Moni OC Env

Monitors of Organic Chemicals
in the Environment
Monitors of Organic Chemicals
in the Environment
Semipermeable Membrane Devices
James N. Huckins
Columbia Environmental Research Center
Center, Missouri, USA
Jimmie D. Petty
Columbia Environmental Research Center
Center, Missouri, USA
Kees Booij
Royal Netherlands Institute for Sea Research
Den Burg, The Netherlands
James N. Huckins Jimmie D. Petty, Ph.D.
USGS Columbia Environmental USGS Columbia Environmental
Research Center Research Center
4200 New Haven Road 4200 New Haven Road
Columbia, Missouri 65201 Columbia, Missouri 65201
jhuckins@usgs.govc jdpetty@usgs.gov
Kees Booij, Ph.D.

Royal Netherlands Institute for Sea
Research
P.O. Box 59
1790 AB Den Burg
The Netherlands
booij@nioz.nl
Cover photos: (Left) Deployment of SPMDs under Antarctica ice, courtesy of

Carl Orazio, USGS. (Right) Exposure chambers with passive samplers deployed in
Australian grasslands, courtesy of Don Butler, Australian Environmental Protection
Agency. The photo above the title is a standard lipid-containing SPMD, courtesy of
Randal Clark, USGS (see Figure 1-1, p. 18, for a complete description).
Library of Congress Control Number: 2005933715
ISBN-10: 0-387-29077-X
ISBN-13: 978-0387-29077-5
Printed on acid-free paper.

C 2006 Springer Science+Business Media, LLC
All rights reserved. This work may not be translated or copied in whole or in part without the
written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring
Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews
or scholarly analysis. Use in connection with any form of information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed is forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms,
even if they are not identified as such, is not to be taken as an expression of opinion as to
whether or not they are subject to proprietary rights.
Printed in the United States of America. (TB/MVP)
9 8 7 6 5 4 3 2 1
springer.com
This book is dedicated to the memory of Jon Lebo. Jon was integrally involved in
many aspects of the development and especially in the field deployment of SPMD
technology. His untimely death cut short a very productive research career. Jon
was a good friend and an excellent colleague, and we will all miss his insight and
research contributions.
Foreword
Modern, industrialized societies depend on a wide range of chemical substances

such as fuels, plastics, biocides, pharmaceuticals and detergents for maintaining
the high quality lifestyle to which we aspire. The challenge is to ensure that while
we enjoy the benefits of these substances, their inevitable release into our biosphere
does not result in unwanted human and ecosystem exposures, and the risk of ad-
verse effects. One response to this challenge has been the extensive effort to detect
and analyze or monitor a multitude of chemicals in a variety of environmental
media, especially toxic organic compounds in air, water, soils and biota. The con-
ventional monitoring strategy of sampling liters or kilograms of the environmental
medium followed by analytical determination of the quantity of chemical in the
sample extract has been the successful cornerstone of investigative environmental
chemistry. No doubt, it will continue to be so. An extensive literature on these
traditional techniques has evolved over the years.
In parallel with conventional techniques, and I believe entirely complemen-
tary to them, a variety of in situ sensing systems have been developed which
operate on the principle of the preferential partitioning of contaminants into a de-
vice, often at concentrations which are large multiples of environmental levels.
Advocates point out that these partitioning devices have the advantage of integrat-
ing chemical concentrations over a prolonged period, thus “averaging” ambient
levels. Their high partition coefficients can yield significant quantities of analyte
and reduce problems arising from short-term pulses of concentration and from
sample contamination. They can be less expensive, require less sample work-up
and can be deployed more widely. They may or may not approach thermodynamic
equilibrium or equi-fugacity with the media they sense, thus interpretation of the
vii
viii Foreword
partitioned quantities can be challenging but possible when appropriate methods

are applied. Proponents have demonstrated ingenuity in designing, modifying and
exploiting the valuable features of these partitioning devices, not only to sense
our environment by absorbing chemical from it, but also by using them to deliver
chemical in controlled quantities in laboratory settings such as bioassays.
Accordingly, as this technology has matured and the literature has expanded,
the need has arisen for a comprehensive and authoritative review of these devices,
the principles on which they operate and the practice of using them. Fortunately and
appropriately, Huckins, Petty and Booij have taken on this task and share with the
reader their long experience in designing, using, and interpreting the performance
of these devices. In this volume, all relevant aspects of these devices are addressed,
thus the reader will find it an invaluable source of information and insight. It
therefore adds significantly to the environmental literature by supplementing the
many texts concerning traditional chemical analysis.
In closing, it is satisfying to note two themes that permeate this work. The
authors have been notable for their willingness to share their expertise and enthu-
siasms with the wider community of environmental scientists and managers. The
result has been a continuing evolution of a variety of partitioning devices in an
innovative, open and constructive atmosphere in which the over-riding goal has
been to improve and maintain environmental quality. Finally, it is satisfying and
perhaps ironic to see acknowledgment that these ingenious artificial devices have
their ancestry in observations of natural bioconcentration of contaminants in plants
and animals. The thermodynamic or partitioning mechanisms that have resulted
in the regrettable contamination of plants and animals, and often in their demise,
are now being exploited to protect them and us. We owe a debt of gratitude to the
authors for the rigor with which they present the science and technology of these
devices and their sensitivity to the need for environmental protection to which this
book contributes.
DON MACKAY
Peterborough, Canada
Preface
The complexity of the modern world is often beyond the explicit understanding
of any individual. Much of this complexity stems from the technology inherent in
our lifestyle, resulting in a standard of living undreamed of by some futurists. The
standard of living, in terms of comfort, convenience, health and safety is however,
not without its cost.
The quality of life mankind has come to expect often comes with a cost to the
environment, which includes the adverse effects of chemical contaminants. These
contaminants are global in nature and are of increasing concern. The sources of
anthropogenic pollution are legion and all too often the release of contaminants
into environmental systems is considered an unavoidable cost of development. As
a result, many areas of the global environment are under stress from a broad array
of chemicals, both waterborne and airborne.
Because of the potential of these chemicals to adversely affect organisms
in diverse ecosystems and ultimately humans, numerous resource management
and regulatory agencies globally require high quality data defining the presence,
identities and potential biological consequences of exposure to environmental con-
taminants. During the past 50 years, successful control measures have been im-
plemented for many well known contaminants. Substitutes are now being used for
many toxic contaminants, e.g., organochlorines pesticides such as DDT, but the
overall number and variety of chemicals used by modern societies has increased,
while measures to stem their inadvertent release into environmental systems have
not been fully successful. Thus, the need continues for new technologies and tech-
niques to provide reliable data for assessing the potential threats associated with
low levels of increasingly complex mixtures of environmental contaminants.
ix
x Preface
The authors have been intimately involved in conducting research to address

many aspects of environmental contaminants for about three decades. Historically,
samples of environmental matrices, particularly water and air have been collected
at narrow windows of time (i.e., minutes or several hours) which are not represen-
tative of the exposure experienced by organisms. Consequently, we initiated the
development of what would ultimately be the semipermeable membrane device
(SPMD). The SPMD has subsequently proven to be an effective passive sampler
for a wide range of hydrophobic contaminants in multiple media. To date, there
are more than 180 peer reviewed publications in the open scientific literature,
where SPMDs are used for a variety of applications. Some of these publications
are critical of the use of passive samplers for certain applications. However, con-
structive criticism has greatly aided in defining information gaps and limitations
of the passive sampling approach.
Clearly, SPMDs are becoming a mature technology, with increasing global
acceptance as an effective and reproducible passive sampling system for individual
contaminants or complex mixtures. However, new applications continue to be
described and analytical methodology continues to be refined. The growing interest
in the passive sampling approach and its many potential applications ensures that
SPMDs and other passive samplers represent a fertile research area for scientists
involved in studies targeting the presence or effects of environmental contaminants.
It is the intent of the authors to provide a general introduction to passive
samplers and a detailed description of SPMD technology. In short, this work is
intended as a guide or handbook for users and managers faced with contaminant
issues. We address the topical areas of study design, field deployment, sample
processing and residue enrichment, analysis of accumulated chemicals, models for
determining ambient environmental concentrations of target compounds, bioassay
of SPMD extracts, quality control/quality assurance approaches, and selected case
studies describing the results of field deployments of SPMDs. Furthermore, the
reader will find that a number of aspects of the technology await additional research
and development to fully utilize the potential of SPMDs to address contaminant
issues.
It is our sincere hope that this book will provide not only details and explana-
tions for those scientists interested in applying SPMDs and other passive samplers,
but that it also serves as a springboard for new research to expand and enhance the
field of passive sampling. Also, it is important that the reader realize that many of
the techniques and mathematical models presented herein apply to other passive
samplers as well. We are confident that in the future, managers and regulators will
increasingly realize the utility of SPMDs and other passive samplers for addressing
site specific and global contaminant issues.
Acknowledgments
“What a person thinks on his own without being stimulated by the thoughts and
experiences of other people is even in the best case rather paltry and monotonous.”
ALBERT EINSTEIN
The authors gratefully acknowledge the many people that have contributed to the
development and application of the SPMD technology. These include environmen-
tal chemists, biologists, engineers and technical specialists who have contributed
their ideas and efforts. We also acknowledge the numerous agencies and scien-
tific organizations worldwide who have demonstrated their faith in our work by
granting funds for the research and development of SPMD technology.
The authors acknowledge the support of the U.S. Geological Survey and the
Royal Netherlands Institute for Sea Research during the research and development
of SPMDs and during the writing of this book. Also, the research in this book
represents the combined efforts of many other colleagues and we are very much
in their debt. In particular, we thank Randal Clark and Lynne Johnson for their
efforts to meet format requirements of this manuscript and Don Mackay for the
Foreword. Last but not least, the authors recognize the patience of their families
for enduring the diversion of their attention during the writing of this book.
xi
Contents
1. Introduction to Passive Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 The Need for Passive In Situ Samplers . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Passive Sampler Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3 Topics Covered in Subsequent Chapters . . . . . . . . . . . . . . . . . . . . . 23
1.4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2. Fundamentals of SPMDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.1 SPMD Description and Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.2 Applicability of SPMDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.3 Accumulation of Chemicals by SPMDs . . . . . . . . . . . . . . . . . . . . . . 36
2.4 Passive Sampler Fundamentals and Terminology . . . . . . . . . . . . . . 38
2.5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3. Theory and Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.1 Uptake Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.2 Kinetic and Equilibrium Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.3 Dissipation of Performance Reference Compounds (PRCs) . . . . . . 50
3.4 Potential Effects of Dissolved Organic Carbon (DOC)
on SPMD Calibration Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.5 SPMD-Water Partition Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.6 Water Sampling Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.7 Pore Water Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
xiii
xiv Contents
3.8 Groundwater Samplng . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

3.9 Air Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4. Study Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.2 Sources of SPMDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.3 Evaluation of Components and Preparation of SPMDs . . . . . . . . . 88
4.4 Specifications of the Standard SPMD . . . . . . . . . . . . . . . . . . . . . . 89
4.5 Pre-Exposure Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.6 Storage, Transport and Retrieval . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.7 Deployment Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5. Analytical Chemistry Related to SPMDs . . . . . . . . . . . . . . . . . . . . . . 101

5.1 Analytical Speed, Selectivity and Quantitation Limits . . . . . . . . . 101
5.2 Quality Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.3 Sample Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.4 Potential Interferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
5.5 Instrumental Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
5.6 Data Format and Comparability . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
5.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6. Bioassay of SPMD Extracts or Diluents . . . . . . . . . . . . . . . . . . . . . . . 121

6.1 Overview and Rationale of the SPMD-Toxicity
Screening Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.2 Microtox and Mutatox . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
6.3 Mixed Function Oxidase-7-Ethoxyresorufin-O-Deethylase
(MFO-EROD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
6.4 Neurotoxicity Endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
6.5 Endocrine Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
6.6 Exposure of Laboratory Animals to SPMD Extracts . . . . . . . . . . . 132
6.7 Overview of Additional Assays . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
6.8 Potential Interferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
6.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
7. Comparisons to Biomonitoring Organisms . . . . . . . . . . . . . . . . . . . . 139

7.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
7.2 Implications of Selected Models Used for SPMDs
and BMOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
7.3 Comparison of SPMDs and BMOs . . . . . . . . . . . . . . . . . . . . . . . . 144
Contents xv
7.4 Relative Amounts Accumulated . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

7.5 Independence of Concentration Factors Relative to
Exposure Concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
7.6 Similarity of Elimination or Equilibration Rate Constants . . . . . . 155
7.7 Comparability of Estimated and Measured BCFs or BAFs . . . . . . 158
7.8 Dietary Uptake and Biomagnification . . . . . . . . . . . . . . . . . . . . . . 160
7.9 Do SPMDs Qualify as Biomimetic Samplers? . . . . . . . . . . . . . . . . 160
7.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
8. Selected Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

8.1 Review of SPMD Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
8.2 Air-Water-Microlayer Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . 169
8.3 Sampling Indoor Air . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
8.4 Further Comparisons of Bivalves and SPMDs . . . . . . . . . . . . . . . . 173
8.5 Estimation of Exposure to Dioxin-Like Compounds
Using Sediments, Caged Fish, and SPMDs . . . . . . . . . . . . . . . . . . 175
8.6 Using SPMDs in Conjunction with Bioassays and
Chemical Analysis to Characterize Toxic Pollutants . . . . . . . . . . . 177
8.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Appendix A:. SPMD Calibration Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

A.1 SPMD-Water Partition Coefficients . . . . . . . . . . . . . . . . . . . . . . . . 183
A.2 Water Sampling Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
A.3 Air Sampling Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
A.4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
Appendix B:. SPMD Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Chapter 1
Introduction to Passive
Sampling
1.1. THE NEED FOR PASSIVE IN SITU SAMPLERS
Tens of thousands of chemicals are in commercial production throughout the

world and the total number increases each year (Mackay et al., 1992a). Unfortu-
nately, significant amounts of many of these anthropogenic chemicals are released
into terrestrial and aquatic ecosystems. Organic compounds represent the largest
portion, in terms of numbers of these chemicals. The distribution and fate of con-
taminants in environmental systems is controlled by three factors: 1) physicochem-
ical properties of the chemical; 2) environmental conditions (e.g., hydrodynamics,
pH, and solar radiation) and physicochemical properties of abiotic components
(e.g., organic content of soil and sediment), and 3) the composition, mass, and
physiological, anatomical and behavioral characteristics of the species inhabiting
exposed ecosystems. Many studies and programs worldwide are designed to mon-
itor for the presence of environmental contaminants and to determine trends in
their distribution and concentrations. In some cases, data from these studies and
programs are used for guidance on chemicals which are of sufficient concern to
warrant further study.
To aid in assessing the risks associated with large numbers of environ-
mental contaminants, quantitative-structure-activity relationships (QSARs) have
been developed covering nearly all biological effects or other endpoints in both
aquatic and terrestrial species (Connell, 1990). QSARs relate chemical structural
1
2 Chapter 1
characteristics or measured physiochemical properties to endpoints such as bio-

concentration factors (BCFs; i.e., the ratios obtained by dividing the equilibrium
concentrations of chemicals in an organism by the concentrations of the same
chemicals in the surrounding medium, where residue uptake is due to water or
air alone) or toxicity. Mackay et al. (1992a) have defined a subset of QSARs as
quantitative-structure-property relationships (QSPRs). QSPRs use chemical struc-
ture alone to predict contaminant physicochemical properties, partitioning behav-
ior, fate, and the tendency to be accumulated in organism tissues. QSPRs can also
be used to reveal likely errors in data representing measured values of physico-
chemical properties, which enhances the quality of data used for QSAR estimates.
Both QSARs and QSPRs are widely accepted as screening tools for prioritizing
research efforts on environmental contaminants. Some QSPRs can be used to es-
timate the most likely routes of chemical exposure. For example, the dissolved
phase is the most likely route of aquatic organism exposure for compounds with
log octanol-water partition coefficients (K ow s) ≤ 6 (Connell, 1990). Similarly, the
vapor phase is the most likely route of exposure for many terrestrial organisms,
when log octanol-air partition coefficients (K oa s) of organic compounds are ≤8.5
(McLachlan, 1999). In general, these predictions have been borne out by labora-
tory and field studies (Connell, 1990; McLachlan, 1999), but there are exceptions
(Mueller, 2004).
The development of multi-media mathematical models (MMMs) by Neely
(1980), Mackay and Paterson (1981), Mackay et al. (1992a, 1992b, 1997) and
others (Klein and Schmidt-Bleek, 1982; Schnoor et al., 1982; Rand and Petro-
celli, 1985) has provided a more holistic means to model chemicals released into
the environment. MMM models permit simultaneous estimation of the transport,
distribution, fate and bioconcentration (uptake from water alone) or bioaccumu-
lation (uptake from water and diet) of contaminants in multiple environmental
compartments. Furthermore, the development of the “fugacity” (i.e., escaping ten-
dency) approach by Mackay and Paterson (1981) provides a realistic mechanism
by which diverse environmental data can be modeled. In general, QSPRs (Mackay
et al., 1992a) provide adequate data for screening chemicals with these models.
The media modeled may include air and associated particulates, water, soil, sus-
pended and benthic sediments, and biota. Use of MMM models beyond chemical
screening requires knowledge of the characteristics of relevant components in the
compartments, and the magnitude of contaminant inputs or residue concentrations
in one or more compartments. Frequently, the quantities of chemicals entering
an environmental system are unknown. Thus, validation of mathematical models
generally requires direct measurements of the concentrations of target compounds
in one or more environmental medium. In cases where concentrations vary tempo-
rally or equilibration times are long, multiple sampling through time is necessary.
Although QSARs, QSPRs, and MMMs generate screening data, which pro-
vide a much-improved focus on the risks associated with environmental contam-
inants, organism exposure must be confirmed by direct determination of contam-
inant identities and measurement of concentrations in water or air or estimated
Introduction to Passive Sampling 3
by indirect measurements (e.g., via sorbents). At this juncture, a definition of

