You are here:
Evaluating the precision of passive sampling methods using PRCs in the water column
Citation:
Joyce, A. AND R. Burgess. Evaluating the precision of passive sampling methods using PRCs in the water column. Society of Environmental Toxicology and Chemistry (SETAC) North America 37th Annual Meeting, Orlando, FL, November 06 - 10, 2016.
Impact/Purpose:
This work presents the results of a passive sampling campaign in New Bedford Harbor, MA, USA, in 2014. Passive samplers of several thicknesses were pre-loaded with performance reference compounds and deployed at three stations within or the harbor. Results from this sampling campaign allowed for the measurement of 28 PCB congeners with sub pg/L detection limits and allowed for spatial comparisons between sites. Overall data provided novel insights into processing and interpreting passive sampler and PRC data.
Description:
Low-Density polyethylene (LDPE) sheets are often used as passive samplers for aquatic environmental monitoring to measure the dissolved concentrations of hydrophobic organic contaminants (HOCs). HOCs that are freely dissolved in water (Cfree) will partition into the LDPE until a thermodynamic equilibrium is achieved; that is, the HOC’s chemical potential in the passive sampler is the same as its potential in the surrounding environment. Unfortunately, achieving equilibrium for high molecular weight or highly hydrophobic compounds can take several months or even years. One way to evaluate equilibrium status or estimate the uptake kinetics is by using performance reference compounds (PRCs). PRCs are partitioned into the LDPE prior to deployment and based on the fraction of each PRC lost during deployment, a sampling rate (Rs) or a fractional equilibrium (feq) can be determined for the target HOCs. Assuming equilibrium or applying a model using PRC data to estimate Cfree concentrations may affect the precision of the measurement. In this work, we investigate how using different PRC models in different states of equilibrium affects the measured Cfree concentrations and their precision for a suite of PCBs. To assess these models, four different thicknesses of LDPE passive samplers were co-deployed for 28 days in the water column at three sites in New Bedford Harbor, MA, USA. Following the deployments, the percent of PRC lost ranged from 0-100%. Fractional equilibrium decreased with increasing PRC molecular weight as well as sampler thickness. These data allow Cfree comparisons to be made in two ways: (1) comparing Cfree derived from one thickness using different models and (2) comparing Cfree derived from the same model using different thicknesses of LDPE. Overall, a total of 27 PCBs (log KOW ranging from 5.07 – 8.09) were measured at Cfree concentrations varying from 0.05 pg/L (PCB 206) to about 200 ng/L (PCB 28) on a single LDPE sampler. Relative standard deviations (RSDs) for total PCB measurements using the same thickness and varying model types range from 0.04-12% and increased with sampler thickness. Total PCB RSD for measurements using the same model and varying thickness ranged from: 6 – 30%. No RSD trends between models were observed but RSD did increase as Cfree decreased. These findings indicate that existing models yield precise and reproducible results when using LDPE and PRCs to measure Cfree.