advancing the use of passive sampling in risk assessment...

Advancing the use of passive sampling in risk assessment and management of

contaminated sediments: Results of an international passive sampling ring test

Michiel T.O. Jonker, Stephan van der Heijden, Yongju Choi, Yanwen Wu, Loretta Fernandez, Robert M. Burgess, Upal Ghosh,

Mehregan Jalalizadeh, Jennifer Apell, Phil Gschwend, Rainer Lohmann, Mohammed Khairy, Dave Adelman, Michael Lydy,

Samuel Nutile, Amanda Harwood, Keith Maruya, Wenjian Lao, Amy Oen, Sarah Hale, Danny Reible, Magdalena Rakowska,

Foppe Smedes, and Mark Lampi

Acknowledgments

Cefic-LRi; ECO22 project

Bruno Hubesch

Mark Lampi

ILSI-HESI

Michelle Embry

ECETOC

Malyka Galay Burgos

Various funding agencies participants

Background (1)

• Numerous sediments & soils are contaminated with organic contaminants,

such as PAHs & PCBs

• Current risk assessment based on total, solvent-extractable concentrations

• Not the total, but only the ‘bioavailable’ concentration is available for

uptake in organisms and causing effects

• Improved risk assessment (less false positives) is possible, based on

bioavailable concentrations

Background (2)

• Several methods have been developed for measuring bioavailable

concentrations

• Most attention currently paid to ‘passive sampling’, i.e., using polymer

samplers to determine ‘freely dissolved concentrations’ in sediments/soils

• Passive sampling is a mature technique in science, but not yet fully

accepted in the regulatory community: there is no scientific consensus on

which technique to apply (range of methods available)

• Need for:

• Scientific consensus: comparison study of different methods

• Information on robustness (variability/accuracy)

• Standardization of method(s)

Objectives

• Map the state of the science of passive sampling (performance) in

sediments: Quantify intermethod and interlab variability

• Investigate how any (unacceptable) variation can be reduced

• Recommend standard method(s)

Setup (1)

General setup

• 11 labs participating in ring test; 1 coordinating lab (UU)

• 14 passive sampling formats

• 3 different sediments

• 25 target compounds

Participants

• Established track record in passive sampling with sediments

• Netherlands, Norway, Czech Republic, Korea, USA

Setup (2)

Setup (3)

Passive sampling formats

• Polyethylene (PE): 6 suppliers; 2 thicknesses (25 and 50 µm)

• Polydimethylsiloxane (PDMS): 5 different SPME fibers (suppliers and

coating thicknesses – 10, 30, 100 µm)

• Polyoxymethylene (POM): 2 suppliers and 3 thicknesses (17, 55, 77 µm)

• Polyacrylate (PAc): 30 µm coated SPME fibers

• Silicone rubber (SSP): 100 µm thickness

Compounds

• 13 PAHs (3-6 rings) and 12 PCBs (tri- to heptachlorinated)

• Range in hydrophobicity, partitioning behavior, freely dissolved concs

Setup (4)

Setup (5)

Sediments

3 sediments differing in complexity:

1. Spiked sediment (SP): high concentrations

spiked; low background; sandy; TOC=1.4

2. Field contaminated sediment (Dutch;

Biesbosch area; BB): homogeneous; low

concentrations PAHs and PCBs; TOC=4.3

3. ‘Composed’ sediment (FD): 2 field sediments

mixed.

- French, sandy sediment; low-high PCB

levels (no PAHs)

- Dutch, clayey sediment; moderate PAH

levels (no PCBs); NAPLs (diesel) present;

TOC=2.3

Setup (6)

Experiments

1. ‘Own procedure’: Participants followed their own approach

2. ‘Standardized procedure’: Participants followed standard protocols (UU)

3. Standardized procedure, but extracts analyzed by UU

4. ‘All @ UU’: all 14 formats applied (standardized) and analyzed by UU

5. Additional tests:

- analysis of analytical standard and weighing test (all participants)

- solvent extraction and recovery tests, homogeneity test (UU)

- Partition coefficients (Kpw’s) for all compd’s and polymers (UU)

Results (1)

1. Own procedure (State of the science in passive sampling)

Without PCB-77

Chemical-averaged variation range factor (95% percentile / 5% percentile)

10 29 9

10 9

BB FD SP

All chemicals

Results (2)

1a. Own procedure (Effect of standardizing Kpw’s)

BB FD SP

Without PCB-77

Chemical-averaged variation range factor (95% percentile / 5% percentile)

12 21 10

10 9

All chemicals

Results (3)

2. Standard procedure (Effect of standardizing protocols & Kpw’s)

BB FD SP

Without PCB-77

Chemical-averaged variation range factor (95% percentile / 5% percentile)

7 9 4

4 5

All chemicals

Results (4)

3. Standard procedure, analyzed @ UU (Impact of analytical chemistry)

BB FD SP

Chemical-averaged variation range factor (95% percentile / 5% percentile)

2.4 2.4 2.6

Results (5)

Standard analytical solution

2.8

Averaged variation range factor (95% percentile / 5% percentile)

Results (6)

4. All @ UU (Intermethod variation)

BB FD SP

Chemical-averaged variation range factor (95% percentile / 5% percentile)

1.6 1.7 1.7

Summary (BB sediment)

Intralab / intermethod

Interlab + intermethod

protocol

s

1.6

10 10 4 2.4

Kpw analytics

Conclusions

• Variation in passive sampling results (current practice) is rather (too) large

• Important contribution to the variation by analytical chemistry!

Identification, integration, calibration

• Variation can be significantly reduced by standardizing protocols

Standardization: polymer washing procedures, polymer/sediment ratio,

sediment/water ratio, way and time of mixing, extraction solvent and procedure

• Standardizing Kpw’s does not reduce variation, but is essential for precision of

Cfree

• Different polymers yield very similar results: Intermethod variability is small

(within a factor of 1.6)

• Passive sampling is a robust method - ready for use within regulatory

applications, provided that standard protocols are used and analytical

chemistry is quality controlled