exposure and related terms is instructive. In the broadest sense, exposure is de-
fined as the concentration of a chemical in the media processed by organisms of
concern (Rand and Petrocelli, 1985), where chemical contact is largely controlled
by factors external to the species of concern. More specifically, exposure is defined
as the amount of chemical residues directly contacting an organism’s membranes
capable of residue exchange, i.e., skin, gills or lungs and gut. Landrum et al. (1994)
describe this level of exposure as the chemical “encounter-volume rate”, which
can be given as L g−1 d−1 of the exposure medium. The “bioavailability” of a
chemical is also a key determinant in chemical exposure. In an environmental
context, bioavailability can be viewed as the ratio of a chemical’s uptake rate (L
g−1 d−1 ) divided by the encounter-volume rate (Landrum et al., 1994). Using this
definition, environmental bioavailability is equivalent to the combined inputs from
skin absorption, gill or lung extraction and gut assimilation. Finally, dose is defined
as the molar concentration of a chemical at a specific site of toxicological action.
Because the dose of a chemical to an organism is often difficult to determine (an
exception may be the use of critical body residues for narcotic acting chemicals),
most risk assessments rely on measurements or measurement-based estimates of
chemical exposure.
Clearly, the choice of analytical methods used for measuring environmen-
tal contaminant concentrations potentially affects mathematical model validation,
the quality and relevance of environmental monitoring data, and the outcomes of
environmental risk assessments. Although method accuracy and precision are im-
portant, other factors should be considered as well. Based on the example QSPRs
given earlier, the method selected should be capable of discriminating between the
relevant (bioavailable) fractions of chemicals from the total amounts of chemicals
present in environmental compartments. Furthermore, in cases where chemicals
bioconcentrate and their environmental concentrations vary temporally, determi-
nation of time-weighted average residue concentrations provides a more complete
picture of organism exposure than those concentrations measured in single or a few
grab samples. In light of the importance of environmentally relevant concentration
data for the assessment of chemical exposure, a brief review of commonly used
sampling or monitoring methods is appropriate.
1.1.1. Active Sampling Methods for Water and Air

Active sampling techniques represent the most widely used approach for the
collection and extraction or trapping of contaminant residues in water. “Active
sampling” refers to those methods that require physical intervention or external
energy input for sample collection and/or residue extraction or trapping. Often,
samples are excised or removed (i.e., grab sampling) from the exposure medium
before residues are extracted. Traditionally, grab samples of water were nearly al-
ways extracted with organic solvents (i.e., liquid-liquid extraction, LLE). Using the
LLE approach, the data generated are limited to the total chemical concentration
4 Chapter 1
in all waterborne phases, which includes microorganisms and algae, particulate

organic carbon (POC), dissolved organic carbon (DOC), inorganic particulates,
and the dissolved phase. Thus, the fraction of the total waterborne residues rep-
resented by the dissolved phase (i.e., the most readily bioavailable phase) is not
distinguished from generally less bioavailable residues in other waterborne phases.
To reduce the use of organic solvents and to discriminate between aqueous residues
associated with POC-DOC and the more bioavailable dissolved phase, solid-phase
extraction (SPE) systems were developed. This method is based on the percola-
tion or pumping of water or air samples through columns, tubes or cartridges of
sorbents consisting of polymeric phases bound to silica cores or various types
of polymeric beads or foam. The SPE approach also includes Empore extraction
disks (Kraut-Vass and Thoma, 1991). These disks or membranes consist of a Teflon
fibril network loaded with SPE sorbents, but the particle size of these sorbents is
smaller than those described for standard SPE columns. Although methods using
SPE cartridges or columns and Empore disks are categorized as active sampling,
the sample extraction involves diffusional and partitioning-sorptive steps. Thus, at
some fundamental level active sampling involves passive (defined later) processes.
In cases where water is turbid, samples are generally filtered through glass
fiber filters (GFF) prior to percolation through sorbents. This step recovers water-
borne particulates and microorganisms with average diameters >0.7 µm, which are
analyzed separately. However, chemicals associated with colloid-sized particulates
and DOC are not removed by GFFs.
More recently, solid-phase micro extraction (SPME) fibers have gained
widespread acceptance as equilibrium samplers for the extraction of water samples
(Arthur and Pawliszyn, 1990). The SPME fiber consists of a small-diameter fused
silica fiber coated with one of several polymeric phases for the sorption of analytes.
The polymeric-film thickness for commercially available SPMEs generally ranges
from 7 to 100 µm with the total phase volume of a 1 cm segment being 0.028 to
0.612 µL. Traditionally, SPME methods are only applied to excised samples and
stirring is used to expedite analyte extraction or times to equilibrium. Also, analyte
concentrations in SPMEs generally represent the total residues in a water sample
(Mayer, 2003). The approach has a number of advantages over active sampling
methods, which include the elimination of pre-filtration and organic solvent ex-
traction steps, typically required for sample preparation, and the direct injection
of the total sample into a gas chromatograph (GC) for analysis.
Nearly all grab sampling methods suffer from potential problems with sample
preservation such as losses due to volatilization, sorption to container walls, and
chemical degradation. For water samples, some of these problems can be avoided
by the addition of an appropriate “keeper” solvent immediately after sample collec-
tion. However, the use of SPEs and SPMEs is thereby precluded and the ability to
discriminate residue distribution among water-borne phases is lost. Data from grab
samples and other active sampling methods represent only a single point or small
window in time, which does not account for temporal variations in contaminant
concentrations at study sites. Thus, adequate assessment of organism exposure
requires labor-intensive multiple sample collections. Generally, the volume of col-

lected samples is limited to ≤5 L, due to the difficulty in the handling, transport,
processing, and extraction of large amounts of water. Consequently, method quan-
titation limits (MQL) may not be adequate for the analysis of trace (≤ 1 µg L−1
or mg m−3 ) or ultra-trace (≤1 ng L−1 or µg m−3 ) levels of hydrophobic organic
chemicals (HOCs). These relatively low quantitation limits are especially needed
for assessing the environmental significance of HOCs that bioconcentrate (uptake
from water by respiration or skin absorption) or bioaccumulate (uptake via skin
absorption, respiration, and diet), and for some highly toxic organic compounds,
which may or may not bioaccumulate.
To overcome the limitations of grab sample preservation and method sen-
sitivity, in situ large-volume SPE systems, equipped with submersible pumps,
have been developed (e.g., the Infiltrix Column System by AXYS Environmental
Systems, Sidney, BC, Canada) for sampling aquatic environments. An alternative
method for ultra-trace analyses of HOCs in water is the Goulden large sample
extractor developed at the Canada Centre for Inland Waters (Forbes and Afghan,
1987). Detection limits are much lower for these types of systems (Rantalainen
et al., 1998), but significant concerns still exist about sample contamination, an-
alyte losses to exposed surfaces, filter plugging in turbid waters, the use of toxic
chlorinated solvent, and procedurally mediated changes in the distribution of con-
taminants among phases constituting environmental waters. For many studies or
programs involving multiple sites, in situ large-volume samplers are often too labor
intensive and costly to use at all sites. Thus, simultaneous replication of sampling
for statistical purposes is seldom performed.
Methods used for the active sampling and extraction or concentration of
organic vapors in air are generally related to those described for water. For ex-
ample, samples of volatile organic compounds (VOCs) and semi-volatile organic
compounds (SVOCs) are often collected and concentrated by in situ pumping
of air through SPE tubes or cartridges. Even when grab sampling is the method
of choice (e.g., VOC sample collection in Summa-polished canisters and Tedlar
bags), SPEs are often used for sample preconcentration or trapping prior to in-
strumental analysis. In all cases, GFFs are used when discrimination between the
vapor and particulate phases is needed to estimate the relative contributions of
the two-exposure pathways (e.g., SVOCs with log K oa s > 8.5). The SPE sorbents
used to concentrate vapors of trace to ultra-trace levels of SVOCs in large vol-
umes of air include polyurethane foam plugs, Tenax and XAD-2 resin (Ockenden
et al., 1998). These sorbents are also used for the extraction of ultra-trace levels of
dissolved-phase waterborne residues (Rantalainen et al., 1998).
Many of the shortcomings listed for active sampling of waterborne residues
apply to the analysis of VOC and SVOC vapors in air. Also, sampling analytes in an
equivalent mass of air and water requires about a 103 larger volume of air, thereby
practically limiting the applicability of grab sampling for the analysis of trace
and ultra-trace SVOCs in air. Furthermore, because sampling sites for assessing
global-atmospheric transport of contaminants are often in rugged terrain in remote
6 Chapter 1
locations, the choice of sampling methods may be limited by the size, weight and
portability of the sampling apparatus and the need for electrical power.
In summary, active sampling and extraction or trapping methods provide rea-
sonably reliable information on the total waterborne and airborne concentrations of
HOCs, but only during one point in time or a relatively brief interval of time, which
is in marked contrast to the exposure duration of most organisms. Most of these
methods permit some discrimination between analyte concentrations in material
trapped by the filter, representing the total residues associated with POC, inor-
ganic particulates, and microorganisms (e.g., algae and spores), and dissolved and
vapor phase residue concentrations in filtrates. Unfortunately, the potential effects
of sampling, transport and filtration on the environmental distribution of contam-
inant residues among the various phases in water and air are difficult to predict
a priori. For example, solutes and vapors can adsorb on GFFs and be misidentified
as part of the particulate phase, while residues associated with fine particulates can
desorb and be incorrectly identified as part of the dissolved phase (Mackay, 1994).
Thus, residue concentrations in filtrates are not necessarily representative of the
dissolved or vapor phases in the undisturbed sample media. With the exception of
programmable in situ active sampling systems (e.g., the Infiltrix water sampling
system and high volume [HiVol] air samplers) sample size may not be adequate
for the analysis of trace and ultra-trace levels of contaminants. Also, active sam-
pling methods (excludes the Infiltrix system) are generally relevant only to a few
points in time and do not provide time-weighted average (TWA) concentrations.
Strictly speaking, measurement of TWA concentrations during a specified time
period requires continuous, additive extraction (i.e., integrative sampling, where
the extraction medium acts as an infinite sink) of an exposure medium. TWA
concentration data are useful indicators of organism exposure to HOCs.
1.1.2. Biomonitoring Organisms for Water and Air

Biomonitoring organisms (BMOs) are often used for chemical risk assess-
ments because they address many of the limitations related to the biological
relevance of data from point in time analytical-based sampling methods. In
particular, residues accumulated in BMO tissues were, by definition, bioavail-
able. Also, many HOCs are highly concentrated in fatty tissues of organisms by
the process of bioconcentration or bioaccumulation. Unfortunately, certain phys-
iological, anatomical and behavioral characteristics specific to the BMO species
used and site-exposure conditions can affect the magnitude and variability of HOC
concentrations accumulated in tissues (Livingstone et al., 1985; Barron, 1990;
Huebner and Pynnönen, 1992; Gobas et al., 1993; Goudreau et al., 1993; Gilek
et al., 1996; Björk and Gilek, 1997; Moring and Rose, 1997; Baumard et al., 1998a,
1998b; Axelman et al., 1999; Wang and Fisher, 1999; Baussant et al., 2001, Gray,
2002). For example, investigators have found that bioaccumulation factors (BAFs;
i.e., the ratios obtained by dividing the equilibrium concentrations of chemicals
in an organism by the concentrations of the same chemicals in the surrounding
medium, where residue accumulation is based on water or air and the diet) of HOCs
in many organisms are affected by residue metabolism (Livingstone et al., 1985;
Moring and Rose, 1997), food ration (Björk and Gilek, 1997), size of the organism
(Gilek et al., 1996), and toxic stress (Huebner and Pynnönen, 1992; Goudreau et al.,
1993). This data suggests that BCFs and BAFs may be site- and species-specific,
and do not necessarily represent thermodynamic equilibrium, where residues in
tissues are proportional to environmental exposure concentrations (Huckins et al.,
2004). Thus, steady state is often a more appropriate descriptor of constant tissue
concentrations than equilibrium. Furthermore, the concentrations of HOC residues
in tissues of BMO species may not be representative of tissue concentrations in
species of concern (Gray, 2002). Altogether, these complications limit the appli-
cability of aquatic BMOs for the extrapolation of exposure concentrations, and for
determining HOC sources and concentration gradients.
Terrestrial BMOs have also been widely used for monitoring environmen-
tal contaminants. In particular, the lipid-like waxy cuticle layer of various types
of plant leaves has been used to monitor residues of HOCs in the atmosphere.
However, some of the problems associated with aquatic BMOs apply to terrestrial
BMOs as well. For example, Böhme et al. (1999) found that the concentrations
of HOCs with log K oa s < 9 (i.e., those compounds that should have attained
equilibrium) varied by as much as 37-fold in plant species, after normalization of
residue concentrations to levels in ryegrass (Lolium spp.). These authors suggested
that differences in cuticular wax composition (quality) were responsible for this
deviation from equilibrium partition theory. Other characteristics of plant leaves
may affect the amount of kinetically-limited and particle-bound HOCs sampled
by plant leaves but to a lesser extent (i.e., <4-fold), these include age, surface area,
topography of the surface, and leaf orientation.
1.2. PASSIVE SAMPLER DEVELOPMENT
In this work, we define “passive” samplers as human-made devices where

sample collection and residue extraction occur simultaneously in a completely
passive manner. The sampling or concentration process is mediated by the dif-
fusion of chemicals from a matrix where chemical fugacity or potential is high
to a matrix (receiving medium or sorbent) where chemical fugacity or potential
is low. Three requirements must be met to derive reasonable estimates of ambi-
ent concentrations of analytes from their concentrations in a passive sampler: 1)
concentrations in the device must be proportional to environmental concentrations
and the associated rate constants for chemical exchange and partition coefficients
must be independent of ambient analyte concentrations; 2) calibration data (rate
constants and partition coefficients) applicable to site conditions must be available
for target compounds; and 3) the sampling process should not significantly reduce
analyte concentrations in the medium sampled. Also, if samplers are calibrated
in the laboratory or by the use of mathematical models, the potential effects of
8 Chapter 1
site-specific exposure conditions must be taken into account. Unfortunately, ef-

