A new analytical method was developed to quantify 26 newly-identified and 21 legacy (e.g. perfluoroalkyl carboxylates, perfluoroalkyl sulfonates, and fluorotelomer sulfonates) per and polyfluorinated alkyl substances (PFAS) in groundwater and aqueous film forming foam (AFFF) formulations. Prior to analysis, AFFF formulations were diluted into methanol and PFAS in groundwater were micro liquid-liquid extracted. Methanolic dilutions of AFFF formulations and groundwater extracts were analyzed by large-volume injection (900 μL) high-performance liquid chromatography tandem mass spectrometry. Orthogonal chromatography was performed using cation exchange (silica) and anion exchange (propylamine) guard columns connected in series to a reverse-phase (C18) analytical column. Method detection limits for PFAS in groundwater ranged from 0.71 ng/L to 67 ng/L, and whole-method accuracy ranged from 96% to 106% for analytes for which matched authentic analytical standards were available. For analytes without authentic analytical standards, whole-method accuracy ranged from 78 % to 144 %, and whole-method precision was less than 15 % relative standard deviation for all analytes. A demonstration of the method on groundwater samples from five military bases revealed eight of the 26 newly-identified PFAS present at concentrations up to 6900 ng/L. The newly-identified PFAS represent a minor fraction of the fluorinated chemicals in groundwater relative to legacy PFAS. The profiles of PFAS in groundwater differ from those found in fluorotelomer- and electrofluorination-based AFFF formulations, which potentially indicates environmental transformation of PFAS.

Poly- and perfluoroalkyl substances (PFASs) are a class of fluorinated chemicals that are utilized in firefighting and have been reported in groundwater and soil at several firefighter training areas. In this study, soil and groundwater samples were collected from across a former firefighter training area to examine the extent to which remedial activities have altered the composition and spatial distribution of PFASs in the subsurface. Log Koc values for perfluoroalkyl acids (PFAAs), estimated from analysis of paired samples of groundwater and aquifer solids, indicated that solid/water partitioning was not entirely consistent with predictions based on laboratory studies. Differential PFAA transport was not strongly evident in the subsurface, likely due to remediation-induced conditions. When compared to the surface soil spatial distributions, the relative concentrations of perfluorooctanesulfonate (PFOS) and PFAA precursors in groundwater strongly suggest that remedial activities altered the subsurface PFAS distribution, presumably through significant pumping of groundwater and transformation of precursors to PFAAs. Additional evidence for transformation of PFAA precursors during remediation included elevated ratios of perfluorohexanesulfonate (PFHxS) to PFOS in groundwater near oxygen sparging wells.

A new method was developed for the analysis of natural and synthetic androgenic steroids and their selected metabolites in aquatic environmental matrices using direct large-volume injection (LVI) high performance liquid chromatography (HPLC) tandem mass spectrometry (MS/MS). Method accuracy ranged from 88 to 108% for analytes with well-matched internal standards. Precision, quantified by relative standard deviation (RSD), was less than 12%. Detection limits for the method ranged from 1.2 to 360 ng/L. The method was demonstrated on a series of 1-hr composite wastewater influent samples collected over a day with the purpose of assessing temporal profiles of androgen loads in wastewater. Testosterone, androstenedione, boldenone, and nandrolone were detected in the sample series at concentrations up to 290 ng/L and loads up to 535 mg. Boldenone, a synthetic androgen, had a temporal profile that was strongly correlated to testosterone, a natural human androgen, suggesting its source may be endogenous. An analysis of the sample particulate fraction revealed detectable amounts of sorbed testosterone and androstenedione. Androstenedione sorbed to the particulate fraction accounted for an estimated five to seven percent of the total androstenedione mass.

The time-honored convention of concentrating aqueous samples by solid-phase extraction (SPE) is being challenged by the increasingly wide spread use of large-volume injection (LVI) liquid chromatography-mass spectrometry (LC-MS) for the determination of traces of polar organic contaminants in environmental samples. Although different LVI approaches have been proposed over the last 40 years, the simplest and most popular way of performing LVI is known as single column LVI (SC-LVI), in which a large-volume of an aqueous sample is directly injected into an analytical column. For the purposes of this critical review, LVI is defined as an injected sample volume that is ≥ 10% of the void volume of the analytical column. Compared to other techniques, SC-LVI is easier to set up, as it only requires small hardware modifications to existing autosamplers and, thus, will represent the main focus of the current review. Although not new, SC-LVI is gaining acceptance and the approach is emerging as a technology that will render SPE nearly obsolete for many environmental applications. In this review, we discuss 1) the history and development of various forms of LVI, 2) the critical factors that one needs to consider when creating and optimizing SC-LVI methods and 3) example applications that demonstrate the range of environmental matrices, to which LVI is applicable such as drinking water, groundwater and surface water including seawater as well as wastewater. Furthermore, we give responses to answer a set of 'frequently asked questions' typically encountered from audiences and we indicate future directions and areas that need to be addressed to fully delineate the limits of SC-LVI.

Environmental analysis by large-volume injection (LVI) was compared to solid-phase extraction (SPE) based methods using matrix effects as a quantitative indicator of analytical signal quality. LVI was performed by the direct injection of 900 μL of wastewater onto a high-performance liquid chromatography (HPLC) column while SPE-based methods utilized octadecyl silane (C18) and hydrophobic-lypophilic balance (HLB) solid phases to preconcentrate wastewater prior to analysis. Model analytes from three classes of environmental contaminants were selected for study including four estrogens (estrone, estradiol, estriol, and ethinylestradiol), eight perfluoroalkyl carboxylates (C4-C11), and five perfluoroalkyl sulfonates (C4, C6-C8, and C10). The matrix effects on analytes were assessed by two approaches (quantitatively by calculating percent matrix effects and qualitatively with postcolumn infusions) and compared across LVI- and SPE-based methods at constant (high and low) analyte-to-matrix mass ratios. The results from this study demonstrated that the LVI-based method produced analytical signals of quality similar to the two SPE-based methods. Furthermore, LVI presented a clear advantage over SPE because it was performed at lower cost, required fewer materials, involved less labor and eliminated the analyte loss associated with SPE.

No abstract

No abstract

An analytical method is presented here that is sensitive to the parts-per-quadrillion (pg/L) for estrogens in surface water. The estrogens included for study were estrone, 17β-estradiol, estriol, 17α-ethinylestradiol, and equilin. The method consisted of the small-scale liquid-liquid extraction of surface water followed by derivation with dansyl chloride. Analyte separation and detection were performed by high-pressure liquid-chromatography and tandem mass-spectrometry. A large volume (100 μL) of the sample was injected on-column to increase the analyte mass sent to the detector. The detection limits of the method were 0.045 ng/L for estrone, 0.086 ng/L for 17β-estradiol, 0.030 ng/L for estriol, 0.049 ng/L for 17α-ethinylestradiol, and 0.13 ng/L for equilin. The whole-method accuracy ranged from 93 ± 5.8% to 105 ± 4.5% for all the analytes at two different spike levels. Similarly, the precision of the method was less than 8.0% relative standard deviation. The final method was used to analyze a series of samples from the Mississippi River spanning 51 river miles. Estrone was detected in all of the samples and 17β-estradiol was detected in one. Concentrations of estrone ranged from between the detection and quantification limits up to 0.63 ng/L. Increases in the concentration of estrone could be observed downstream from potential sources including a drinking water treatment plant. 17β-estradiol was detected below its quantitation limit in a sample taken downstream from a wastewater treatment plant.