Source apportionment of perfluoroalkyl substances in surface sediments from lakes in Jiangsu Province, China: Comparison of three receptor models

Abstract

Receptor models have been proved as useful tools to identify source categories and quantitatively calculate the contributions of extracted sources. In this study, sixty surface sediment samples were collected from fourteen lakes in Jiangsu Province, China. The total concentrations of C₄–C₁₄-perfluoroalkyl carboxylic acids and perfluorooctane sulfonic acid (∑₁₂PFASs) in sediments ranged from 0.264 to 4.44 ng/g dw (dry weight), with an average of 1.76 ng/g dw. Three commonly-applied receptor models, namely principal component analysis-multiple linear regression (PCA-MLR), positive matrix factorization (PMF) and Unmix models, were employed to apportion PFAS sources in sediments. Overall, these three models all could well track the ∑₁₂PFASs concentrations as well as the concentrations explained in sediments. These three models identified consistently four PFAS sources: the textile treatment sources, the fluoropolymer processing aid/fluororesin coating sources, the textile treatment/metal plating sources and the precious metal sources, contributing 28.1%, 37.0%, 29.7% and 5.3% by PCA-MLR model, 30.60%, 39.3%, 22.4% and 7.7% by PMF model, and 20.6%, 52.4%, 20.2% and 6.8% by Unmix model to the ∑₁₂PFASs, respectively. Comparative statistics of multiple analytical methods could minimize individual-method weaknesses and provide convergent results to enhance the persuasiveness of the conclusions. The findings could give us a better knowledge of PFAS sources in aquatic environments.

Introduction

The mass production and widespread usage of perfluoroalkyl substances (PFASs) have begun since the late 1940s (Kim et al., 2012). PFASs are ubiquitous in various environmental media, such as water, soils, sediments, sewage sludge, biota and human bodies (Wang et al., 2015, Zhao et al., 2012). It was inferred that a large proportion of PFASs would be released to the surface waters, and sediments, as the natural environment of benthic organisms, are considered as one of the most important environmental sinks of PFASs (Prevedouros et al., 2006). PFASs have been found in sediments from several countries and regions, and PFAS concentrations have reached up to 800 ng/g dry weight (dw) (Ahrens et al., 2015, Campo et al., 2015, Zhou et al., 2013). Source apportionment of PFASs in sediments is of great significance for pollution control and ecological protection of aquatic environments.

Qualitative and semiquantitative methods have been widely employed to identify PFAS sources. Xiao et al. (2012) adopted cluster analysis, correlation analysis, ANOVA and per capita discharge to distinguish different PFAS patterns in influent samples from thirty-seven wastewater treatment plants in a multi-city survey. Murakami et al. (2009) used ratio methods to evaluate PFAS contributions from street runoff and wastewater to the aquatic environments. A few methods, e.g., principal component analysis (PCA), multiple linear regression model (MLR), positive matrix factorization (PMF) and Unmix models, have been applied to quantitative source apportionment of PFASs. However, only one or two of these models are usually applied to source apportionment of PFASs, and it is not enough to provide comprehensive information for PFAS sources (Kuroda et al., 2014, Qi et al., 2016).

Jiangsu Province, located in the lower reaches of the Yangtze River and Huaihe River, is traversed by the Beijing-Hangzhou Grand Canal from north to south. It forms part of the Yangtze River Delta urban agglomeration, which is one of the six world-class city clusters. As one of the most intensively industrialized provinces in China, Jiangsu Province contributed the largest portion of perfluorooctane sulfonic acid (PFOS) emissions in China (Xie et al., 2013). PFOS concentrations in Taihu Lake in Jiangsu Province even exceeded US EPA standards of 0.2 μg/L for PFOS in drinking water (US EPA, 2009, Yang et al., 2011). Our published work showed that lake sediments from Jiangsu Province have also suffered the heaviest PFAS pollution (mean 1.76 ng/g dw) among all provinces studied (Qi et al., 2016). In this study, PCA-MLR, PMF and Unmix models were applied to identify source categories and quantitatively calculate source contributions of PFASs in lake sediments from Jiangsu Province. Results from three models were evaluated and compared in order to improve source apportionment of PFASs. Comparative statistics could enhance the persuasiveness of the conclusions and offer us a better knowledge of PFAS sources in lake sediments.