forts to validate the accuracy of estimates of environmental concentrations based
on passive samplers are often impeded by the lack of an accurate independent
method for measuring trace or ultra-trace residues of analytes in environmental
media. Furthermore, the discrimination of dissolved and vapor phase residues from
those associated with particulates, aerosols, micelles or macromolecules can be
problematic for both active and passive sampling methods.
In the following sections we highlight only selected works that have con-
tributed toward the further development of passive samplers for SVOCs and/or
HOCs. The literature related to the development and use of passive samplers for
monitoring gases or VOCs in occupational environments is large. However, these
publications are discussed only briefly, because lipid-containing semipermeable
membrane devices (SPMDs) are primarily designed for SVOCs.
1.2.1. Air
The first passive samplers were small personal diffusional monitors (PDMs).
These devices were developed in the early seventies and were designed to deter-
mine occupational exposure to VOCs in air using linear uptake kinetics. PDMs
contain a sorbent or reactive material separated from the sampled medium by a
rate limiting air filled diffusional zone or a rate limiting semipermeable membrane
(Fowler, 1982). The sorbent or reactive material acts as an infinite sink during an
exposure. Analytes are accumulated in an integrative manner because no signifi-
cant losses of accumulated residues occur during an exposure, regardless of any
decreases in ambient concentrations. This performance characteristic permits the
determination of TWA analyte concentrations. The American Conference of Gov-
ernmental Industrial Hygienists (ACGIH) has accepted PDMs as the best available
technology to gauge human exposure to most VOCs in occupational environments
(ACGIH, 1990). In particular, PDM-derived TWAs provide the most satisfactory
way of determining the compliance of the work atmosphere to threshold limit
values of VOCs (ACGIH, 1990).
More recently, Harner et al. (2003) coated ethylene vinyl acetate (EVA) onto
glass (polymer coated glass [POG]) for use as fugacity sensors or equilibrium
samplers of SVOCs in indoor and outdoor air. The EVA film thickness was 1.1 and
2.4 µm depending on the application and as expected, SVOC sorption capacity
and times to equilibrium were shown to be directly proportional to film thickness.
The clearance capacity (E v ; volume of sample medium cleared of chemical) of a
sorbent for an analyte is given by
E v = K pa Aδp = K pa Vp (1.1)
where K pa is the polymer-air partition coefficient, A is the surface area, δ p is film

thickness and Vp is the volume of the polymeric phase. Equation 1.1 indicates
that sorption capacity for vapor-phase samplers is the volume of air cleared of
vapor-phase chemical for a specified volume of sorbent or partitioning phase, and

that for a fixed A, E v is directly proportional to both film thickness and K pa .
For a 1 µm film thickness, the surface-area (A; cm2 )-to-sorbent-volume (V ;
cm )-ratio (AV −1 ) of the Harner et al. (2003) device is 1 × 104 cm−1 with a total
3
sorbent volume of 2.9 µL. Earlier, Wilcockson and Gobas (2001) devised a POG
with a 0.05 µm EVA film thickness, but it was not used for air sampling. Assuming
first-order kinetics, times to 95% of equilibrium (t95 ) are given by
t95 = − ln 0.05/(ka /K pa δp ) = 2.99/ke (1.2)
where ka is the mass transfer coefficient for the air boundary layer (ABL) and ke is
the release or elimination rate constant. The effect of the magnitude of compound
K pa on time to equilibrium is underscored by the finding that polychlorinated
biphenyl (PCB) congeners 28 and 153 required 7 and 217 days (d), respectively,
to reach equilibrium with the 1 µm thick EVA coated POG. The ABL was shown
to control the uptake rates of SVOC vapors and thus wind speed and turbulence
largely mediates exchange kinetics. To minimize flow-induced variability in ex-
change kinetics, Harner et al. (2003) designed and applied a deployment device
that dampened flow differences. Clearly, the use of these samplers can be extended
to outdoor air, assuming adequate analyte mass can be sampled for quantitation of
trace to ultra-trace levels of HOC vapors.
Coutant et al. (1985) first extended the application of small PDM-like devices
to VOCs in soils. Also, Zabik et al. (1992) first reported the development of passive
samplers for SVOCs in soil. Unlike PDMs for VOCs, the device of Zabik et al.
(1992) contained XAD-4 resin or C18 enclosed in a polymeric Whirlpak Bag made
of low-density polyethylene (LDPE). Vapors with moderate- to high-K oa s were
sampled in an integrative manner during exposures of 21 d. Also, analytes were
recovered by solvent elution of sorbents, as opposed to thermal desorption often
used for PDMs. The amount of chemical in the LDPE was not measured. Johnson
et al. (1995) further developed this approach utilizing C18 alone in Whirlpack
bags for sampling SVOC vapors in soils. Again only the residues in the sorbent
were measured. They were able to accurately estimate areal distribution of a PCB
mixture in soil (concentrations varied by 4-orders of magnitude) at a hazardous
waste site from PCB concentrations in deployed samplers. Although Johnson et al.
(1995) noted that sampling rate (in this case, µg PCBs g−1 C18 ) was inversely
proportional to moisture content; no analysis of the rate-limiting step in vapor
sampling was performed. The exchange kinetics of SVOC vapors or solutes with
this type of sampler are more complicated than with the POG design, as both the
membrane and the sorbent phases are involved and analytes must diffuse through
the membrane and desorb as vapors from the inside wall of the bag (i.e., partial
pervaporation) to reach the sorbent.
Shoeib and Harner (2002) and Wania et al. (2003) separately developed large
capacity passive samplers for integratively monitoring the atmospheric transport
of HOCs. The sorbents used in these devices act as an infinite sink for HOC
vapors, and have been used earlier to actively sample large volumes of air and
10 Chapter 1
water (i.e., HiVol samplers). The Shoeib and Harner (2002) device is based on
the use of a polyurethane foam (PUF) disk, while the Wania et al. (2003) device
uses XAD-2 resin in a finely perforated stainless steel column. We are unable to
compute the AV −1 ratio of the sampler because the area for chemical exchange
could not be determined with the available information. The Wania et al. (2003)
device showed a steady uptake of HOC vapors in outdoor air over the course of
one year and the rate of HOC uptake was controlled by the effective thickness of
the boundary layer associated with the sorbent. Similar to Harner et al. (2003), a
deployment device was used which dampens air flow-turbulence, minimizing the
effects of these variables among sites. However, field calibration of the same set of
chemicals at three different sites showed that apparent sampling rates (Rs ; m3 or
cm3 d−1 cleared of analyte) varied by two to as much as 5-fold for α-HCH. Wania
et al. (2003) suggested that some of the observed variation in Rs values among
sites was due to problems associated with the HiVol samplers such as analyte
breakthrough of the PUF sorbent at sites with higher temperatures and differences
in sampling time scales.
1.2.2. Water
In 1980, Byrne and Aylott (1980) were the first to patent a simple device
that passively sampled organic contaminants from water. The device consists of
a reservoir of nonpolar organic solvent separated from water by selected “non-
porous” polymeric membranes. The word nonporous refers to polymeric films or
membranes, where solutes essentially dissolve into rubbery or amorphous regions
of the polymer, as there are no fixed holes (other than defects) in the matrix for
diffusive transport. The membranes used in the Byrne and Aylott device included
cellulose, vinyl chlorides, polyvinylidene fluoride, and polytetrafluoroethylene.
No publications other than the patent are known to exist for the use of this de-
vice. Södergren (1987) first reported the development and testing of an in situ
mimetic (mimics key elements of complex biological processes in simple media)
passive sampling device for the accumulation of waterborne HOCs. This device
consisted of 3 mL of hexane sealed in a hydrophilic regenerated cellulose dialysis
bag (A = 12.6 cm2 ). The AV −1 ratio for the device was 4.2 cm−1 . The cellu-
lose membrane bag had a molecular-weight cutoff of 1000 Daltons, which allows
only dissolved phase HOCs to concentrate in the hexane. Field-testing showed
that PCBs were accumulated in the hexane and that the hexane could be directly
injected into a gas chromatograph equipped with an electron capture detector (GC-
ECD) for sample analysis. Using the Södergren device, the estimated linear uptake
rate for several HOCs (i.e., DDTs and PCBs) over a 7-d exposure period was about
0.05 L d−1 . Unfortunately, the stability of cellulose membranes is poor in
warm (i.e., >20 ◦ C) turbulent aquatic environments, which may be caused
by its fragility and/or the presence of microorganisms with cellulase enzymes.
Prest et al. (1992) suggested that the very polar hydrophilic nature of cellu-
lose reduces the permeability of HOC solutes relative to nonpolar membranes
such as nonporous LDPE. Sabaliūnas and Södergren (1996) found that iden-
tical size (A ≈ 25 cm2 ) cellulose and LDPE membrane tubes, containing a
4-mL mixture of cyclohexane and the lipid triolein (AV −1 of both devices is
6.2 cm−1 ), exhibited marked differences in their uptake rates of HOCs. More
specifically, the HOC uptake rates of the solvent-containing LDPE bags were
24 to 84-times higher than the solvent containing cellulose bags (Sabaliūnas and
Södergren, 1996), but the loss of cyclohexane was much greater from LDPE bags.
This finding confirmed that the cellulose dialysis membranes (thickness not spec-
ified) tested had much greater resistance to mass (HOC solutes) transfer when
compared to 75 µm thick LDPE membranes.
In an effort to optimize the solvent-containing passive sampler design,
Zabik (1988) and Huckins (1988) evaluated the organic contaminant permeability
and solvent compatibility of several candidate nonporous polymeric membranes
(Huckins et al., 2002a). The membranes included LDPE, polypropylene (PP),
polyvinyl chloride, polyacetate, and silicone, specifically medical grade silicone
(silastic). Solvents used were hexane, ethyl acetate, dichloromethane, isooctane,
etc. With the exception of silastic, membranes were <120-µm thick. Because
silicone has the greatest free volume of all the nonporous polymers, thicker mem-
branes were used. Although there are a number of definitions of polymer free
volume based on various mathematical treatments of the diffusion process, free
volume can be viewed as the free space within the polymer matrix available for
solute diffusion.
The criteria used for membrane evaluation in these studies included organic
solvent compatibility, durability, HOC uptake rates, and cost. The results of the
solvent-polymer compatibility tests ranged from no apparent impact on membrane
properties to dissolution. For example, hexane-filled silastic tubing swelled to
greater than twice its original volume, but returned to its original volume in a
matter of minutes in air (i.e., hexane pervaporated) and in a few hours in water.
Of the polymers tested, LDPE and PP demonstrated the best overall compatibility
with organic solvents (Huckins et al., 2002a). Subsequent to the above experimental
work, Huckins et al. (1990a) suggested the use of Hildebrand and Hansen solubility
parameters (i.e., cohesive energy density based numerical values that indicate the
relative tendency of a specific solvent to solvate a material such as a polymer) as a
screening tool for matching of appropriate combinations of polymer and solvent.
An extensive list of solubility parameter values and an in-depth discussion of
solubility parameter theory and use are available (Grulke, 1989).
Even when using LDPE and PP, the diffusive losses of most nonpolar sol-
vents may be unacceptably large for AV −1 designs that exceeded 1 cm−1 . This is
especially true at higher exposure temperatures because both solute diffusion and
polymer free volume increase with temperature (Comyn, 1985). Even when sol-
vent losses were not excessive, uptake of HOCs by membrane-enclosed solvents
appeared to become curvilinear well before thermodynamic equilibrium was ap-
proached. This phenomenon is likely due to the outward flux of sampler solvent
with elevated HOC levels (relative to water), which appears to facilitate residue
12 Chapter 1
export to the exterior surface of the membrane (Huckins et al., 1990a, 1993; Prest
et al., 1992). This process is similar to a known membrane transport process de-
scribed in the membrane separations literature as facilitated diffusion (Hwang and
Kammermeyer, 1984), where analyte transport through a membrane is enhanced
by the mass transfer of a carrier substance (solvent in this case).
Independently, Hassett et al. (1989) developed the passive in situ concen-
tration and extraction sampler (PISCES), which consists of a chrome-plated
brass reservoir containing 200 mL of hexane or isooctane and dual LDPE mem-
branes having a combined A of about 23 cm2 . This design features a low AV −1
(≈0.1 cm−1 ), which reduces excessive solvent losses. Litten et al. (1993) success-
fully field-tested PISCES for locating the source of PCB inputs into a riverine
system, where most conventional active sampling methods were ineffectual. Lin-
ear uptake (sampling) rates ranged from 0.5 to 0.9 L d−1 for individual devices
and did not differ among sampled PCB congeners. Unfortunately, gas or vapor
buildup inside the reservoir caused bulging of the membranes and potential leaks,
and non-reproducible solvent losses of replicates were noted.
More recently, Kingston et al. (2000) developed a passive sampling sys-
tem (PSS), which consists of several types of diffusion (rate)-limiting membranes
and Empore disks as the analyte sequestration phase. In this study, polysulfone
and polyethylene membranes were used as rate-limiting barriers. Both designs
used a C18 Empore disk as a receiving or sequestration phase. The polysulfone
membrane was best suited for target compounds with log K ow s between 2.0 and
4.0, and a polyethylene membrane was best suited for target compounds with
log K ow s > 4.0. Using these designs, they successfully concentrated chemicals
as polar as atrazine and chemicals as hydrophobic as PCB congener 153. Be-
cause the PSS is not limited to the membranes listed above or only one type
of Empore disk, sampling of a wide range of chemicals is possible. Alvarez
et al. (2004) designed a polar organic chemical integrative sampler (POCIS)
for a broad spectrum of hydrophilic organic chemicals. The POCIS is suited
for polar pesticides, pharmaceuticals, hormones, or organic compounds with log
K ow s < 4.0. The device is commonly used in conjunction with SPMDs to enable
a more holistic assessment of the presence of organic contaminants in aquatic
systems (Petty et al., 2004). POCIS units are constructed by forming a membrane-
sorbent-membrane sandwich, which maximizes the surface area for chemical up-
take. The membrane is polyethersulfone and the sorbent used is a triphasic admix-
ture of polymeric resins and a small amount of activated carbon, or Oasis HLB.
The choice of sorbent depends on the classes of chemicals targeted but is not nec-
essarily limited to the two types listed. Table 1.1 provides additional details on
both the PSS and the POCIS.
Huckins (1989) first evaluated the use of LDPE strips (PESs) alone as passive
samplers. The PESs (75 µm thick; A = 400 cm2 ) were exposed to three concentra-
tions (0.8, 8.2, and 61 ng L−1 ) of 14 C-2,2 ,5,5 -tetrachlorobiphenyl (TCB) for 28 d
in flow-though, constant concentration exposures. Replicate PESs (n = 4) were
collected on day 1, 3, 7, 14, 21 and 28. Considerable biofouling was observed on
TABLE 1.1 Comparison of Passive Sampler Characteristics and Applications for Organic Compounds
SPMDs SPMEs POGs LDPE strips PISCES POCIS PSSs
integrative equilibrium equilibrium integrative integrative integrative integrative

Classification (log K ow s > 4.5) (log K ow s < 6) (logK oa s < 10) (log K ow s > 5)
surface water, surface water, surface water, air surface water surface water surface water, surface water
Sample media groundwater, air groundwater, groundwater, groundwater, groundwater, air,
sediment, soil sediment, soil air, sediment sediment sediment, soil
Chemicals HOCs HOCs HOCs HOCs HOCs HPOCsa HOCs,

