Assessment of PDMS-Water Partition Coefficients: Implications for Passive Environmental Sampling of Hydrophobic Organic Compounds

Abstract

Solid-phase microextraction (SPME) has shown potential as an in situ passive-sampling technique in aquatic environments. The reliability of this method depends upon accurate determination of the partition coefficient between the fiber coating and water (K(f)). For some hydrophobic organic compounds (HOCs), K(f) values spanning 4 orders of magnitude have been reported for polydimethylsiloxane (PDMS) and water. However, 24% of the published data examined in this review did not pass the criterion for negligible depletion, resulting in questionable K(f) values. The range in reported K(f) is reduced to just over 2 orders of magnitude for some polychlorinated biphenyls (PCBs) when these questionable values are removed. Other factors that could account for the range in reported K(f), such as fiber-coating thickness and fiber manufacturer, were evaluated and found to be insignificant. In addition to accurate measurement of K(f), an understanding of the impact of environmental variables, such as temperature and ionic strength, on partitioning is essential for application of laboratory-measured K(f) values to field samples. To date, few studies have measured K(f) for HOCs at conditions other than at 20° or 25 °C in distilled water. The available data indicate measurable variations in K(f) at different temperatures and different ionic strengths. Therefore, if the appropriate environmental variables are not taken into account, significant error will be introduced into calculated aqueous concentrations using this passive sampling technique. A multiparameter linear solvation energy relationship (LSER) was developed to estimate log K(f) in distilled water at 25 °C based on published physicochemical parameters. This method provided a good correlation (R(2) = 0.94) between measured and predicted log K(f) values for several compound classes. Thus, an LSER approach may offer a reliable means of predicting log K(f) for HOCs whose experimental log K(f) values are presently unavailable. Future research should focus on understanding the impact of environmental variables on K(f). Obtaining the data needed for an LSER approach to estimate K(f) for all environmentally relevant HOCs would be beneficial to the application of SPME as a passive-sampling technique.