PFSA (pentafluoromercaptoethane sulfonic acid), and O-UPFSiA (unsaturated perfluoroalkylsulfinic acid) as examples, their mass spectrum analysis was as follows:

[0097] FIG. 3 showed a structural elucidation of H-O-U-PFCA; FIG. 4 showed a chromatogram of H-O-U-PFCA. Table 2 showed the diagnostic fragment information of H-O-U-PFCA.

[0098] Table 2 Diagnostic fragment information of H-O-U-PFCA

[0099] Ionic formula Exact mass[M-H] Theoretical mass[M-H] Error (ppm) Error (mDa) CF3 68.9958 68.9952 8.7 0.6 C4H2F3 107.0114 107.0109 4.7 0.5 C5F5 154.9927 154.9920 4.5 0.7 CsHFe 174.9989 174.9982 4.0 0.7 GHoUO 202.9940 202.9931 4.4 0.9 C7HF8O 252.9909 252.9900 3.6 0.9 C7H2F9O 272.9971 272.9962 3.3 0.9 C7H3F1oO 293.0041 293.0024 5.8 1.7 CsH3F wO3 336.9932 336.9923 2.7 0.9

[0100] FIG. 5 showed a structural elucidation of H-PFCAs; FIG. 6 showed a c H-PFCAs. Table 3 showed the diagnostic fragment information of H-PFCAs.

[0101] Table 3 Diagnostic fragment information of H-PFCAs iromatogram of

[0102] Ionic formula Exact mass[M-H] Theoretical mass[M-H] Error (ppm) Error (mDa) c3f3- 92.9959 92.9952 7.5 0.7 C7F9- 254.9862 254.9856 2.4 0.6 C7HFio- 274.9924 274.9918 2.2 0.6 C7H3F12O2 358.9951 358.9941 2.8 1.0

[0103] FIG. 7 showed a structural elucidation of F5S-PFSA; FIG. 8 showed a chromatogram of F5S-PFSA. Table 4 showed the diagnostic fragment information of F5S-PFSA.

[0104] Table 4 Diagnostic fragment information of F5S-PFSA

[0105] Ionic formula Exact mass[M-H] Theoretical mass[M-H] Error (ppm) Error (mDa) SO3‘ 79.9574 79.9568 7.5 0.6 SF5- 126.9646 126.9641 3.9 0.5 C3F6SO3- 229.9473 229.9472 0.4 0.1 c4f8so3- 279.9444 279.9440 1.4 0.4 C6Fi2SO3- 379.9378 379.9376 0.5 0.2 C6Fi7S2O3- 506.9023 506.9017 0.2 0.6

[0106] FIG. 9 showed a structural elucidation of O-UPFSiA; FIG. 10 showed a chromatogram of O-UPFSiA. Table 5 showed the diagnostic fragment information of O-UPFSiA.

[0107] Table 5 Diagnostic fragment information of O-UPFSiA

[0108] Ionic formula Exact mass[M-H] Theoretical mass[M-H] Error (ppm) Error (mDa) so3- 79.9573 79.9568 6.3 0.5 SO2F- 82.9609 82.9603 7.2 0.6 cf3o- 84.9906 84.9901 5.9 0.5 C2F3O' 96.9906 96.9901 5.2 0.5 so3f- 98.9557 98.9552 5.1 0.5 c3f5- 130.9924 130.992 3.1 0.4 C3F5O- 146.9875 146.9869 4.1 0.6 C3FsO2‘ 162.9824 162.9819 3.1 0.5 C4F7CU 212.9792 212.9787 2.3 0.5 C4F7SO4- 276.9408 276.9406 0.7 0.2

[0109] Example 2

[0110] A serum sample to be tested was treated and detected by UPLC online solid-phase extraction chromatography coupling system-HRMS instrument, including sample injection, online solid-phase extraction, elution, chromatographic signal detection, and mass spectrometry signal detection to obtain mass spectrum data; where the processing and detection included the following parameters: UPLC online solid-phase extraction chromatography coupling system using CHRONECT Symbiosis Advance, Axel Semrau; chromatographic column using ZORBAX Eclipse Plus Cl8 chromatographic column (2.1x50 mm, 1.8 pm), Aglient; column temperature at 50°C; mobile phase flow rate at 0.25 mL / min; injection volume at 180 pL. The mobile phase included mobile phase A and mobile phase B, where mobile phase A was 5 mM ammonium acetate aqueous solution, and mobile phase B was methanol, and a gradient elution program was adopted, as shown in Table 6.