Introduction to Passive Sampling
sampled (SVOCs) (VOCs, SVOCs) (SVOCs) (SVOCs) (SVOCs) HPOCs
Sampling rate EBLb EBL EBL EBL EBL EBL diffusive

control (log K ow s > 4.5) (log K ow s >?) (log K ow s > ?) (log K ow s >?) membrane
A/V c 90 cm−1 1,400 cm−1 104 cm−1 230 cm−1 0.1 cm−1 150 cm−1 110 cm−1
(91.4 × 2.5 cm) (7 µm film) (1 µm film) (85 µm film)
A 460 cm2 0.4 cm2 , 7 µm 290 cm2 , 460 cm2 23 cm2 41 cm2 16 cm2
(91.4 × 2.5 cm) film × 10 cm 6.8 × 7.0 cm (91.4 × 2.5cm)
Sorbent-liquid 5,000 µL 0.28 µL 29 µL 2,000 µL 2.0 × 105 µL 240 µL 145 µL C18

phase volume 7 µm film 1 µm film 85 µm film
Clearance 560 L (w) 0.027 L (w) 16 L (w) 79 L (w) 1.3 × 105 ? 75 L (w)
capacityd 360 m3 (a) 18 L (a) 10.4 m3 (a) L (w)
a
Hydrophilic organic chemical.
b
EBL is external boundary layer, water or air.
c
A /V is the surface area (cm2 ) of the sampler divided by the volume (cm3 ) of the accumulating phase.
d
Volume of water or air cleared (E v ) by a specific sampler configuration at equilibrium. We use PCB congener 52 in this example.
13
14 Chapter 1
PESs by the end of the exposure. Differences in exposure concentrations did not
affect the sampling rates of PESs, which indicate that these devices obey first-order
uptake kinetics. The sampling rate of 14 C-2,2’,5,5’-TCB by PESs (4.8 L d−1 ) was
similar to that observed for 1 mL triolein SPMDs, with the same surface area.
By using a PES with a different thickness, one can conveniently change
the AV −1 ratio. This approach permits some control over the time required to
reach equilibrium concentrations. Bartkow et al. (2004) has reported an excellent
example of the impact of AV −1 ratio or thickness on the time to equilibrium.
These investigators showed that a 200 µm thick PE sheet took twice as long to
reach equilibrium in air as a 100 µm thick PE sheet. In theory, changing membrane
thickness will not affect polymer diffusivity and equilibrium membrane-water
partition coefficients (K mw s) or solubility coefficients (Sp ). However, in practice
different values of K mw , K ma (membrane-air partition coefficient) and membrane
diffusivity may be obtained from films of different thickness because of changes
in polymer chain orientation and crystallinity (Pauly, 1989). For example, Rogers
(1985) has shown that Sp , which is given as the K mw when Henry’s Law convention
applies, is dependent on polymer crystallinity as shown by
Sp = Sa φa (1.3)
where Sa is the solubility coefficient for a completely amorphous or rubbery poly-
mer with due consideration of surface and void defects and φ a is the amorphous
volume fraction of the polymer. In particular, this relationship has been shown to
hold true for PE. Also, some variations in these parameters are expected for LDPE
films of the same thickness from different manufactures due to a lack of uniformity
in fabrication conditions. Recently Booij et al. (2003a) have shown that PES parti-
tion coefficients are affected by temperature (2 to 30 ◦ C) while temperature effects
on SPMDs are insignificant over the same range of temperatures. Thus, the reliabil-
ity of environmental concentration estimates from polymeric film-based samplers
is contingent on the availability of reproducible polymer, and a means to estimate
the effects of biofouling, flow-turbulence, and temperature on the times to equilib-
rium. Booij et al. (2002) have made progress in this area by developing a method
to spike PESs with performance reference compounds (PRCs; see discussion on
PRCs in Section 1.2.3.1. and Chapter 3). PRCs are analytically non-interfering
compounds with moderate- to relatively high-fugacities, which are added to a pas-
sive sampler (e.g., the lipid of SPMDs) prior to deployment. The rate of PRC loss
during an exposure can be used to estimate in situ sampling rates of analytes of
interest.
We also examined the use of medical-grade silicone or silastic tubing (ST)
alone as a passive sampler medium (Huckins and Petty, 1994a). The ST had the
following dimensions: 30.5 cm long, a wall thickness of 0.24 mm, an outside
diameter of 1.9 mm, and an A of 18.7 cm2 . Replicate STs were exposed to 8.4 ng
L−1 of 14 C-2,2’,5,5’-TCB for 7 d in a flow-though, constant concentration system.
STs (n = 3) were collected on day 2, 5, and 7. The STs contained about 82% of
the amount of 14 C-2,2’,5,5’-TCB concentrated in lipid-containing STs of the same
dimensions (i.e., silastic SPMDs; see Section 1.2.3.1.).
According to the polymer literature (Flynn and Yalkowsky, 1972; Rogers,

1985), silicone has the greatest free volume of the commonly available nonporous
polymers. However, partition coefficients and diffusion coefficients for this type
of polymer may not necessarily be independent of chemical concentration, i.e.,
Henry’s law convention may not apply (Rogers, 1985) when ambient environ-
mental HOC concentrations are very high (e.g., oil spills). As HOC solute con-
centrations in the silicone rise, free volume increases more rapidly than in other
polymers, which at some undefined point, results in increases in both polymer
diffusivity and partition coefficients. The Flory-Huggins equation likely fits this
type of behavior. Under membrane control, one would observe increasing sam-
pling rates with rising solute concentrations, while under water boundary layer
(WBL; a thin hydrodynamically complex region separating the SPMD membrane
from the bulk water, where molecular diffusion dominates mass transfer and re-
sistance to mass transfer) control, analyte sampling rates would remain constant
for extended periods unless extensive biofouling occurs. Fortunately, this type of
scenario is unlikely for trace levels of environmental contaminants. Furthermore,
unlike sorbents used in some samplers, residues are readily recovered by soaking
silicone or silastic in the appropriate solvent.
In regard to the potential effects of silicone membrane or film thickness
(e.g., polydimethylsiloxane [PDMS]) on partition coefficients, Paschke and Popp
(2003) have shown that that at equilibrium an SPME fiber with a 7 µm thick
film of PDMS had about a 6-fold higher K pw than a similar fiber with a 100 µm
thick film. However, this could be the result of interactions with the silica core.
Recent research (Smedes, 2004) has shown that silicone sheeting with PRCs can
be employed for water sampling with good results.
As suggested earlier, SPMEs are essentially equilibrium sampling devices but
are typically used in an active sampling mode. However, Górecki and Pawliszyn
(1997) and Mayer et al. (2000) have used SPMEs in an in situ passive extraction
mode in the laboratory, which is described by Mayer et al. (2000) as matrix-SPME.
In this case, sediment was collected from the field and sieved prior to SPME
exposures. The goal was to estimate the concentrations of a number of HOCs
in field-sediment pore water using SPMEs. The approach relies on non-depletive
extraction, where the total residues extracted by an SPME at equilibrium do not
significantly change the concentrations of the sampled medium (i.e., pore water
and sediment particles). Thus, equilibrium concentrations of analytes in the fiber’s
polymer can be readily related to their pre-exposure concentrations in sediment
pore water by a simple equilibrium partition coefficient.
Because SPMEs are equilibrium samplers similar to POGs, the thickness of
the fiber’s polymeric phase must be kept relatively small to ensure that times to
equilibrium are reasonable (e.g., ≤1 month) for high K ow analytes (log K ow s > 6),
under conditions of minimal turbulence. Also, to ensure non-depletive extraction
of analytes from excised sediments, the sorption capacity of the organic carbon in
the excised sediment sample should be at least 50-fold greater than the sorption
capacity of the SPME polymeric phase (i.e., m oc K oc K pw Vp ; where m oc rep-
resents the mass of the organic carbon, K oc is the equilibrium sediment organic
16 Chapter 1
carbon-water partition coefficient, p represents the SPME polymeric phase, and

K pw is the equilibrium polymer-water partition coefficient). The δ p of the 7 cm
PDMS fiber used by Mayer et al. (2000) was 15 µm and the AV −1 ratio was
670 cm−1 . The total volume of the fiber’s PDMS phase was 0.71 µL. For com-
parison purposes, the δ p values of commercially available SPMEs generally range
from 7 to 100 µm and the sorbent phase volumes of a 1 cm segment are generally
<0.7 µL. It is interesting to note that SPMEs with 15 µm film thickness appear
to reach steady state with high K ow compounds in sediments in less than 30 d
with some agitation, while 1 µm thick POGs failed to reach steady state with high
K oa vapors (log K oa > 9.3) in 100 d. The findings of Booij et al. (2003b) offer a
potential explanation for this difference in times to equilibrium. They found that
sampling rates for PESs (70 µm film thickness) were much higher in sediment
slurries (≈0.1 cm s−1 flow at the polymer surface) than in stagnant sediments. The
authors proposed that resistance to solute uptake posed by the WBL is greatly re-
duced by contaminant desorption from particles (i.e., POC) very near or in contact
with the polymer. It seems likely that upon close contact or collision of parti-
cles with the PESs or the SPME sorbent, both the WBL around the polymeric
films and the WBL around the particles are effectively thinned, thus resistance
to mass transfer is reduced (Eq. 3.9). Contaminant desorption is induced by the
depleted levels of these solutes at the exterior surface of the sorbent phase (i.e.,
the concentration gradient across the LDPE or SPME boundary layer). Further-
more, computations by Mackay (1994) indicate that chemicals in small particles
(≤1 mm) desorb rapidly, reaching equilibrium in milliseconds to minutes for com-
pounds with log particle-water partition coefficient ranging from 1 to 6. Even
though this analysis assumed that desorption was dominated by the facile por-
tion of the desorption isotherm, it appears quite likely that particle desorption
contributions may greatly reduce times to equilibrium in sediment slurries.
Although SPMEs, and POGs are useful tools for many sampling scenarios,
their strengths for one application can be the source of their limitations for an-
other application. This characteristic applies to all sampling devices designed to
accumulate compounds with a broad range of physicochemical properties. For ex-
ample, the low sorbent phase volumes and high surface areas of thin-film SPMEs
and POGs enable the attainment of equilibrium in relatively short exposure times,
but the volumes of water cleared of chemical are relatively low compared to sys-
tems with higher sorbent volumes. Using a 7 cm SPME fiber coated with a 15 µm
thick polymeric phase (0.7 µL V ) to sample a compound with a log K pw of 6
permits the extraction of only 0.7 L of water at equilibrium. Using a 1 µm thick
POG with a total A of 290 cm2 (2.9 µL V ) to sample the same compound with
a log K pa of 8, permits the extraction of only 0.3 m3 of air at equilibrium. These
sample volumes may be inadequate for the quantification of ultra-trace analytes.
Although the 1 µm POGs, as described by Harner et al. (2003), would be expected
to reach equilibrium more rapidly and have a greater capacity, the method detec-
tion limits may not be as low as those described for the 15 µm SPMEs. This is
because the entire SPME sample is injected into the GC or the instrument used
for analyte quantification. However, the potentially lower method detection limits
of SPMEs are contingent on minimal co-extracted interferences and on the abil-
ity of the GC column to resolve different classes of chemicals without additional
chromatographic separations.
Because high AV −1 ratios of SPMEs and POGs also mean rapid desorption
of accumulated residues, samplers must be frozen at sub-zero temperatures and
transport and storage times must be minimized. Mayer et al. (2000) analyzed
SPMEs by GC within 20 s of sample removal in the matrix-SPME study. To enable
rapid turn-around times for sample data, and to minimize transport and storage
problems, Górecki and Pawliszyn (1997) developed a field-portable SPME/fast GC
system for sampling VOCs. However, at more remote sites the need for electrical
power is problematic.
An inherent problem for all passive samplers designed to attain equilibrium
is the very wide range of K ow s and K oa s of the broad array of organic compounds
of interest. However, a significant body of evidence suggests that K pw s and K pa s
of large, very nonpolar molecules appear to be less than their respective K ow s
and K oa s (Huckins et al., 1990a, 1993; Lefkovitz et al., 1994; Harner et al., 2003;
Paschke and Popp, 2003). Chiou (1985) has discussed one potential factor that may
account for this phenomenon, which relates to the effect of a decrease in the size
difference between a macromolecular phase and solutes with relatively large molar
volumes. Even if K pw s and K pa s are less than the corresponding K ow s and K oa s,
equilibrium values of target analytes may vary as much as 6-orders-of-magnitude.
Table 1.1 compares key aspects and performance characteristics of selected
passive samplers, including the triolein-containing SPMD. Of the eight devices
examined, only a few appear to have overlapping functions. Clearly, no one device
can provide the desired data for all exposure scenarios.
1.2.3. Lipid-Containing SPMDs and Closely Related Devices

Similar to the previous section, we discuss only selected works to highlight the
development of SPMDs. Also, we include some discussion of several unpublished
pilot studies (Huckins, 1989) that influenced our early development of SPMDs.
These pilot studies were directed solely toward sampling the aqueous phase. The
first application of SPMDs for sampling organic vapors did not occur until several
years later (Petty et al., 1993). To our knowledge, only SPMDs, PESs and SPMEs
are being applied in both air and water, because the use of many passive samplers
is limited to a specific medium and exposure scenario.
1.2.3.1. Aqueous Phase Sampling

