Non-Targeted Analysis of Per- and Polyfluoroalkyl Substances in Environmental Extracts by Microflow LC-QToF Coupled with Multi-spray ESI

Significance of the topic

Per- and polyfluoroalkyl substances (PFAS) are a large, chemically diverse and continuously evolving class of environmental contaminants with persistence, mobility and toxicity concerns. Established targeted assays cover known PFAS but cannot capture novel, emerging or transformation products. Non-targeted analysis (NTA) expands the capability to detect and characterize unknown PFAS in complex environmental matrices; however, NTA is often limited by instrumental sensitivity and by data-processing complexity. This study demonstrates how microflow liquid chromatography combined with a multi-nozzle ESI emitter and high-resolution QToF mass spectrometry, together with dedicated software workflows, can increase sensitivity and streamline discovery and confirmation of PFAS in environmental extracts.

Objectives and study overview

• Evaluate microflow LC coupled to a multi-nozzle emitter and a high-resolution QToF (DIA acquisition) for non-targeted PFAS screening in environmental extracts.
• Develop a workflow for feature detection, Kendrick mass defect (KMD) filtering, library-assisted identification and statistical prioritization of candidates across multiple samples.
• Demonstrate the approach on real-world matrices: AFFF-impacted soil extracts (from a NIST interlaboratory study) and landfill leachate extracts, including a comparison of two WAX SPE cartridges for sample cleanup.

Methods and instrumentation

Sample introduction used microflow LC (approximately 5 µL/min) with direct injection onto a reversed-phase C18 column (short, sub-2 µm particle format) to minimize bias introduced by trap columns. Mobile phases were aqueous ammonium acetate (low millimolar) and acetonitrile, and a shallow gradient increased organic content from low to high over tens of minutes to separate PFAS homologous series. A Newomics multi-nozzle DuoESI emitter improved desolvation and ionization efficiency in negative electrospray mode. A Shimadzu LCMS-9050 Quadrupole Time-of-Flight (QToF) mass spectrometer acquired full-scan data (m/z ~100–1000) and performed data-independent acquisition (DIA) with broad isolation windows and wide collision-energy variation to collect MS/MS information for many coeluting precursors. Key acquisition details were optimized with PFAS standards and included low-flow conditions, negative ionization, and DIA windows sized to balance coverage and spectral complexity.

Instrumentation used

• Shimadzu Nexera Mikros microflow liquid chromatograph.
• Newomics DuoESI source with M3 multi-nozzle emitter to enhance desolvation and sensitivity.
• Shimadzu LCMS-9050 QToF mass spectrometer operated in negative ion mode with DIA (wide isolation windows and stepped collision energies).
• Software suite: Insight Explore, Insight Discovery, and Profiler for feature detection, KMD filtering, deconvolution of DIA MS/MS, library searching and multivariate statistics.

Main results and discussion

• Sensitivity gains and detection: The microflow + multi-nozzle emitter configuration improved desolvation and increased signal for PFAS species, enabling detection of low-abundance features in environmental extracts that might be missed in conventional flow regimes.
• Kendrick mass defect (KMD) filtering: Using CF2 as the repeating unit and an 80 mDa KMD tolerance, KMD filtering effectively reduced data complexity and highlighted homologous PFAS series. This approach helped prioritize features for further interrogation.
• Example identification: An extracted ion chromatogram (m/z 499.0033) revealed a distinct peak absent from existing libraries. DIA deconvoluted MS/MS showed a diagnostic FSO3− fragment, supporting a tentative assignment to a sulfated PFAS species.
• Multivariate prioritization: Principal component analysis (PCA) aligned over 1,000 features across sample extracts and blanks; loading plots pointed to features contributing to sample separation. One identified contributor was 6:2 fluorotelomer sulfonate (6:2 FTS) at m/z 426.966 (RT ~10.7 min).
• SPE cartridge comparison: Profiler aligned 641 features for an extraction comparison. Volcano-plot filtering highlighted differential cleanliness/retention between two WAX cartridge sources—302 features enriched in cartridge 2 and 129 enriched in cartridge 1—illustrating the method’s ability to reveal extraction-dependent chemical space.
• Software-enabled workflow: Insight Explore was used for screening, KMD filtering and tentative assignments; Insight Discovery and Profiler enabled batch processing, feature picking, library searches and statistical comparisons, which together produced a streamlined NTA pipeline for PFAS in complex matrices.

Benefits and practical applications of the method

• Improved sensitivity for low-abundance PFAS and transformation products due to microflow LC and multi-nozzle ESI, increasing the chance of detecting novel contaminants in complex matrices.
• DIA on a high-resolution QToF provides MS/MS evidence for many precursors, facilitating structural elucidation without prior target lists.
• KMD filtering and homologous-series detection rapidly reduce candidate lists and focus interpretation on chemically plausible PFAS series.
• Integrated software workflows enable batch processing, statistical prioritization (PCA, volcano plots), and library-building—accelerating confirmation of knowns and discovery of unknowns across multiple samples.
• Practical use cases include environmental monitoring of AFFF-impacted soils, landfill leachate screening, remediation monitoring, and suspect/non-targeted surveys following contamination events.

Future trends and potential uses

• Expanded spectral libraries and community-shared suspect lists will increase the rate of confident identifications when combined with sensitive microflow-QToF platforms.
• Hybrid data-acquisition strategies (combining DIA and targeted MS/MS triggers) and improved DIA deconvolution algorithms will further improve structural annotation of coeluting species in complex matrices.
• Automated chemoinformatics (in silico fragmentation prediction, retention-time prediction, and AI-based prioritization) will reduce manual curation and accelerate discovery-to-confirmation workflows.
• Miniaturized LC-MS configurations and optimized emitter geometries will continue to push sensitivity while reducing solvent consumption, making routine NTA more feasible for monitoring programs.
• Standardized reporting, interlaboratory comparisons (such as NIST initiatives) and harmonized sample-preparation protocols will improve comparability of NTA PFAS data across labs.

Conclusion

Microflow LC coupled with a multi-nozzle ESI emitter and high-resolution QToF in DIA mode enhances the ability to perform non-targeted PFAS screening in challenging environmental extracts. When combined with Kendrick mass defect filtering and a suite of processing tools for feature alignment, deconvolution, library search and statistics, the workflow supports effective prioritization and tentative identification of known and unknown PFAS. The approach is well suited for discovery-oriented environmental monitoring and for supporting the expansion of PFAS libraries and suspect lists.

References

• Place BJ, et al. Per- and Polyfluoroalkyl Substances—Non-Targeted Analysis Interlaboratory Study Final Report. US Department of Commerce, National Institute of Standards and Technology; 2024.
• Place B. Suspect List of Possible Per- and Polyfluoroalkyl Substances. National Institute of Standards and Technology; 2021. DOI: 10.18434/mds2-2387 (Version 1.7).