Analytical and Toxicological Analysis of Perfluorinated Compounds Present in the Environment

Abstract

Perfluorinated compounds (PFCs) have been produced in relatively large quantities since the 1950s for a wide range of applications such as carpet coatings, food packaging, shampoos, paper, and fire-fighting foams. PFCs are globally ubiquitous in both remote and urban environments. PFCs are present in various matrices including; human blood (whole, plasma and serum), sediments, water, and wildlife. The primary objective of the research described herein is to address data-gaps which have hindered attempts to effectively characterize the risks of PFCs to both humans and wildlife. The hypotheses of this thesis are that (1) current analytical methods for the detection of PFCs are flawed and are in need of improvement, (2) concentrations of PFCs in South Korea are elevated and there is associated risk to both humans and wildlife, and (3) model PFCs such as perfluorooctanesulfonate (PFOS) and perfluorooctanoic acid (PFOA) are poor predictors of effects caused by exposure to other PFCs. Accurate and precise measurement is essential for effective decision making regarding the production and usage of PFCs. Many issues such as impure standards (either homologs of the standards or unrelated compounds), interactions between isotopically labeled and non-labeled analytes, and the presence of multiple isomers complicate the accurate and precise quantification of PFCs. It has been reported that the relative response factors of isotopically labeled standards and unlabeled standards of the same PFC could be different. Individual (100 ng mL-1) solutions of PFOA and PFOS were analyzed using HPLC-MS/MS under negative-ion-electrospray to detect any impurities present down to 0.5 to 0.1% relative to the major component. Purity of the standards ranged from approximately 86% to ≥ 97%. Standard solutions of unlabeled and isotopically labeled materials were analyzed to compare response factors of isotopically labeled analytes versus their non-labeled counterparts in three different matrices at equivalent concentrations: organic solvent (methanol), serum extract, and water present individually and iii concurrently. Not all labeled analytes have the same response factor as their non-labeled complement, and in at least one case the matrix in which the standard is present may cause significant suppression of response. Standard solutions of electrochemical fluorination produced PFOA and PFOS were quantified under multiple reaction monitoring (MRM) mode, using calibration curves prepared from standards consisting primarily of linear standards only, and our results have shown that in general, the use of linear only standards may cause under prediction of concentrations, and that the working range of these standards may be limited. Previous studies have reported concentrations of PFCs in Asia to be relatively great and in particular Korea has shown to have some of the highest environmental levels ever detected. Despite this fact, relatively little was known about sources, distribution and fate among matrixes. In 2008 and 2009 soil, sediment, water, and biota were collected from the western coast of Korea and were analyzed to determine occurrence and sources of PFCs. PFCs were significantly concentrated in some water and biological samples, while concentrations of PFCs in soils and sediments were relatively low. The most widely detected compound was found to be PFOS, with maximum in water of 450 ng/L and in fish of 612 ng/g, dw. PFOS in water and biota were both less than those thought to cause toxicity, however; in both cases concentrations were within a factor of 10 to a possible toxicity threshold concentration. Although in general the concentrations of PFCs in all three media were reduced from 2008 levels, the calculated bioconcentration factor for PFOS in fish was among the highest ever reported. Overall, the detection of PFCs at relatively great concentrations in various environmental matrices from this region of Korea suggests that further study and characterization of these chemicals and their potential risk to both humans and wildlife is needed. iv While PFOS has been extensively studied, other PFCs including replacement chemicals such as perfluorobutanesulfonate (PFBS) and perfluorobutyrate (PFBA), have not been well characterized. Despite the relative lack of data available describing these other PFCs it has been assumed that they will cause similar or lesser effects than PFOS. To test the null hypothesis that all PFCs, including shorter chain-length replacements, act via similar modes of action, rat H4IIE cells were exposed to 10 PFCs, all of which are routinely found in the environment, and the mRNA abundance of 7 target genes was quantified using real-time PCR. Significant changes in mRNA abundance were observed. Effects caused by the shorter chain replacement chemicals differed significantly from those caused by PFOS or PFOA. Furthermore, not all of the PFCs caused the same effects, and changes could not simply be attributed to chain-length or functional group. These differences support that these replacement chemicals do not act through the same mechanisms as the more studied PFOS and PFOA.