Based on earlier work (Lieb and Stein, 1969; Chiou, 1985; Södergren, 1987;
Zabik, 1988) Huckins et al. (1989, 1990a, 1993) first developed and tested two types
of lipid-containing semipermeable membrane devices (SPMDs) for in situ passive
sampling of bioavailable dissolved aqueous-phase HOCs. The lipid-containing
18 Chapter 1
FIGURE 1.1 A standard lipid-containing SPMD with three molecular welds near each end. Note
the low interfacial tension causes intimate contact (i.e., the presence of a lipid film on the
membrane interior surface) between the triolein and the membrane even where air bubbles
exist. Reprinted with permission from the American Petroleum Institute (Huckins et al., 2002a).
SPMDs consisted of 99% triolein in layflat-LDPE tubing (Figure 1.1) and 99% tri-
olein in small-bore ST. The following tests were conducted in these studies: 1) a 28 d
flow-though, constant concentration exposure of LDPE triolein-containing SPMDs
to three concentrations (0.8, 8.2, and 61 ng L−1 ) of 14 C-2,2 ,5,5 -TCB; 2) a 7 d flow-
though, constant concentration exposure of silastic triolein-containing SPMDs to
8.4 ng L−1 of 14 C-2,2 ,5,5 -TCB; and 3) a 21 d static exposure of LDPE SPMDs
(three types of lipid) to 14 C-2,2 ,5,5 -TCB (single application). The goals of these
tests were to determine the relative uptake rates of LDPE and silastic SPMDs, the
independence of LDPE SPMD concentration factors to exposure concentrations,
and the steady-state partition coefficients between each of the three types of SPMD
lipid used and the water.
The reasons for the selection of high purity triolein as a test lipid in these
pilot studies are given in Chapter 2. Two other types of lipid were examined in this
pilot study as well. Grass carp (Ctenopharyngodon idella) lipids (liquid phase)
were obtained by dichloromethane-extraction of whole-body grass carp tissues,
and were used as a representative for complex lipid mixtures found in fish and
other organisms. The phospholipid lecithin (MW ≈ 787 Daltons, ≈85% purity,
wax at room temperature) was used as a representative of organisms such as algae.
Also, lecithin enclosed in an SPMD membrane may provide a reasonable surrogate
for estimating partitioning of HOCs between phospholipids in biomembranes and
biological fluids. In light of current research (Leslie et al., 2002), this type of data
is needed for assessing the potential impacts of narcotic-acting HOCs because
toxicity is only elicited when biomembrane concentrations attain or exceed the
critical body threshold.
The layflat-LDPE tubing used for making SPMDs was 2.6 cm wide and had a
wall thickness of about 75 µm. The A of the 0.5 mL triolein SPMDs used in these
flow-through tests was about 200 cm2 with an AV −1 (lipid + membrane) ratio of
about 80 cm−1 . The A of the 1 mL triolein SPMDs used in the static exposures
was about 360 cm2 (i.e., AV −1 = 72 cm−1 ). The surface area of the 1 mL lipid
SPMDs was about 20% less than the currently used 1 mL triolein SPMDs. The
silastic SPMDs contained 0.5 mL triolein and the AV −1 (membrane + lipid) was
22 cm−1 . Replicate SPMDs (n = 4) used in the 28 d flow-though tests (well wa-
ter) were collected at d 1, 3, 7, 14, 21 and 28. Some biofouling was observed
on SPMDs near the end of the exposure, but less than that observed on similarly
exposed PESs, as discussed earlier. The uptake rates of 14 C-2,2 ,5,5 -TCB by the
SPMDs was about 4.2 L d−1 (Huckins, 1989). This uptake rate is 4.7- to 8.8-fold
greater than the PISCES, but the difference between SPMDs and PISCES is less
than predicted on the basis of the relative surface areas (i.e., SPMDs have about
a 10-fold greater surface area). Greater than 50% of the total 14 C-2,2 ,5,5 -TCB
residues accumulated in SPMDs at 28 d were present in the triolein (i.e., triolein-
membrane partition coefficient [K Lm ] ≈ 4) even though triolein represented only
about 20% of the sampler mass. Differences in exposure concentrations did not
affect the sampling rates of SPMDs, which indicates that these devices obey first-
order exchange kinetics. Based on an equivalent weight of SPMD and PES (see ear-
lier discussion), the estimated steady state concentrations of 14 C-2,2 ,5,5 -TCB in
triolein-containing SPMDs would be about two-fold greater than those in the PES.
Lipid-containing silastic SPMDs (n = 3) used in 7 d flow-through tests (well
water) were collected on days 1, 3, and 7. After normalizing the surface area of
the silastic SPMDs to that of the standard design LDPE SPMD used in the 28 d
exposures (7 d, 8.2 ng L−1 treatment), it was found that the 7-d silastic SPMDs
accumulated about 1.4 times as much 14 C-2,2 ,5,5 -TCB as the corresponding 7-
d LDPE SPMDs. Surprisingly, the K Lm for the silastic SPMD configuration at
7 d was 16, which is about 4-fold higher than the LDPE SPMD configuration.
Because the same flow-through exposure system was used for both the LDPE
and silastic SPMD exposures and test temperatures were identical, this disparity
between silastic SPMD and LDPE SPMD uptake rates is likely due to one or
both of the following factors: 1) differences in flow-turbulence at the LDPE and
silastic membrane surfaces, which affect the effective thickness of the WBL; and 2)
analytical error. In light of more recent research indicating that small variations in
flow-turbulence at the membrane surface can affect SPMD sampling rates (Vrana
and Schüürmann, 2002), the most likely factor causing the observed differences
in sampling rates is the effective thickness of the WBL.
Finally, LDPE SPMDs with grass carp lipid were exposed for 21 d to
14
C-2,2 ,5,5 -TCB, 14 C-3,3 ,4,4 -TCB, 14 C-mirex and 14 C-fenvalerate, whereas
SPMDs with triolein or lecithin were exposed to only 14 C-2,2 ,5,5 -TCB. After
21 d, the largest mass fraction of these test chemicals (14 C-mirex was an exception)
was in the triolein. The 14 C-2,2 ,5,5 -TCB log triolein-water partition coefficient
was 6.01, whereas the 14 C-2,2 ,5,5 -TCB partition coefficients for the grass carp
lipid-water and lecithin-water systems were 30% and 35% lower, respectively.
Comparison of these data to literature log K ow s of 2,2 ,5,5 -TCB showed that
the partition coefficients for the grass-carp lipid and the lecithin were not signifi-
cantly different from median values reported for the log K ow of 14 C-2,2 ,5,5 -TCB.
However, the partition coefficient of 2,2 ,5,5 -TCB in triolein and water in direct
20 Chapter 1
contact (Chiou, 1985) is only 41% of the membrane enclosed triolein-water parti-
tion coefficient derived from SPMDs (Huckins et al., 1990b). A possible explana-
tion for the discrepancy in these partition coefficients is that the SPMD membrane
maintains true binary integrity of the triolein and water phases during tests due
to the low permeation of the relatively high molecular weight triolein and polar
hydrogen-bonding water molecules through LDPE. Note that direct mixing of wa-
ter with triolein during partitioning results in a triolein layer with about 11% water
at 25 ◦ C, which likely affects the magnitude of the partition coefficients of very
hydrophobic compounds (Chiou, 1985).
Petty and Orazio (1996) developed an interesting variation in SPMD liquid
phases. The approach consisted of both LDPE and silastic tubes containing sili-
cone fluid (50 cSt or 3200 MW) with 3% by weight PX-21 activated carbon. The
presence of the activated carbon enhanced retention of planar molecules such as
PAHs and the silicone fluid remains liquid at temperatures below freezing. How-
ever, the partition coefficients of HOCs for this type of silicone fluid are much
lower than for triolein.
At about the same time that the lipid-containing SPMD was introduced, Huck-
ins et al. (1990b) first described an organic solvent dialysis (OSD) method for the
recovery of HOCs accumulated in intact SPMDs and in the LDPE membrane alone.
Essentially, the process involves submersion of an exposed lipid-containing SPMD
into a glass container with a suitable solvent such as hexane or cyclopentane for
at least 24 hrs. Solvation of the nonporous LDPE-SPMD membrane enhances the
outward diffusive flux of analytes (compared to non-solvated membrane) into the
bulk solvent of the reservoir, whereas the diffusion of triolein through the swollen
polymer is still very slow. By matching the Hildebrand or Hansen solubility pa-
rameters (commonly available in the literature) of a solvent to LDPE or other
nonporous polymers of interest, the amount of polymer swelling and dissolution
can be predicted (Brandrup and Immergut, 1989). The amount of lipid carried over
with analytes of interest during OSD is dependent on the amount of low-molecular
weight impurities (e.g., oleic acid and methyl oleate) in the triolein used in the
SPMD, the solvent used, temperature, and duration of the dialytic procedure. When
performing hexane dialysis of standard SPMDs with ≥95% triolein for ≤48 hr,
only 3–4% of the lipid components are co-dialyzed with the analytes. Interestingly,
analyte residues contained in PES and many other nonporous polymers are readily
recovered by the same method.
OSD also provides an enrichment method for the extraction and separation
of HOCs from tissue or egg extracts with large volumes of lipids (Huckins et al.,
1990a; Meadows et al., 1993). The OSD method has been further advanced by the
work of others (Bergqvist et al., 1998; Strandberg et al., 1998). The bags employed
are made of several sizes of LDPE tubing that are heat sealed at one end and subse-
quently charged with the lipid to be treated by the OSD procedure. Quality control
procedures for the LDPE bags are similar to those used for SPMDs. Application of
the OSD method to silicone or silastic SPMDs or tubes (wall thickness ≈ 500 µm)
with lipid extracts was unsuccessful. Solvation of silicone or silastic membranes
with solvents typically employed in OSD greatly expands the polymeric free vol-
ume, which results in the co-dialysis of most of the triolein (or other lipid) with the
analytes. However, analytes were quantitatively recovered in a matter of minutes
instead of hours (i.e., LDPE) thus future examination of much thicker membranes
and different solvents deserves additional attention.
Until early 1993, lipid-containing SPMDs were only used to determine the
presence and identities of contaminants, relative differences in SPMD concen-
trations among deployment sites, and to compare HOC residues accumulated in
SPMDs to those accumulated in biomonitoring organisms. Extrapolation of SPMD
concentrations to HOC concentration in ambient water requires appropriate math-
ematical models and calibration data relevant to exposure conditions. In 1993,
several mathematical models were developed for SPMDs providing the basis for
the estimation of water concentrations (Huckins et al., 1993). These models con-
tained most of the variables affecting SPMD uptake and elimination rates, but the
rate-limiting step in the residue exchange process was not well defined. Factors
potentially affecting overall mass transfer rates of HOCs include the additive re-
sistances of the membrane, the WBL and any biofilm (i.e., periphyton and any
imbibed detritus) on the exterior surface of the membrane, and the effects of tem-
perature. Huckins et al. (1993) also provided calibration data for a model PCB
and a model polycyclic aromatic hydrocarbon (PAH). However, in this paper the
authors incorrectly hypothesized that diffusion of most HOCs through the SPMD
membrane is the rate-limiting step in HOC exchange. Subsequently, Booij et al.
(1998) demonstrated that resistance in the WBL is the rate-limiting step in the up-
take and elimination of most HOCs by SPMDs. This finding has had considerable
impact on the direction of SPMD research.
While diffusion of HOCs in water is much less complex and more predictable
than diffusion in nonporous polymeric membranes, accounting for the effects of
site-specific variations in flow-turbulence on effective thickness of the aqueous
boundary layer is not straightforward. In light of the number of variables potentially
affecting the sampling rates of this new generation of passive monitoring devices,
the acceptability of their use beyond screening level assessments for dissolved
phase concentrations of HOCs is dependent on the ability of investigators to predict
in situ sampling rates. Recent work by Booij et al. (2003a and 2003b) has greatly
improved our ability to use mathematical models to predict in situ SPMD sampling
rates.
The use of PRCs to determine the effects of SPMD membrane biofouling on
the uptake rates of HOCs was first proposed in 1991 (Huckins et al., 2002a). The
PRC method can also account for the effects of flow-turbulence and temperature
differences (Huckins et al., 2002b). Implicit in the PRC approach is that the over-
all uptake and elimination rates of HOCs in SPMDs are governed by first-order
kinetics and that residue exchange with the sampled medium is isotropic. Isotropic
exchange can be described by analogy as a two-way door (where the door is the
rate limiting step in chemical exchange), with equal resistances to mass transfer
in either direction. Huckins et al. (1993) showed that the ke determined from the
22 Chapter 1
uptake of phenanthrene by replicate SPMDs was within 25% of the ke determined

from the loss of phenanthrene from replicate SPMDs, which is slightly less than the
error propagated by the derivation procedure. However, similar ke s derived from
the uptake and release of a chemical does not necessarily mean that an exchange
process is isotropic (especially in the case of BMOs). Finally, using the strict defi-
nition of isotropic exchange, any change in the linear uptake rate constants ku and
ke should not result in a significant change in the equilibrium partition coefficient
of the sampling matrix.
Prest et al. (1992) first used phenanthrene as a PRC in a field exposure of
SPMDs. The results of their study appeared to be consistent with PRC theory.
However, it was several years later before the validity of the use of PRCs for bio-
fouled SPMDs was further substantiated in controlled laboratory studies (Huckins
et al., 1994b).
Because SPMD membranes often biofoul during extended exposures in warm
surface waters and because this biofilm may impede the uptake rates of HOCs by
unknown amounts, Huckins et al. (1990a) hypothesized that the use of certain
antifouling chemicals spiked into SPMD triolein might reduce the severity of this
potential problem. This hypothesis was contingent on the availability of an envi-
ronmentally safe, non-interfering (i.e., analytically) and effective biocide-biostat
with appropriate LDPE diffusional characteristics (i.e., controlled release). A pilot
experiment (unpublished data) was conducted with a candidate organo-arsenical.
Unfortunately, the results of this experiment suggested that very high concentra-
tions (i.e., almost one percent of the SPMD weight) are necessary for controlling
biofouling. This effort was discontinued due to the hazards associated with high
levels of these chemicals and the potential licensing requirements for this type of
application.
1.2.3.2. Atmospheric Sampling

Petty et al. (1993) were the first to use SPMDs as passive samplers of SVOC
vapors. Standard (subsequently defined) SPMDs were exposed to laboratory air
containing background levels of PCBs. An active sampling method based on pump-
ing air through a GFF and a Florisil column, which is a modification of the Na-
tional Institute of Occupational Safety and Health (NIOSH) Method 5503, was
used to independently monitor air levels during 14- and 28-d SPMD exposures.
The mean temperature during the exposure periods was 26 ◦ C and air around
the samplers was relatively quiescent. The uptake of total PCB vapors was linear
throughout the 28 d exposure period. Based on a 1 mL triolein SPMD configura-
tion, total PCB congeners were sampled at a rate of 4.6 m3 d−1 (i.e., volume of
air cleared of chemical per d). After fourteen d exposure, SPMDs concentrated
total PCB congeners by about 2×104 fold (mass basis) or 2 × 107 (volume basis).
Congener-specific analysis of PCBs concentrated in SPMDs exposed to laboratory
air for 28 d (Petty et al., 1993), indicated that the uptake rates generally increased
with the degree of chlorination. Congeners with fewer ortho-Cls within a homolog
group (e.g., tetrachlorobiphenyls) had higher sampling rates than other isomeric
positions. Sampling rates or Rs s were determined for fourteen of the congeners,
which ranged from 0.9 m3 d−1 for a mono-ortho dichlorobiphenyl (congener 007)
to 9.6 m3 d−1 for a tri-ortho octachlorobiphenyl (congener 201).
These data seem to support resistance to diffusion across the membrane
(LDPE) as a significant factor in sampling airborne PCBs. Under membrane con-
trol, sampling rates increase with membrane-air partition coefficients (K ma s) or
K oa s, whereas under ABL control, sampling rates decrease slightly with increas-
ing molar volume or molecular weight (see Section 3.9.2.). Decreased ortho-Cl
substitution enhances rotational freedom to assume a more planar configuration,
which increases heats of fusion, heats of adsorption (e.g., charcoal) and polymer
permeability, and decreases vapor pressure. Also, the K oa s of PCB congeners in-
crease with the degree of PCB chlorination (Harner and Bidleman, 1996). Because
some tetra- and pentachlorobiphenyls, most hexachlorobiphenyls and more highly
chlorinated congeners have log K oa s > 9, a significant portion of the total residues
is associated with airborne particulates (McLachlan, 1999).
A thin film of oil-like material was visible after 28 d on the exterior surfaces
of the SPMD membrane. Analysis of this film indicated that the triolein impurities,
oleic acid and methyl oleate, were the major constituents. This external lipid film
(Petty et al., 1993) appeared to contain imbibed particulates. Although the film
was removed from the SPMDs by solvent rinsing and analyzed separately, some
lipid-mediated desorption of particle-associated PCBs and subsequent diffusion
into the SPMD may have occurred prior to solvent-removal of the film. This
observation suggests the potential for SPMD concentrations to reflect both vapor
phase concentrations and to a lesser extent, lipid-extracted particulate-associated
residues (see Section 3.9.2.). Unfortunately, concentrations of more chlorinated
congeners in particulates collected on GFFs from the NIOSH method were often
below quantitation limits, because only a small volume of air was sampled (1 m3 )
using this active method.
1.3. TOPICS COVERED IN SUBSEQUENT CHAPTERS
In subsequent chapters, we provide an overview of SPMD fundamentals and

