A passive sampler based on solid phase microextraction (SPME) for quantifying hydrophobic

Sediment quality assessment is often hindered by the lack of agreement between chemical and biological lines of evidence. One limitation is that the bulk sediment toxicant concentration, the most widely used chemical parameter, does not always represent the bioavailable concentration, particularly for hydrophobic organic compounds (HOCs) in highly contaminated sediments. In this study, a porewater sampler that utilizes solid phase microextraction (SPME) to measure freely dissolved (“bioavailable”) HOC concentrations was developed and tested. A single polydimethylsiloxane (PDMS) coated SPME fiber was secured in a compact protective housing that allows aqueous exchange with whole sediment while eliminating direct contact with sediment particles. Fibers with three PDMS coating thicknesses were first calibrated for 12 model polycyclic aromatic hydrocarbons (PAH), polychlorinated biphenyls (PCBs), DDTs, and chlordanes, representing HOCs of current regulatory concern. Pre-calibrated samplers were exposed to spiked estuarine sediment in lab microcosms to determine the time to equilibrium and equilibrium concentrations across a range of sediment contamination. Time to equilibrium ranged from 14 to 110 days, with 30 days being sufficient for more than half of the target HOCs. Ranging from 0.009 to 2400 ng/L, equilibrium SPME measurements were highly correlated with, but generally lower than HOC porewater concentrations determined independently by liquid-liquid extraction. This concept shows promise for directly measuring the freely dissolved concentration of HOCs in sediment porewater, a previously difficult-to-measure parameter that will improve frameworks for assessing the impacts of contaminated sediments. INTRODUCTION Sediments are the largest sink of HOCs in the aquatic environment. Accumulation of HOCs to high levels in sediments poses a risk to both ecological and human health via direct and indirect pathways. For example, chlordanes are associated with toxic effects exerted on benthic organisms that directly inhabit contaminated sediments. Conversely, marine and terrestrial mammals (including humans) are indirectly exposed to HOCs via food web transfer. The resulting biomagnification of toxicants like DDT, PCBs, and brominated flame retardants (e.g., polybrominated diphenyl ethers (PBDEs)) can result in and/or exacerbate reproductive and immunosuppressive impacts. The bulk sediment concentration of a suspected toxicant is a logical and thus commonly used indicator of potential HOC exposure for benthic organisms. Coupled with biological endpoints like benthic community condition and sediment toxicity, bulk “sediment chemistry” has been widely used in sediment quality assessment (Chapman et al. 1987, Bay et al. 2007). To account for the affinity of HOCs with organic matter (Karickhoff et al. 1979), models relating aqueous and solid phase partitioning were created to help explain differences in bioavailability observed in toxicological and bioaccumulation endpoints (DiToro et al. 1991). More recently, heterogeneities within the organic subcomponent of soils and sediments have also been shown to have a profound effect on the partitioning and bioavailability of HOCs. Condensed, sooty materials known collectively as black carbon (BC) were first shown to reduce the bioavailability of PAH in contaminated harbor sediments (McGroddy A passive sampler based on solid phase microextraction (SPME) for quantifying hydrophobic organic contaminants in sediment porewater Keith A. Maruya, Eddy Y. Zeng1, David Tsukada and Steven M. Bay I Chinese Academy of Sciences, Guangzhou Institute of Geochemistry, State Key Laboratory of Organic Geochemistry, Guangzhou, China SPME passive sampler for hydrophobic organic compounds 39 SPME passive sampler for hydrophobic organic compounds 40 and Farrington 1995). Several subsequent studies have shown a similar influence of BC on the binding strength and bioavailability of additional classes of HOCs, including PCBs (Jonker et al. 2004), as well as legacy and current-use pesticides (Xu et al. 2008). The effectiveness of BC in altering HOC bioavailability, at least in the short term, has resulted in its consideration in remedial strategies for highly contaminated sediments (Tomaszewski et al. 2007). A growing body of evidence suggests that the freely dissolved phase of HOCs represents the highly bioavailable fraction (Mayer et al. 2000a, Kraaij et al. 2003). By definition, however, determination of this parameter must be made at ultralow levels (e.g., in the sub-parts per billion range). Moreover, this determination is made difficult by the presence of dissolved organic matter (DOM), a competing binding phase for HOCs in natural waters (Means and Wijayaratne 1982), including sediment interstitial or “porewater” (Brownawell and Farrington 1986). Because separation of freely dissolved from colloidal and particulate HOC fractions is exceedingly difficult, little data is available and measurements representing total aqueous phase HOC concentrations are much more common. Unfortunately, this latter parameter can be of limited utility in assessing bioavailability of HOCs in highly impacted, urban sediments containing elevated levels of soot carbon. Solid phase microextraction is a passive sampling technology (Arthur and Pawliszyn 1990) that has shown promise for ultralow level detection of HOCs in aqueous media. Moreover, SPME is selective for the freely dissolved fraction in complex aqueous matrices (Mayer et al. 2000a). As evidence of this phenomenon, in situ samplers incorporating hydrophobic PDMS coated SPME fibers (Zeng et al. 2004) were deployed for several weeks to detect subparts per trillion concentrations of DDT in coastal seawater (Zeng et al. 2005). Concentrations determined by this sampler agreed well with independently measured, operationally defined dissolved phase measurements. Based on this concept, a mass balance and HOC partitioning model predicted that the minimum sediment volume (V s ) required to maintain non-depletive conditions for a SPME based porewater sampler was independent of HOC concentration in both sediment and porewater, and that relatively small sediment volumes (<10 ml) participate in exchange equilibria among the solid, aqueous and SPME phases (Yang et al. 2007a). This work also demonstrated that the sensitivity of such a sampler was proportional to Kow, increased with increasing sorbent volume (Vf), and most importantly, that sub ng/L detection limits were possible. The objective of this study was to develop and test the performance of a compact SPME passive sampler that can measure freely dissolved HOCs of regulatory concern in sediment porewater. With a focus on in situ measurement, an additional requirement for this sampler was to protect the SPME fiber from damage, fouling, and/or sediment contact. To achieve this, a prototype sampler adapted from a previous water column design (Zeng et al. 2004) was developed and fabricated. Key parameters (effect of PDMS coating thickness on sensitivity, time to equilibrium, and agreement with alternative porewater measurements) were then determined for selected prototypes using spiked sediment exposures.