[0111] Table 6 Mobile phase gradient elution program

[0112] Time (min) Volume of mobile phase A (%) Volume of mobile phase B (%) 0.00 65 35 2.00 90 10 3.00 90 10 8.00 55 45 15.00 5 95 15.20 5 95 19.00 90 10

[0113] The HRMS preferably included the following parameters: HRMS instrument using Q Exactive Focus-Orbitrap, Thermo Fisher; ion source type using ESI; ionization mode using negative mode; electrospray voltage at -4,000 V; sheath gas flow rate at 10 a.u; capillary temperature at 320°C; nebulizer temperature at 350°C; S lens RF level at 50; full-scan mass scan range at (150-1,100) Da; full scan resolution at 70,000; MS / MS spectra resolution at 17,500; fragment normalized collision energy at 35±15 eV; dynamic exclusion time at 50 s; vertex excitation range at 30%.

[0114] (2) The mass spectrum data obtained in step (1) were processed using ProteoWizard and written R script, including:

[0115] (21) The mass spectrum data (RAW format file) were preferably converted into the file in mzXML format using an MSConvert module in ProteoWizard (version 3.0.19039); parameters included: Output format of mzXML, Filters of Peak Picking, and Algorithm of Vendor; peakPicking was added in a workflow area, and other filtering means were removed.

[0116] (22) The peak extraction was conducted using a centWave algorithm (version 3.5.2) with XCMS package of R (version 3.4.2); parameters of the centWave algorithm included: peakwidth=c(5, 20), noise=5000, ppm=5, mzdiff=0.003; and a peak table in csv format was obtained after the peak extraction was completed.

[0117] (23) The MS / MS spectra extraction was conducted using a chromPeakSpectra algorithm and an R script with XCMS package of R (version 3.4.2); parameters of the chromPeakSpectra algorithm included: expandRt=15 and expandMz=0.001; and a MS / MS spectra table in csv format was obtained after the MS / MS spectra data were converted by R script.

[0118] (24) The isotope detection and the adduct detection included: isotopes and adducts generated by a same compound were detected with CAMERA package of R (version 1.38.1) with default parameters.

[0119] (25) The constructing the local database included: the peak table and the MS / MS spectra table were converted into the local database through MySQL (Community Server 8.0.11), where a structure of the local database was a relational database model.

[0120] (3) Nontarget screening of PFAS was conducted based on the data processed in step (2), including:

[0121] (31) Preparation of a suspect list: the suspect list was made in accordance with the databases US EPA CompTox database, the commercial PFAS database with 95 MS / MS spectra from AB SCIEX, the 1,030 PFAS from the Norman Suspect List Exchange, the 3,706 PFAS from the OECD PFAS list, and the PFAS Master List (9,525 PFAS included).

[0122] (32) The suspect screening included: in the SQL script, an expected [M-H]' precursor of each compound was searched for using a Querymsl statement according to the suspect list, with a mass tolerance set to 1.5 mDa and 5 ppm; and results having MS / MS spectra records within a mass tolerance of 1.5 mDa were searched for using a mass-to-charge ratio as an input in a Query_ms2 statement.

[0123] (33) The fragment-based nontarget screening included: searching was conducted using a Query fragment statement in the SQL script, with a mass tolerance set to 1.5 mDa, where the diagnostic fragments included CnF2n+f, CnF2n-f, CnF2n+iO", CnF2n-iO‘, SO2F', and SOsF', and the neutral losses included HF and CHFO2.

[0124] (34) The homolog-based nontarget screening included: filtering was conducted with a mass defect>0.85 or a mass defect<0.15, the homologous series of the PFAS were calculated through a Python script, and then a homologous PFAS was screened.

[0125] (35) The structural identification included: literature or database matching identification was conducted on reported PFAS; an elemental composition of potential PFAS precursors and fragments on unreported positive substances was determined using an online calculation tool ChemCalc, where parameters included: C<50, H<50, F<50, O<10, S<5, Cl<3, Br<3, P<3, and N<10; and a mass tolerance was set to 1.5 mDa.

[0126] (36) The candidate PFAS was subjected to true positive testing, including: whether the compound was an impurity was determined based on detection frequency, chromatographic peak shape, and response, and then whether in-source cleavage occurred was preliminarily determined based on whether a co-elution effect occurred and the MS / MS diagram to determine a final screening result.