applications (Chapter 2); the theory and modeling which includes the extrapola-
tion of SPMD concentrations to ambient environmental concentrations (Chapter
3); study considerations such as the necessary precautions and procedures during
SPMD transport, deployment, and retrieval (Chapter 4); the analytical chemistry
and associated quality control for the analysis of SPMD dialysates or extracts
(Chapter 5); a survey and brief description of bioassays-biomarkers used to screen
the toxicity of SPMD environmental extracts (Chapter 6); discussions on how HOC
concentrations in SPMDs may or may not relate to similarly exposed biomoni-
toring organisms (Chapter 7); and selected examples of environmental studies
using SPMDs (Chapter 8). In addition, two appendices are included which provide
24 Chapter 1
SPMD calibration data for many HOCs (Appendix A) and sources of additional
information on SPMDs (Appendix B).
1.4. REFERENCES
Alvarez, D.A.; Petty, J.D.; Huckins, J.N.; Jones-Lepp, T.L.; Getting, D.C.; Goddard, J.P.; Manahan, S.E.
2004, Development of a passive, in situ, integrative sampler for hydrophilic organic contaminants
in aquatic environments. Environ. Toxicol. Chem. 23: 1640–1648.
American Conference of Governmental Industrial Hygienists (ACGIH) 1990, Threshold Limit Values
for Chemical Substances and Physical Agents and Biological Exposure Indices; Cincinnati, OH.
Arthur, C.L.; and Pawliszyn, J. 1990, Solid phase microextraction with thermal desorption using fused
silica optical fibers. Anal. Chem. 62: 2145–2148.
Axelman, J.; Næs, K.; Näf, C.; Broman, D. 1999, Accumulation of polycyclic aromatic hydrocarbons in
semipermeable membrane devices and caged mussels (Mytilus edulis) in relation to water column
phase distribution. Environ. Toxicol. Chem. 18: 2454–2461.
Barron, M.G. 1990, Bioconcentration: Will water-borne organic chemicals accumulate in aquatic or-
ganisms. Environ. Sci. Technol. 24: 1612–1618.
Bartkow, M.E.; Hawker, D.W.; Kennedy, K.E.; Müller, J.K. 2004, Characterizing uptake kinetics of
PAHs form the air using polyethylene-based passive air samplers of multiple surface area-to-
volume ratios. Environ. Sci. Technol. 38: 2701–2706.
Baumard, P.; Budzindki, H.; Garrigues, P. 1998a, PAHs in Arcachon Bay, France: Origin and biomon-
itoring with caged organisms. Mar. Pollut. Bull. 36: 577–586.
Baumard, P.; Budzinski, H.; Garrigues, P.; Sorbe, J.C.; Burgeot, T. Bellocq, J. 1998b, Concentrations
of PAHs (polycyclic aromatic hydrocarbons) in various marine organisms in relation to those in
sediments and to trophic level. Mar. Pollut. Bull. 36: 951–960.
Baussant, T.; Sanni, S.; Jonsson, G.; Skadsheim, A.; Børseth, J.F. 2001, Bioaccumulation of polycyclic
aromatic compounds: 1. Bioconcentration in two marine species and in semipermeable membrane
devices during chronic exposure to dispersed crude oil.Environ. Toxicol. Chem. 20: 1175–1184.
Bergqvist, P.-A.; Strandberg, B.; Rappe, C. 1998, Lipid removal using semipermeable membranes in
PCDD and PCDF analysis of fat-rich environmental samples. Chemosphere 38: 933–943.
Björk, M. and Gilek, M. 1997, Bioaccumulation kinetics of PCB 31, 49, and 153 in the blue mussel,
Mytilus edulis L. as a function of algal food concentration. Aquat. Toxicol. 38: 101–123.
Böhme, F.; Welsch-Pausch, K.; Mclachlan, M.S. 1999, Uptake of airborne semivolatile organic com-
pounds in agricultural plants: Field measurements of interspecies variability. Environ. Sci. Technol.
33: 1805–1813.
Booij, K.; Sleiderink, H.M.; Smedes, F. 1998, Calibrating the uptake kinetics of semipermeable mem-
brane devices using exposure standards. Environ. Toxicol. Chem. 17: 1236–1245.
Booij, K.; Smedes, F.; van Weerlee, E.M. 2002, Spiking of performance reference compounds in low
density polyethylene and silicone passive water samplers. Chemosphere 46: 1157–1161.
Booij, K.; Hofmans, H.E.; Fischer, C.V.; van Weerlee, E.M. 2003a, Temperature-dependent uptake rates
of nonpolar organic compounds by semipermeable membrane devices and low-density polyethy-
lene membranes.Environ. Sci. Technol. 37: 361–366.
Booij, K.; Hoedemaker, J.R.; Bakker, J.F. 2003b, Dissolved PCBs, PAHs, and HCB in pore waters and
overlying waters of contaminated harbor sediments. Environ. Sci. Technol. 37: 4213–4220.
Brandrup, J. and Immergut, E.H., Eds. 1989, Polymer Handbook; 3rd ed.; Wiley & Sons: New York,
NY.
Byrne, B.A. and Aylott, R.I. 1980, Concentrator for Removing Organic Materials from Aqueous
Systems. British Patent 1566253.
Chiou, C.T. 1985, Partition coefficients of organic compounds in lipid-water systems and correlations
with fish bioconcentration factors. Environ. Sci. Technol. 19: 57–62.
Comyn, J., Ed. 1985, Polymer Permeability; Elsevier Applied Science LTD: London, England.
Connell, D.W. 1990, Bioaccumulation of Xenobiotic Compounds. CRC Press: Boca Raton, FL.
Coutant, R.W.; Lewis, R.G.; Mulik, J. 1985, Passive sampling devices with reversible adsorption. Anal.
Chem. 57: 219–223.
Flynn, G.L. and Yalkowsky, S.H. 1972, Correlation and prediction of mass transport across membranes
I: Influence of alkyl chain length on flux-determining properties of barrier and diffusant. J. Pharm.
Sci. 61: 838–852.
Forbes, M. and Afghan, B.K. 1987, Analytical Protocol for Monitoring Ambient Water Quality at
the Niagra-On-The-Lake and Fort Erie Stations. Canada Centre for Inland Waters: Burlington,
Ontario, Canada.
Fowler, W.K. 1982, Fundamentals of passive vapor sampling. Am. Lab. 14: 80–87.
Gilek, M.; Björk, M.; NäF, C. 1996, Influence of body size on the uptake, depuration, and bioaccu-
mulation of polychlorinated biphenyl congeners by Baltic Sea blue mussels, Mytilus edulis. Mar.
Biol. 125: 499–510.
Gobas, F.A.P.C.; Zhang, X.; Wells, R. 1993, Gastrointestional magnification: the mechanism of bio-
magnification and food chain accumulation of organic chemicals. Environ. Sci. Technol. 27: 2855–
2863.
Goudreau, S.E.; Neves, R.J.; Sheehan, R. 1993, Effects of wastewater treatment plant efflu-
ents of freshwater mollusks in the upper Clinch River, Virginia, USA. Hydrobiologia 252:
211–230.
Gray, J.S. 2002, Perceived and real risks: Produced water from oil extraction. Mar. Pollut. Bull. 44:
1171–1172.
Grulke, E.A. 1989, Solubility Parameter Values. In Polymer Handbook, 3rd ed.; Brandrup, J., Immergut,
E.H., Eds.; Wiley & Sons: New York, NY; Section 7, pp. 519–559.
Górecki, T. and Pawliszyn, J. 1997, Field-portable solid-phase microextraction/fast GC system for
trace analysis. Field Anal. Chem. Tech. 1: 227–284.
Harner, T. and Bidleman, T.F. 1996, Measurements of octanol-water partition coefficients for poly-
chlorinated biphenyls. J. Chem. Eng. Data. 41: 895–899.
Harner, T.; Farrar, N.J.; Schoeib, M.; Jones, K.C.; Gobas, F.A.P.C. 2003, Characterization of polymer-
coated glass as a passive air sampler for persistent organic pollutants. Environ. Sci. Technol. 37:
2486–2493.
Hassett, J.P.; Force, M.; Song, H. 1989, A Passive In situ Sampler for Organic Compounds in Water.
Extended Abstracts of the 198th National Meeting of American Chemical Society; Miami Beach,
FL; 5: 496–397.
Huckins, J.N. 1988, Solvent and polymer interactions. USGS, Columbia Environmental Research
Center, Columbia, MO; Unpublished work.
Huckins, J.N. 1989, A Novel Approach for Monitoring and Subsequent Cleanup of Nonpolar Con-
taminants in Water. The 103r d Association of Official Analytical Chemists Annual International
Meeting; St. Louis, MO. Abstract 327.
Huckins, J.N.; Tubergen, M.W.; Manuweera, G.K. 1990a, Semipermeable membrane devices contain-
ing model lipid: A new approach to monitoring the availability of lipophilic contaminants and
estimating their bioconcentration potential. Chemosphere 20: 533–552.
Huckins, J.N.; Tubergen, M.W.; Lebo, J.A.; Gale, R.W.; Schwartz, T.R. 1990b, Polymeric film dialysis
in organic solvent media for cleanup of organic contaminants. J. AOAC 73: 290–293.
Huckins, J.N.; Manuweera, G.K.; Petty, J.D.; Mackay, D.; Lebo, J.A. 1993, Lipid-containing semiper-
meable membrane devices for monitoring organic contaminants in water. Environ. Sci. Technol.
27: 2489–2496.
Huckins, J.N. and Petty, J.D.1994a, Use of medical-grade silicone or silastic tubing alone as a passive
sampler medium. USGS, Columbia Environmental Research Center, Columbia, MO; Unpublished
work.
Huckins, J.N.; Petty, J.D.; Orazio, C.E. ; Lebo, J.A.; Clark R.C.; Haverland, P.S. 1994b, A Lab-
oratory Study to Demonstrate the Feasibility of the use of SPMD Permeability Reference
26 Chapter 1
Compounds (PRCs) to Correct for the Effects of Fouling on the Uptake of PAHs. USGS, Columbia
Environmental Research Center, Columbia, MO; Unpublished report to American Petroleum
Institute (API), Washington, DC.
Huckins, J.N.; Petty, J.D.; Prest, H.F.; Clark, R.C.; Alvarez, D.A.; Orazio, C.E.; Lebo, J.A.; Cranor,
W.L.; Johnson, B.T. 2002a, A Guide for the Use of Semipermeable Membrane Devices (SPMDs)
as Samplers of Waterborne Hydrophobic Organic Contaminants; Publication No. 4690; American
Petroleum Institute (API): Washington, DC.
Huckins, J.N.; Petty, J.D.; Lebo, J.A.; Almeida, F.V.; Booij, K.; Alvarez, D.A.; Cranor, W.L.; Clark,
R.C.; Mogensen, B.B. 2002b, Development of the permeability/performance reference compound
(PRC) approach for in situ calibration of semipermeable membrane devices (SPMDs). Environ.
Sci. Technol. 36: 85–91.
Huckins, J.N.; Prest, H.F.; Petty, J.D.; Lebo, J.A.; Hodgins, M.M.; Clark, R.C.; Alvarez, D.A.; Gala,
W.R.; Steen, A.; Gale, R.W.; Ingersoll, C.G. 2004, Overview and comparison of lipid-containing
semipermeable membrane devices (SPMDs) and oysters (Crassostrea gigas) for assessing chem-
ical exposure. Environ. Toxicol. Chem. 23: 1617–1628.
Huebner, J.D. and Pynnönen, K.S. 1992, Viability of glochida of two species of Anodonta exposed to
low pH and selected metals. Can. J. Zool. 70: 2348–2354.
Hwang, S-T. and Kammermeyer, K. 1984, Membranes in Separations. Krieger Publishing: Malabar,
FL.
Johnson, K.A.; Naddy, R.B.; Weisskopf, C.P. 1995, Passive sampling devices for rapid determination
of soil contaminant distributions. Toxicol. Environ. Chem. 51: 31–44.
Kingston, J.K.; Greenwood, R.; Mills, G.A., Morrison, G.M.; Björklund-Persson, L. 2000, Develop-
ment of a novel passive sampling system for the time-averaged measurement of a range of organic
pollutants in aquatic environments. J. Environ. Monit. 2: 487–495.
Klein, A. W. and Schmidt-Bleek, F. 1982, Significance and Limitations of Environmental Compart-
mentalization Models in the Control of New Chemicals Based on the Organization for Economic
Cooperation and Development Minimum Premarketing Set of Data. In Modeling The Fate of
Chemicals in the Aquatic Environment; Dickson, K. L., Maki, A. W., Cairns, J. Jr., Eds.; Ann
Arbor Science Publishers: Ann Arbor, MI.; pp. 73–92.
Kraut-Vass, A. and Thoma, J. 1991, Performance of an extraction disk in synthetic organic chemical
analysis using gas chromatography-mass spectrometry. J. Chromatogr. 583: 233–240.
Landrum, P.F.; Hayton, W.L.; Lee, H., II; McCarty, L.S.; Mackay, D.; McKim, J.M. 1994, Synopsis
of Discussion Session on the Kinetics behind Environmental Bioavailability. In Bioavailability:
Physical, Chemical and Biological Interactions; Hamelink, J.L., Landrum, P.F., Bergman, H.L.,
and Benson W., Eds.; CRC Press: Boca Raton, FL.; pp. 203–219.
Lefkovitz, L.F.; Crecelius, E.A.; Gilfcil, T.J. 1994, The Use of Semipermeable Membrane Devices
to Predict Bioaccumulation of Hydrophobic Organic Contaminants. The 15th Annual meeting of
Society of Environmental Toxicology and Chemistry, Denver, CO; Abstract WE25.
Leslie, A.L.; Oosthoek, A.J.P.; Busser, F.J.M.; Kraak, M.H.S.; Hermens, P.L.M. 2002, Biomimetic
solid-phase microextraction to predict body residues and toxicity of chemicals that act by narcosis.
Environ. Toxicol. Chem. 21: 229–234.
Lieb, W.R. and Stein, W.D. 1969, Biological membranes behave as non-porous polymeric sheets with
respect to the diffusion of non-electrolytes. Nature 224: 240–243.
Litten, S.; Mead, B.; Hassett, J. 1993, Application of passive samplers (PISCES) to locating a source
of PCBs on the Black River, New York, NY. Environ. Toxicol. Chem. 12: 639–647.
Livingstone, D. R.; Moore, M.N.; Lowe, D.M.; Casci, C.; Farrar, S. 1985, Responses of the cytochrome
P-450 monooxygenase system to diesel oil in the common mussel Mytilus edulis L. and the
periwinkle, Littorina littorea L. Aquat. Toxicol. 7: 79–91.
Mackay, D. and Paterson, S. 1981, Calculating fugacity. Environ. Sci. Technol. 15: 1006–1013.
Mackay, D.; Shiu, W.Y.; Ma, K.C. 1992a, Illustrated Handbook of Physical-Chemical Properties and
Environmental Fate for Organic Chemicals: Monoaromatic Hydrocarbons, Chlorobenzenes, and
PCBs; Lewis Publishers: Chelsea, MI.; Vol. I.
Mackay, D.; Shiu, W.Y.; Ma, K.C. 1992b, Illustrated Handbook of Physical-Chemical Properties and
Environmental Fate for Organic Chemicals: Polynuclear Aromatic Hydrocarbons, Polychlori-
nated Dioxins, and Dibenzofurans; Lewis Publishers: Chelsea, MI.; Vol. II.
Mackay, D. 1994, Unraveling the Choreography of Contaminant Kinetics: Approaches to Quantifying
the Uptake of Chemicals by Organisms. In Bioavailability: Physical, Chemical and Biological
Interactions; Hamelink, J.L., Landrum, P.F., Bergman, H.L., Benson W., Eds.; CRC Press: Boca
Raton, FL.; pp. 171–177.
Mackay, D.; Shiu, W.Y.; Ma, K.C. 1997, Illustrated Handbook of Physical-Chemical Properties and
Environmental Fate for Organic Chemicals: Pesticide Chemicals; Lewis Publishers: Boca Raton,
FL; Vol. V.
Mayer, P.; Vaes, W.H.J.; Wijnker, F.; Legierse, K.C.H.M.; Kraaij, R.; Tolls, J.; Hermens, J.L.
2000, Sensing dissolved sediment porewater concentrations of persistent and bioaccumula-
tive pollutants using disposable solid-phase microextraction fibers. Environ. Sci. Technol. 34:
5177–5183.
Mayer, P. 2003, National Environmental Research Institute (NERI), Roskilde, Denmark. Personal
communication.
McLachlan, M. 1999, Framework for the interpretation of measurements of SVOCs in plants. Environ.
Sci. Technol. 33: 1799–1804.
Meadows, J.C.; Tillitt, D.E.; Huckins, J.N.; Schroeder, D. 1993, Large-scale dialysis of sample lipids
using a semipermeable membrane device. Chemosphere 26: 1993–2006.
Moring, J.B. and Rose, D.R. 1997, Occurrence and concentration of polycyclic aromatic hydrocarbons
in semipermeable membrane devices and clams in three urban streams of the Dallas-Fort Worth
Metropolitan Area, Texas. Chemosphere 34: 551–566.
Mueller, J. 2004, National Research Centre for Environmental Toxicology (NRCET), Coopers Plains,
Australia. Personal communication.
Neely, W.B. 1980, Chemicals in the Environment: Distribution, Transport, Fate and Analysis; Marcel
Dekker: New York, NY.
Ockenden, W.A.; Prest, H.F.; Thomas, G.O.; Sweetman, A.; Jones, K.C. 1998, Passive air sampling
of PCBs: Field calculation of atmospheric sampling rates by triolein-containing semipermeable
membrane devices. Environ. Sci. Technol. 32: 1538–1543.
Paschke, A. and Popp, P. 2003, Solid-phase microextraction fibre-water distribution constants of more
hydrophobic organic compounds and their correlations with octanol-water partition coefficients.
J. Chromatogr. A 999: 35–42.
Pauly, S. 1989, Permeability and Diffusion Data. In Polymer Handbook, 3rd ed., Brandrup, J. Immergut,
E.H., Eds.; Wiley & Sons: New York.; Section 6, pp. 435–449.
Petty, J.D.; Huckins, J.N.; Zajicek, J.L. 1993, Application of semipermeable membrane devices (SP-
MDs) as passive air samplers. Chemosphere 27: 1609–1624.
Petty, J.D. and Orazio, C.E. 1996, Application of Semipermeable Membrane Devices (SPMDs) As
Passive Monitors of the Environment of Antarctica. USGS, Midwest Science Center, Columbia,
MO; Unpublished report to National Science Foundation: Washington, DC.
Petty, J.D.; Huckins, J.N.; Alvarez, D.A.; Brumbaugh, W.G.; Cranor, W.L.; Gale, R.W.; Rastall, A.C.;
Jones-Lepp, T.L.; Leiker, T.J.; Rostad C.E.; Furlong, E.T. 2004, A holistic passive integrative
sampling approach for assessing the presence and potential impacts of waterborne environmental
contaminants.Chemosphere 54: 695–705.
Prest, H.F.; Jarman, W.M.; Burns, S.A.; Weismuller, T.; Martin, M.; Huckins, J.N. 1992, Passive
water sampling via semipermeable membrane devices (SPMDs) in concert with bivalves in the
Sacramento/San Joachin river delta. Chemosphere 25: 1811–1824.
Rand, G.M. and Petrocelli, S.R., Eds. 1985, Fundamentals of Aquatic Toxicology, Methods and Appli-
cations; Hemisphere Publishing: Washington, DC.
Rantalainen, A.-L.; Ikonomou, M.G.; Rogers, I.H. 1998, Lipid-containing semipermeable-membrane
devices (SPMDs) as concentrators of toxic chemicals in the Lower Fraser River, British
Columbia.Chemosphere 37: 1119–1138.
28 Chapter 1
Rogers, C.E. 1985, Permeation of Gases Vapours in Polymers. In Polymer Permeability, Comyn, J.,
Ed.; Elsevier Applied Science LTD: London, England; pp. 11–73.
Sabaliūnas, D. and Sodergren, A. 1996, Uptake of organochlorine pesticides by solvent-filled cellulose
and polyethylene membranes. Ecotox. Environ. Safe. 35: 150–155.
Schnoor, J.L.; Rao, N.B.; Cartwright, K.J.; Noll, R.M. 1982, Fate and Transport Modeling for Toxic
Substances. In Modeling The Fate Of Chemicals In The Aquatic Environment; Dickson, K. L.,
Maki, A. W., Cairns, j. Jr., Eds.; Ann Arbor Science Publishers: Ann Arbor, MI.; pp. 145–163.
Shoeib, M. and Harner, T. 2002, Characterization and comparison of three passive air samplers for
persistent organic pollutants. Environ. Sci. Technol. 36: 4142–4151.
Smedes, F. 2004, National Institute for Coastal and Marine Management (RIKZ), The Netherlands;
Unpublished Work.
Södergren, A. 1987, Solvent filled dialysis membranes simulate uptake of pollutants by aquatic organ-
isms. Environ. Sci. Technol. 21: 855–859.
Strandberg, B.; Bergqvist, P.-A.; Rappe, C. 1998, Dialysis with semipermeable membranes as an
efficient lipid removal method in the analysis of bioaccumulative chemicals. Anal. Chem. 70:
526–533.
Vrana, B. and Schüürmann, G. 2002, Calibrating the uptake kinetics of semipermeable membrane
devices in water: Impact of hydrodynamics. Environ. Sci. Technol. 36: 290–296.
Wang, W-X. and Fisher, N.S. 1999, Assimilation efficiencies of chemical contaminants in aquatic
invertebrates: A synthesis. Environ. Toxicol. Chem. 18: 2034–2045.
Wania, F.; Shen, L.; Lei, Y.D.; Teixeira, C.; Muir, D.C.G. 2003, Development and calibration of a resin-
based passive sampling system for monitoring persistent organic pollutants in the atmosphere.
Environ. Sci. Technol. 37: 1352–1359.
Wilcockson, J. B. and Gobas, F.A.P.C. 2001, Thin-film solid-phase extraction to measure fugacities of
organic chemicals with low volatility in biological samples. Environ. Sci. Technol. 35: 1425–1431.
Zabik, J.M.; Aston, L.S.; Seiber, J.N. 1992, Rapid characterization of pesticide residues in contaminated
soils by passive sampling devices. Environ. Toxicol. Chem. 11: 765–770.
Zabik, M. 1988, Michigan State University, East Lansing, MI. Personal communication.
Chapter 2
Fundamentals of SPMDs
2.1. SPMD DESCRIPTION AND RATIONALE
From the discussions thus far, the reader can infer that SPMDs are de-
signed to mimic the passive diffusional and partitioning steps of bioconcentration
while providing semi-quantitative to quantitative estimates of hydrophobic or-
ganic chemicals (HOC) concentrations in the ambient exposure medium. SPMDs
(see Figure 1.1) generally consist of a thin film of the neutral triglyceride triolein
(1,2,3-tri-[cis-9-octadecenoyl] glycerol) sealed in a layflat, thin-walled tube of
low-density polyethylene (LDPE). Although fish lipid (Huckins et al., 1990a) and
silicone fluids (Petty and Orazio, 1996) have been successfully used as SPMD
liquid phases, triolein was chosen as the standard for use in SPMDs for the fol-
lowing reasons: 1) it is a major storage lipid found in most organisms; 2) its
high-molecular weight (885.5 Daltons) results in extremely low LDPE membrane
permeability, even during dialytic recovery of analytes; 3) triolein is commercially
available in synthetic, high purity forms; 4) triolein-water partition coefficients
and octanol-water partition coefficients (K ow s) are similar in magnitude and are
well correlated (Chiou, 1985); 5) it is a liquid down to about −4 ◦ C; and 6) it
provides a convenient reservoir for performance reference compounds (PRCs; for
information on PRCs see Section 3.3.). Nonpolar liquid phases such as triolein
have very low interfacial tension with LDPE, which enables the formation of a
thin film with intimate membrane contact. Because solute diffusivity is 102 to 103
greater in liquids than in solids, the use of a liquid phase ensures rapid mixing of
accumulated residues. In contrast, solid phase sorbents in LDPE and other non-
porous hydrophobic polymer bags or enclosures are difficult to configure with
29
30 Chapter 2
relatively high surface-area-to-sorbent-volume (AV −1 ) ratios, and solutes in the

