Date of Graduation
12-2023
Document Type
Dissertation
Degree Name
Doctor of Philosophy in Engineering (PhD)
Degree Level
Graduate
Department
Civil Engineering
Advisor/Mentor
Julian Fairey
Committee Member
Wahman, David
Second Committee Member
Wickramasinghe, Ranil
Third Committee Member
Zhang, Wen
Keywords
DGT, Diffusion, Environmental Sampling, Passive Sampling, PFAS
Abstract
Environmental sampling is the backbone of regulatory compliance in public water systems. Coupled with continuously improving analytical capabilities, grab sampling has revealed the ubiquity of emerging contaminants in water resources such as per- and polyfluoroalkyl substances (PFAS). The Environmental Protection Agency’s (EPAs) Method 533 and 537.1 specify 0.25 L grab samples undergo a solid phase extraction (SPE) to enrich PFAS up to 250-fold in 1 mL of methanol prior to analysis by liquid chromatography-tandem mass spectrometry (LC-MS/MS). This process can achieve lowest concentration minimum reporting levels (LCMRLs) between 0.53–16 ng•L−1 for the twenty-nine prioritized PFAS in drinking waters. However, toxicity assessments of PFAS spurred updated EPA health advisory levels 2–3 orders of magnitude below the Methods 537.1 and 533 LCMRLs for perfluorooctanoic acid (PFOA, 0.004 ng•L−1) and perfluorooctane sulfonic acid (PFOS, 0.02 ng•L−1). Further, the EPA has proposed maximum contaminant levels for PFOA and PFOS at 4 ng•L−1, motivating the need to develop improved sampling methodologies to quantify PFAS near the health advisory levels. Diffusive gradients in thin-films (DGT) passive samplers consist of a binding layer containing a sorbent with rapid uptake kinetics and large affinity for the target analytes overlain by a gel layer of precise thickness that restricts mass transport from the bulk water to the binding layer to molecular diffusion. During deployment, DGT passive samplers continuously accumulate analyte mass in the binding layer at a rate proportional to the analyte diffusion coefficient in the gel, DGel. DGTs are deployed for a specified duration, tD, which is typically between 3–90 days. Following tD, analytes are extracted into methanol for LC-MS/MS analysis. While DGT passive samplers enrich analytes in the binding layer, like the SPE process in Methods 537.1 and 533, the extent of PFAS enrichment is unknown, requiring development as an alternative sampling approach. In this work, development of a DGT passive sampler began with method development to support determination of DGel values. The standard analysis of diffusion cell (D-Cell) test data relied on an assumption of pseudo-steady-state flux which proved overly simplistic, motivating the development of a finite difference model (FDM) that accurately captured non-steady-state diffusive flux. Tests with nitrate validated the D-Cell design and operation and resulted in improved estimates of DGel ± 95 % confidence intervals (CIs). This method was applied to 32 PFAS to measure their DGel ± 95 % CIs. Surfactant properties of many of the larger molecular weight PFAS proved problematic, leading to aggregation at the air-water interface, and necessitating D-Cell tests in methanol. The resultant DGel ± 95 % CI values were adjusted to water using the Stokes-Einstein equation. Last, DGT passive samplers were deployed in a laboratory-based carousel test in which they accurately quantified 21 PFAS at low nanogram per liter levels. These results showed the DGT technique could be used as an alternative to grab sampling and analysis by EPA Methods 537.1 and 533 and has the potential to enable PFAS quantitation at levels near the health advisory limits.
Citation
Hodges, S. (2023). Diffusive Gradients in Thin-Films Passive Samplers for Quantitation of Per- and Polyfluoroalkyl Substances in Water: Kinetic Modeling, Experimental Methods, and Outlook for Regulatory Monitoring. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/5101