[0127] (37) Compound Discoverer and MS-DIAL were used for software-assisted verification: the Compound Discoverer included the following parameters: (global) mass-to-charge ratio error 5 ppm, and an atomic range for compound molecular formula calculation being: C<90, H<190, Br<3, Cl<8, F<50, I<2, N<10, O<18, P<3, and S<5; and MS-DIAL preferably included the following parameters: ionization mode: soft ionization; separation mode: chromatography; mass spectrometry mode: legacy or data-dependent type; primary ion data type: centroid; ion mode: negative mode; in the Analysis parameter setting dialog box, the peak alignment file (Alignment), possible adduct (Adduct), and primary identification reference file (MSP file) were set.

[0128] The method for comprehensive nontarget screening of PFAS in this example was used to screen 50 serum samples of patients diagnosed with thyroid cancer and 46 serum samples of the general population. The results showed that 31 PFAS of 13 categories were identified in the 96 serum samples. Taking the 4 PFAS screened as examples, they were 5:2 FTCA (5:2 fluorotelomer carboxylic acid, a novel biotransformation intermediate, which was formed by the rapid decarboxylation of a-OH 5:3 FTCA to generate 5:2 FTAL (fluorotelomer aldehyde), and then oxidized), 2H-PFMEOH (dihydrogen-substituted perfluoromonoether alcohol, first found in serum), 3:2 FTOH (3:2 fluorotelomer alcohol, first found in human body), and H-PFMEOH (hydrogen-substituted perfluoromonoether alcohol). At present, there was a lack of standard products for these substances, and these substances could not be detected by targeted screening. The method of the present disclosure could detect these substances through nontarget screening. Taking the above 4 PFAS as an example, their mass spectrum analysis was as follows:

[0129] FIG. 11 showed a structural elucidation of 5:2 FTCA; FIG. 12 showed a chromatogram of 5:2 FTCA. Table 7 showed the diagnostic fragment information of 5:2 FTCA.

[0130] Table 7 Diagnostic fragment information of 5:2 FTCA

[0131] Ionic formula Exact mass[M-H] Theoretical mass[M-H] Error (ppm) Error (mDa) c3f3- 92.9957 92.9952 5.49 0.51 C6F9- 242.9863 242.9856 2.64 0.64 CeHFio' 262.9934 262.9919 5.79 1.52 C7H2F11O2 326.9985 326.9879 1.76 0.58

[0132] FIG. 13 showed a structural elucidation of 2H-PFMEOH; FIG. 14 showed a chromatogram of 2H-PFMEOH. Table 8 showed the diagnostic fragment information of 2H-PFMEOH.

[0133] Table 8 Diagnostic fragment information of 2H-PFMEOH

[0134] Ionic formula Exact mass[M-H] Theoretical mass[M-H] Error (ppm) Error (mDa) c2f5- 118.9926 118.9920 4.99 0.59 C3F7- 168.9897 168.9888 5.37 0.91 C4F9- 218.9867 218.9856 4.80 1.05 C5Fii- 268.9833 268.9824 3.14 0.85 CsFn' 418.9747 418.9729 4.41 1.85

[0135] FIG. 15 showed a structural elucidation of 3:2 FTOH; FIG. 16 showed a chromatogram of 3:2 FTOH. Table 9 showed the diagnostic fragment information of 3:2 FTOH.

[0136] Table 9 Diagnostic fragment information of 3:2 FTOH

[0137] Ionic formula Exact mass[M-H] Theoretical mass[M-H] Error (ppm) Error (mDa) cf3- 68.9981 68.9952 41.17 2.84 C4H2F3O- 123.0089 123.0058 25.17 3.10 C4H2F7- 183.0047 183.0045 1.08 0.20 C5H2F7- 195.0047 195.0045 1.32 0.26 C5H2F7O' 210.9999 210.9994 2.20 0.46 C5H4F7O- 213.0155 213.0150 2.03 0.43

[0138] FIG. 17 showed a structura chromatogram of H-PFMEOH. Table elucidation of H-PFIV 10 showed the diagnos EOH; FIG. 18 showed a tic fragment information of H-PFMEOH.