membrane generally must vaporize to make contact with the sorbent. This step
adds another potential barrier to the mass transfer or uptake of analytes.
As indicated earlier, the selection of nonporous LDPE layflat tubing for
SPMDs was based on it’s stability in organic solvents (required for dialysis and
membrane cleaning), the low diffusion rates of triolein relative to HOCs in LDPE
(both uptake and dialytic recovery processes), and it’s resistance to abrasion and
puncturing. The results of this research also enabled the development of polymeric
film (LDPE) dialysis in organic solvent, which has been shown to be a highly ef-
fective method for separating organic contaminants from lipids (Huckins et al.,
1990b; Meadows et al., 1993; Bergqvist et al., 1998). Thin-walled layflat LDPE
is widely available and, because it is a thermoplastic, the lipid phase can be sealed
inside the membrane tube using molecular welding (heat seals).
Although SPMDs are simple in design, the mechanisms governing their per-
formance as passive samplers of HOCs can be quite complex (see Chapter 3). The
underlying principle of molecular-size discrimination in the uptake and loss of
chemicals by SPMDs is shown in Figure 2.1. The sizes of the molecules shown in
the illustration are scaled to the postulated ≈10 Å diameter of the transient pores
in the membrane. Temperature and the presence of plasticizers/solvent will affect
the effective pore sizes.
In nonporous polymers such as LDPE, free volume is formed by random
thermal motion of polymer chains in rubbery regions of the matrix (LDPE is about
50% crystalline and 50% rubbery). The volume associated with “fixed pores,”
which exist only in the crystalline regions of the polymer, is largely insignificant
(Rogers, 1985) relative to the volume associated with the rubbery regions of the
polymer. Thus, the passive sampling of dissolved and vapor phase analytes involves
the dissolution of individual molecules into the rubbery regions of the polymer. The
diameters of the transient polymeric cavities range up to ≈10 Å (Comyn, 1985),
which precludes sampling of the waterborne residues associated with particulate
organic carbon or dissolved organic carbon such as humic acids. The frequency of
cavity formation is largely controlled by temperature-dependent chain segmental
motility. Also, it is noteworthy that the postulated size of transient cavities in
biomembranes is 9.8 Å (Opperhuizen et al., 1985). The molecular size limitation of
nonporous polymers suggests that only readily bioavailable or dissolved chemicals
(molecular weights <600 Daltons) will be sampled by SPMDs, which has been
corroborated by the work of Ellis et al. (1995). This size exclusion characteristic
of nonporous polymers is the reason for extremely low diffusion rates of triolein
in LDPE (i.e., losses from SPMDs).
Ions of organic and inorganic chemicals are not sampled by SPMDs be-
cause charged species are hydrophilic and are essentially insoluble in nonpo-
lar LDPE. Water quality variables, such as pH and salinity (Huckins et al.,
1999), may affect the dissolved concentrations of some compounds in envi-
ronmental waters (e.g., the residue concentrations of organic compounds with
pK a s > 4 and < 9).
Fundamentals of SPMDs 31
FIGURE 2.1 Exploded views showing the nonporous membrane size-exclusion phenomenon in
the uptake and loss of organic compounds. Middle illustration shows the movement of contam-
inant molecules through transient pores in the membrane and retention (membrane exclusion)
of much larger lipid molecules. Upper illustration shows similarly scaled space-filled molecular
models of some organic contaminants and triolein, along with the hypothetical polymer pore
(transient) size. Reprinted with permission from the American Petroleum Institute (Huckins et al.,
2002).
32 Chapter 2
Conceptually, SPMD data fills a gap between exposure assessments based on

direct analytical measurement of total residues in water and air, and the analysis
of residues present in biomonitoring organisms. SPMDs provide a biomimetic
approach (i.e., processes in simple media that mimic more complex biologi-
cal processes) for determining ambient HOC concentrations, sources, and gra-
dients. Residues accumulated in SPMDs are representative of their environmental
bioavailability (see Section 1.1.) in water and air and the encounter-volume rate as
defined by Landrum et al. (1994) is expected to be proportional to the uptake rate.
SPMD-based estimates of water concentrations can be readily compared to aquatic
toxicity data (generally based on dissolved phase concentrations) and SPMD ex-
tracts can be used to screen for toxic concentrations of HOCs using bioassays or
biomarker tests.
2.2. APPLICABILITY OF SPMDs
Although SPMDs concentrate a very wide range of hydrophobic organic

compounds, they are not suitable for all environmental contaminants. Table 2.1
lists chemicals classes or selected compounds shown to concentrate in SPMDs,
but is not all inclusive.
Examination of Table 2.1 shows that SPMDs are not suitable for sampling
very large organic molecules, ionized organic compounds, and metals. For com-
pounds such as chlorinated phenols with different pKa s, the environmental pH
determines the ratio of ionized to neutral species, which directly impacts the ca-
pacity of an SPMD to concentrate the chemical. Thus, the selectivity of SPMD
sampling is limited to size exclusion properties of the low density polyethylene
membrane (see Figure 2.1) and the polarity/ionization potential of the analyte. Hy-
drophobic or nonpolar compounds are characterized by a lack of polar functional
groups and a very low potential for ionization at environmental pHs (i.e., a range of
about 4.5 to 9). SPMDs will significantly concentrate ambient levels of nearly all
TABLE 2.1 Classes or Specific Chemicals Known to Concentrate in SPMDs
priority pollutant PAHs and alkylated PAHs chlorinated dibenzodioxins, including

many heterocyclic aromatics, cyclic 2,3,7,8-TCDD
hydrocarbons (e.g., decalin and alkylated polybrominated diphenyl ethers
decalins) and aliphatics chlorinated benzenes
organochlorine pesticides chlorinated anisoles and veratroles
other pesticides: includes diazinon, endosulfans, alkyl phenols (nonyl phenol)
pyrethroids, toxaphene, and trifluralin triclosan
PCB congeners tributyl tin
chlorinated naphthalenes sulfur
chlorinated dibenzofurans, including essentially, any compound
2,3,7,8-TCDF with log K ow ≥ 3.0a
a
See Table 8.1 for additions to this list.
FIGURE 2.2 Various applications of SPMDs reported in the literature.
hydrophobic compounds with log K ow s ≥ 3 and the sampling rates (Rs s) of most
HOCs are controlled by the “encounter volume”, as defined for aquatic organisms
in Chapter 1. Water quality variables, such as salinity (Brown, 1978), can affect the
dissolved concentrations of hydrophobic compounds in environmental waters, and
thus the amounts of residues accumulated by an SPMD. However, water quality
should have no effect on sampling rate constants (see Section 2.3.).
For compounds with log K ow s < 3, SPMDs may not perform as well as
other sampling procedures such as purge and trap methods for volatile organic
compounds and the polar organic chemical integrative sampler (POCIS) for hy-
drophilic organic compounds (Alvarez et al., 2004). Also, for compounds with log
K ow s and octanol-air partition coefficients (K oa s) larger than about 7.5 and 10.5,
respectively, only vanishingly small amounts will be available for uptake, because
of sorption to particulates and dissolved organic carbon. However, SPMDs have
been successfully used for determining chemicals with very high K ow s and K oa s
in environmental systems (McCarthy and Gale, 2001; Booij et al., 2002; Bartkow
et al., 2004) but may require the use of composite SPMD samples (e.g., three to
nine 1-mL triolein SPMDs).
Figure 2.2 illustrates a number of potential SPMD applications. More specif-
ically, SPMD technology has been used for the following: 1) determination of
the presence, sources, and the transport/fate of hydrophobic semi-volatile organic
pollutants; 2) estimation of ambient time-weighted average (TWA) dissolved or
vapor phase chemical concentrations; 3) determination of time-integrated fluxes of
dissolved and vapor phase chemicals in environmental media; 4) in situ biomimetic
sample extracts of readily available chemicals for toxicity screening (bioassays or
34 Chapter 2
biomarkers), immunoassay, and toxicity identification evaluation; 5) estimation

of organism exposure and bioconcentration factors (BCFs) for dissolved and va-
por phase compounds; and 6) polymeric membrane organic solvent dialysis for
enriching a wide variety of hydrophobic analytes in environmental sample ex-
tracts. Some of these applications and example studies are covered in subsequent
Chapters. Herein, we briefly discuss some general considerations associated with
SPMD applications.
Before choosing SPMDs for a project, data quality requirements must be
considered. Two extreme levels are litigation quality data (i.e., legally admissible)
and screening data (note that rigorous quality control can be applied to screening
tests). The SPMD approach can be readily used in screening projects, such as the
presence/absence, sources, and relative amounts of chemicals (ranking) measured
in SPMDs at different sites, to more in-depth studies designed to estimate the
ambient concentrations of chemicals. For projects in the USA requiring litigation
quality data, study results are typically generated by the US EPA or industry
standard methods in conjunction with a formal set of quality assurance (QA) and
quality control (QC) guidelines/parameters. Particular attention must be made
to security issues (QA) such as sample chain of custody. Because US EPA and
industry standard methods are often more than a decade behind the best available
technology, there has been increased use of more current, but well-established,
nonstandard methods (so-called “performance based methods”) in litigation.
The SPMD approach is widely used by environmental investigators and is
beginning to gain acceptance from regulatory and resource management agen-
cies (e.g., certain EPA regions and states, the United Kingdom, and the Czech
Republic). However, the authors are not aware of any studies conducted with pro-
tocols adequate for litigation. The SPMD studies presented herein may meet the
criteria based on QC parameters but typically fail to meet the QA requirements for
litigation, such as chain of custody documentation. However, as a priori accep-
tance of SPMD technology becomes more widespread, and studies are conducted
with more stringent QA standards, the likelihood of the successful incorporation
of SPMD data in litigation will increase.
Other issues of SPMD applicability relate to the type of matrix sampled. In
particular, the ability to extrapolate ambient concentrations from analyte concen-
trations in SPMDs differs significantly depending on the matrix sampled and the
variables affecting analyte concentrations in the matrix. An assumption, funda-
mental to the use of mathematical models for concentration extrapolations, is that
the sampling process does not significantly alter ambient solute or vapor concen-
trations of analytes. Theory and studies to date show that this assumption is not
violated when sampling surface waters and the atmosphere. However, some excep-
tions may occur when sampling sediments, groundwater and small, enclosed indoor
spaces. To maintain pore water concentration during sampling, solute resupply via
desorption from particulate and dissolved organic carbon phases of sediment must
be faster than the sampler uptake rates. In the only test of this assumption in the
literature, Booij et al. (2003) used LDPE strip samplers in sediments (collected
from two marine harbors) and found that pore water concentration estimates, based
on linear uptake rates and PRC loss rates, corresponded well to those based on
sediment-pore water equilibrium partition coefficients. These data suggest that
chemical resupply of the pore water was rapid enough to offset sampler uptake
or clearance rates. Because SPMDs and LDPE strips with similar surface areas
sample at essentially the same rate during linear uptake, this finding likely applies
to SPMDs as well (see Chapter 3 for more details).
Monitoring wells in fine grained strata often have low coefficients of perme-
ability or recharge rates or hydraulic conductivities (see Chapter 3 for more details).
In this case, SPMD sampling may significantly reduce well water concentrations
of the chemicals of concern. However, knowledge of SPMD uptake rates for target
compounds (see Appendix A), the groundwater hydraulic conductivity at the well
site, the cross-sectional surface area of the well and the approximate volume of
water in the well, should enable investigators to determine if water concentration
will be significantly reduced during sampling. If so, the size (i.e., surface area) or
numbers of SPMDs used in a well can be reduced as long as acceptable detection
and quantitation limits can be achieved. When very low quantitation limits are re-
quired and the well’s hydraulic conductivity is low, it may be possible to increase
the numbers or the surface area of the SPMDs used to ensure that the extraction
efficiency of target compounds from well water (dissolved phase) is >90% during
an exposure period. Thus, depending on the nature of the well, SPMD sampling
may deplete, moderately affect or have little effect on groundwater concentrations
of target solutes.
SPMDs are biomimetic only when partitioning-diffusion processes mediate
bioconcentration. The appropriateness of using SPMD data to predict equilibrium
concentrations of bioconcentratable contaminants in aquatic organisms is depen-
dent on a number of factors, as discussed in Chapter 7 and by Huckins et al.
(2004). Briefly, SPMDs and other passive samplers cannot account for physiolog-
ical and behavioral differences among species such as residue metabolism, dietary
uptake and trophic transfer, which can cause residue concentrations in tissues
to vary considerably from equilibrium partition levels (Connell, 1990; Huckins
et al., 2004). Also, unlike many aquatic invertebrates, shellfish and finfish, SPMDs
generally do not reach equilibrium with hydrophobic chemicals (i.e., for com-
pounds with log K ow s >5) during exposures of 42 d or less. Thus, direct com-
parisons of SPMD-water partition coefficients (K sw s) and BCFs often are not
feasible. However, SPMDs provide reasonably accurate estimates of in situ TWA
concentrations of dissolved-phase chemical concentration. Use of SPMD-derived
water concentrations and biomonitoring organism (BMO) tissue concentrations
may enable the development of improved regression models for estimating HOC
BCFs. This statement is based on the assumption that some of the scatter in BCFs
derived from existing regression models relates to the inability of previous inves-
tigators to determine TWA concentrations of bioavailable residues in exposure
waters. Regardless of the difficulties in directly relating SPMD concentrations
to BCFs, SPMDs provide reasonable estimates of aquatic organism exposure to
36 Chapter 2
persistent HOCs (e.g., Meadows et al., 1998; Huckins et al., 2004), via the dissolved
phase.
The case for using SPMDs as a biomimetic device for estimating TWA at-
mospheric exposure of HOCs to terrestrial organisms is less well developed. The
possible exception to this statement is the exposure of humans to semi-volatile or-
ganic compounds (SVOCs) in indoor air. Determination of TWA values for volatile
organic compounds using passive samplers is widely accepted as the method of
choice for assessing occupational exposure. Because K oa s are very large for hy-
drophobic SVOCs, sampling is generally integrative for months. Note that TWAs
can only be determined by integrative passive samplers. Furthermore, the sampling
rates and capacities of SPMDs for vapors of SVOCs are much higher than tradi-
tional passive samplers. This permits the isolation of sufficient target compound
mass for bioassay and lower quantitation limits.
2.3. ACCUMULATION OF CHEMICALS BY SPMDs
Although “Theory and Modeling” is more extensively discussed in Chapter 3,

it is helpful to briefly discuss some basic concepts related to the accumulation of
chemicals by SPMDs. Huckins et al. (1993) have shown that the uptake process
obeys first-order kinetics (Figure 2.3). This type of exchange kinetics is character-
ized by “half-lives” (t1/2 ), which are constant for a particular set of conditions and
Equilibrium
1.0
ar
iline
Concentration in SPMD
Curv
0.5
rea
Lin
0
t1/2
Time
FIGURE 2.3 Plot of the three phases of SPMD uptake, illustrating first-order exchange kinetics.
Time is given in halflives or t1/2 , which in this case is the time required to reach half of the
equilibrium concentration of a chemical. This figure is reproduced courtesy of the American
Petroleum Institute (Huckins et al., 2002).
chemicals, and “rate constants” that are independent of chemical concentration.