[0139] Table 10 Diagnostic fragment information of H-PFMEOH

[0140] Ionic formula Exact mass[M-H] Theoretical mass[M-H] Error (ppm) Error (mDa) cf3- 68.9957 68.9952 7.69 0.53 C4F5O- 146.9875 146.9869 4.01 0.59 C3HF6O' 166.9940 166.9932 4.86 0.81 C3F7O' 184.9843 184.9837 3.04 0.56 C5F9O2 262.9760 262.9755 2.10 0.55 C6HF12O2' 332.9792 332.9785 2.10 0.70

[0141] The above descriptions are merely preferred implementations of the present disclosure. It should be noted that a person of ordinary skill in the art may further make several improvements and modifications without departing from the principle of the present disclosure, but such improvements and modifications should be deemed as falling within the protection scope of the present disclosure.

Claims

26WHAT IS CLAIMED IS:

1. A method for nontarget screening of per- and polyfluoroalkyl substances (PFAS), comprising the following steps:subjecting a sample to be tested to pretreatment, chromatographic separation, and mass spectrometry detection in sequence to obtain mass spectrum data;converting the mass spectrum data into a file in mzXML format, conducting peak extraction and tandem mass spectrometry (MS / MS) spectra extraction, and then conducting isotope detection and adduct detection; wherein a peak table in csv format is obtained after the peak extraction is completed; and a MS / MS spectra table in csv format is obtained after the MS / MS spectra extraction is completed;constructing a local database according to data obtained from the peak extraction and the MS / MS spectra extraction, specifically comprising: converting the peak table and the MS / MS spectra table into the local database, wherein a structure of the local database is a relational database model; wherein when the MS / MS spectra table is converted and imported into the local database, the MS / MS spectra table is specifically split into a MS / MS spectra maximum response table, a MS / MS spectra basic information table, and a fragment table and then imported into the local database, and the MS / MS spectra maximum response table, the MS / MS spectra basic information table, and the fragment table are connected by a same column; the local database comprises the peak table, a sample information table, the MS / MS spectra basic information table, the MS / MS spectra maximum response table, and the fragment ion table; the peak table comprises peak serial number (SN), mass-to-charge ratio, retention time, response intensity, and sample SN; the sample information table comprises sample SN and sample information; the MS / MS spectra basic information table comprises MS / MS spectra SN, parent ion mass-to-charge ratio, sample SN, retention time, and parent ion response intensity; the MS / MS spectra maximum response table comprises MS / MS spectra SN and maximum response value of fragments; and the fragment ion table comprises response ID, MS / MS spectra SN, ion mass-to-charge ratio, and response intensity; finally, indexes for the aforementioned five SQL tables are created to enhance query efficiency;providing a suspect list, and then conducting suspect screening in the suspect list according to a structured query language (SQL) script and the local database, wherein the suspect screening comprises: in the SQL script, searching for an expected [M-H]' precursor of each compound using a Query_msl statement according to the suspect list, with a mass tolerance set to 1.5 mDa and 5 ppm; and searching for results having MS / MS spectra records within a mass tolerance of27 03 261.5 mDa using a mass-to-charge ratio as an input in a Query _ms2 statement;conducting fragment-based nontarget screening according to the SQL script and diagnostic fragments and neutral losses of the PFAS, wherein the fragment-based nontarget screening comprises: conducting searching using a Query fragment statement in the SQL script, with a mass tolerance set to 1.5 mDa, wherein the diagnostic fragments comprise CnF2n+f, CnF2n-f, CnF2n+iO', CnF2n-iO', SCLF', and SO3F, and the neutral losses comprise HF and CHFO2;conducting homolog-based nontarget screening according to homologous series of the PFAS, wherein the homolog-based nontarget screening comprises: conducting filtering with a mass defect greater than 0.85 or a mass defect less than 0.15, calculating the homologous series of the PFAS through a Python script, and then screening a homologous PFAS; andconducting structural identification and true positive testing based on screening results obtained from the suspect screening, the fragment-based nontarget screening, and the homolog-based nontarget screening as well as detection results obtained from the isotope detection and the adduct detection, thereby achieving the nontarget screening of the PFAS in the sample to be tested, wherein the structural identification comprises: conducting literature or database matching identification on reported PFAS; determining an elemental composition of potential PFAS precursors and fragments on unreported positive substances, wherein parameters comprise: C<50, H<50, F<50, O<10, S<5, Cl<3, Br<3, P<3, and N<10; and a mass tolerance is set to 1.5 mDa.

2. The method for nontarget screening of PFAS according to claim 1, wherein software-assisted verification is further conducted after the true positive testing is completed.

Images