In this case, the rate of change of the concentration in an SPMD (Cs ) is given by
dCs /dt = ku Cw − ke Cs (2.1)
where ku is the uptake rate constant, ke is the elimination rate constant, Cw is the
concentration in the water phase, and t is time. In the case of SPMD-air exchange,
it is only necessary to replace the subscript “w” by “a”. In the initial stages of the
uptake, the term ke Cs is much smaller than ku Cw and Eq. 2.1 reduces to
dCs /dt ≈ ku Cw (2.2)
Equation 2.2 shows that Cs increases linearly with time when the aqueous
concentration is constant. This is why the initial stage of the uptake process is
called the “linear uptake phase” (Figure 2.3). Integrating Eq. 2.2 over time shows
that sampling is “integrative”, and that Cs is linearly proportional to the TWA
concentration in the water phase (Cw,TWA )

Cs (t) = dCs = ku Cw dt ≡ ku Cw,TWA t (2.3)
When equilibrium is attained, the rate of uptake balances the rate of loss, and
Eq. 2.1 reduces to
Cs = Cw ku /ke (2.4)
This stage of the uptake process is therefore called the “equilibrium sampling
phase”.
The time it takes to reach 50% of the equilibrium concentration (t1/2 ) is related
to the elimination rate constant (ke ) by
t1/2 = ln 2/ke (2.5)
where ln 2 is the natural logarithm of 2. Figure 2.3 shows that analytes accumu-
lated by an SPMD may be in the linear (integrative), curvilinear, or equilibrium
partitioning phases of uptake, depending on the chemical sampled, environmental
conditions, and the duration of the exposure. Also, Figure 2.3 shows that sampling
is essentially integrative up to t < t1/2 . For t > 4t1/2 , equilibrium is essentially
complete (>94%). Although the limits between the linear uptake phase, the curvi-
linear phase, and the equilibrium phase are somewhat arbitrary, these times can be
used to get a feeling for the extent to which sampling is integrative.
Modeling SPMD residue exchange as two compartments (membrane and
lipid) adds complexity (Huckins et al., 1993, 1999). A single compartment model
can be applied to SPMD residue exchange when using K sw , resistance is controlled
by the boundary layer, and equilibrium exists between the membrane and lipid
phases. The K sw is the volume-averaged partition coefficient of the membrane
and lipid phases and is given by Eq. 3.11. In simple one-compartment models
(Figure 2.4), the concentration at any moment in time is determined by competing
38 Chapter 2
FIGURE 2.4 Single compartment model for the uptake and release of hydrophobic organic com-
pounds. The a/w subscript refers to air or water.
rates of chemical uptake and release, as given in Eq. 2.1. This common model-
ing approach is widely used for estimates of the concentration of hydrophobic
chemicals in the lipids of aquatic organisms.
2.4. PASSIVE SAMPLER FUNDAMENTALS AND TERMINOLOGY
Until the advent of SPMDs and solid phase microextraction (SPME) fibers,
passive sampling devices were generally limited to integrative “diffusion” or “per-
meation” samplers (Fowler, 1982), with engineered barriers that control uptake
rates. The engineered rate-limiting barriers of these classical samplers consist of a
structural feature with stagnant air or water (diffusional samplers) or a nonporous
polymeric membrane (permeation samplers). In both cases, these barriers are de-
signed to account for >90% of the total resistance to solute or vapor uptake by the
sampler. The advantage of using the engineered barrier approach is that changes in
facial velocity and turbulence have little effect on sampling rates and thus can be
neglected. Also, the diffusion samplers are relatively simple to calibrate because
equations for calculating diffusion coefficients in air and water are well developed
and the relevant diffusional pathway or length is fixed by design. The disadvantage
of both of these engineered diffusion and permeation samplers is that their uptake
fluxes (e.g., ng cm−2 d−1 ) are generally more than an order of magnitude lower
than the uptake rates of samplers under external boundary layer control such as
SPMEs and SPMDs.
All passive monitoring devices operate on the basis of diffusive transfer,
regardless of whether they are classified as diffusion, permeation or unclassified
(e.g., SPMDs), and the rate-limiting barrier is the step with the greatest resistance to
mass transfer (see Figure 3.1). Fick’s first law is the fundamental law of diffusion.
It states that the flux of a chemical in the x-direction ( jx , e.g., ng cm−2 d−1 ) is
proportional to the concentration gradient (∂C/∂x)
jx = −Di (∂C/∂ x) = −Di C/δ i (2.6)
where Di is the diffusion coefficient in the rate limiting barrier, δ i is the effec-
tive thickness of the rate limiting barrier, and C is the concentration difference
across the barrier. Fick’s first law appears to apply to diffusion of trace levels
of HOCs through SPMD membranes and associated boundary layers. However,
the polymer permeability literature contains many references (e.g., Comyn, 1985)
where membrane-diffusion coefficients are not constant, requiring the application
of Fick’s second law.
Unlike the aforementioned classical samplers, the barrier limiting chemi-
cal uptake by SPMDs is dependent on physicochemical properties of the target
compound and the exposure conditions. For example, under conditions of low
water flows and turbulence (i.e., <1 cm s−1 ), the water boundary layer (WBL) is
relatively thick and compounds with log K ow s ≥ 4.5 are generally under WBL
control and δ w represents the effective thickness of the WBL. In this case, SPMDs
act as diffusion samplers (Huckins et al., 1999). However, under the same con-
ditions, compounds with log K ow s < 4.5 are under membrane control (δ m ), and
SPMDs act as permeation samplers. The reason for this bimodal rate control is
that the magnitude of the membrane-water or membrane-air partition coefficient
affects the resistance to mass transfer across the membrane (Eqs. 3.8 and 3.9).
More specifically, high membrane-water partition coefficients effectively reduce
resistance to mass transfer across the membrane. The transition point between
membrane and boundary layer rate control varies (see Figure 7.2) depending on
flow and turbulence conditions at the external surface of the membrane (i.e., thin-
ning of the boundary layer reduces resistance to mass transfer). Because SPMD
sampling rates are affected by environmental conditions, in situ sampling rates
may vary greatly (see Section 3.6.) across sites. As mentioned in Chapter 1
and discussed in Chapter 3, PRCs were developed to provide a means of de-
termining the effects of environmental exposure conditions on SPMD sampling
rates.
Some introductory comments on the conceptual basis of SPMD uptake (ku )
and release (ke ) rate constants and the associated sampling rates (i.e., Rs ) are
in order. The ku can be conceptualized as the volume of air or water cleared of
chemical per unit sampler mass or volume per unit time (e.g., mL g−1 d−1 or
mL mL−1 d−1 ) and Rs is the volume of air or water cleared per unit time (e.g.,
L d−1 ). Thus, the only difference between ku and Rs is that Rs is not normalized
to a unit mass or unit volume of sampler. In the context of organism exposure
(see Section 1.1.), the SPMD ku is equivalent to the “encounter volume” times
the fractional bioavailability of the chemical (which excludes dietary uptake). The
release rate constant (d−1 ) is equal to ku K sw
−1
.
Equation 1.1 gives the “clearance or sorption capacity” (E v ) of a thin poly-
meric film sampler for nonpolar organic compounds, which equals Vs K sw in the
case of water sampling by SPMDs. E v can be visualized as the volume of water
cleared of a target compound, when an SPMD has attained equilibrium with the
ambient environment. For moderate to high K ow compounds, the E v of an SPMD
is generally not approached in most exposures, but E v is often attained for rela-
tively low K ow compounds, exposed under similar conditions. In these cases, an
investigator can estimate E v volumes by using measured or estimated values of
K sw, or by assuming that K ow ≈ K sw . The E v volumes thus derived can be used to
compare to the volumes of air or water extracted by other methods, to determine if
40 Chapter 2
analyte mass is sufficient for analytical determination or bioassay screening, and

to evaluate the need for compositing SPMDs. For the case of air sampling, Cao
(1991) has proposed that sorbents capable of clearing >0.1 m3 g−1 (i.e., SPMD-
air partition coefficient [K sa ] ≈ 105 ) are suitable for the integrative sampling of
organic vapors. In aquatic environments, the minimal value is equivalent to about
0.12 L g−1 (i.e., K sw ≈ 120). For most passive samplers, this K s(a/w) value is far
too low to maintain linear uptake for periods greater than one week and the cor-
responding E v s would be inadequate to accumulate sufficient residues for trace to
ultra-trace analyses.
If the aim of a study is to estimate TWA concentrations, an integrative sampler
must be used. In this case, the response time (tr ) provides useful information on
sampler performance in environments where concentrations vary through time.
Following a step change in ambient exposure concentration, tr can be defined as
the time required for the sampling flux (Rs Cw ) of a passive monitoring device to
largely adjust to the full concentration change in the ambient environment (Fowler,
1982). Values of tr are representative of the average time an analyte spends within
the rate-limiting barrier. If a linear concentration gradient is assumed across the
rate-limiting barrier, then
tr = δi2 /2Di (2.7)
where tr is the response time for both integrative and steady state samplers and
subscripts were defined earlier. Other non-linear derivations of tr using Fick’s
second law show that tr is the time required to achieve approximately 63% increase
(relative to full change induced) in the concentration of a chemical in the rate
limiting zone or region due to a step change in ambient exposure concentration.
Using Eq. 2.7, values of D in water for phenanthrene, benzo[g,h,i]perylene and
decachlorobiphenyl (PCB congener 209) from Hofmans (1998), and an estimate
of SPMD boundary layer thickness under low flow conditions (<1 cm s−1 ) by
Gale (1998), tr s for these compounds were 131, 157 and 197 s, respectively. Note
that the compounds used in this example are known to be under WBL control
at low-flow rates. Under higher flow conditions, these response times would be
expected to be reduced by at least 4-fold. Fowler (1982) has suggested that a tr of
a few minutes or less is satisfactory for most applications of passive samplers.
If the aim of an investigator is to determine equilibrium concentrations in
samplers, then the “residence time” (tm ) is a logical parameter to compare among
samplers. The tm is the mean length of time that a molecule spends in a passive
sampling device, where solute exchange follows first-order kinetics. Residence
time is given by
tm = 1/ke (2.8)
where tm is about 1.5 t1/2 s. This parameter can be determined by curve fitting when
analyte concentrations reach the curvilinear or equilibrium phases of exchange
kinetics (Figure 2.3) or it can be calculated when the ku and K sw are known.
Residence times of chemicals in an SPMD are much larger than response times.
FIGURE 2.5 The amount of a chemical absorbed by a sampler through time, where the lag time (L)
is the time represented by the x-intercept of the extension (dashed line) of the steady state line AB.
For example, under low flow conditions and at a temperature of 18 ◦ C, the SPMD
residence time for phenanthrene is 45.4 d and ke values for benzo[g,h,i]perylene
are too small to measure, which suggests a residence time of >103 d.
The lag time tL is a closely related parameter to tr but is generally used for
diffusional processes under membrane control. This term is given by
tL = δi2 /6Di (2.9)
The meaning of this term is shown by Figure 2.5 and it is essentially the
time required to attain steady state flux across a barrier. When the resistance in the
boundary layer is negligible, the lag-time equation provides a convenient means
of calculating membrane or polymer-diffusion coefficients.
2.5. REFERENCES
Alvarez, D.A.; Petty, J.D.; Huckins, J.N.; Jones-Lepp, T.L.; Getting, D.C.; Goddard, J.P.; Manahan, S.E.
2004, Development of a passive, in situ, integrative sampler for hydrophilic organic contaminants
in aquatic environments. Environ. Toxicol. Chem. 23: 1640–1648.
Bartkow, M.E.; Huckins, J.N.; Müller, J.F. 2004, Field-based evaluation of semipermeable membrane
devices (SPMDs) as passive air samplers of polyaromatic hydrocarbons (PAHs). Atmos. Environ.
38: 5983–5900.
Bergqvist, P.-A.; Strandberg, B.; Rappe, C. 1998, Lipid removal using semipermeable membranes in
42 Chapter 2
Booij, K.; Zegers, B.N.; Boon, J.P. 2002, Levels of some polybrominated diphenyl ether (PBDE) flame
retardants along the Dutch coast as derived from their accumulation in SPMDs and blue mussels
(Mytilus edulis). Chemosphere 46: 683–688.
Booij, K.; Hoedemaker, J.R.; Bakker, J.F. 2003, Dissolved PCBs, PAHs, and HCB in pore waters and
Brown, R.A. 1978, Fate and Effects of Polynuclear Aromatic Hydrocarbons in the Aquatic Environment.
Report No. 4297; American Petroleum Institute (API): Washington, DC.
Cao, X-L.; 1991, Determination of specific retention volumes at 20 ◦ C for hydrocarbons on microporous
carbons. J. Chromatogr. 586: 161–165.
Comyn, J , Ed. 1985, Polymer Permeability; Elsevier Applied Science LTD: London, England.
Ellis, G.S.; Huckins, J.N.; Rostad, C.E.; Schmitt, C.J.; Petty J.D.; MacCarthy, P. 1995, Evaluation
of lipid-containing semipermeable membrane devices (SPMDs) for monitoring organochlorine
contaminants in the Upper Mississippi River. Environ. Toxicol. Chem. 14: 1875–1884.
Fowler, W.K. 1982, Fundamentals of passive vapor sampling. Am. Lab. 14: 80–87.
Gale, R.W. 1998, Three-compartment model for contaminant accumulation by semipermeable mem-
brane devices. Environ. Sci. Technol. 32: 2292–2300.
Hofmans, H.E. 1998, Numerical Modeling of the Exchange Kinetics of Semipermeable Membrane
Devices. M.S. Thesis, Netherlands Institute for Sea Research: Den Burg, The Netherlands.
Huckins, J.N.; Tubergen, M.W.; Manuweera, G.K. 1990a, Semipermeable membrane devices contain-
Huckins, J.N.; Tubergen, M.W.; Lebo, J.A.; Gale, R.W.; Schwartz, T.R. 1990b, Polymeric film dialysis
27: 2489–2496.
W.L.; Johnson, B.T. 2002, A Guide for the Use of Semipermeable Membrane Devices (SPMDs) as
Samplers of Waterborne Hydrophobic Organic Contaminants; Publication No. 4690; American
Huckins, J.N.; Petty, J.D.; Orazio, C.E.; Lebo, J.A.; Clark, R.C.; Gibson, V.L.; Gala, W.R.; Echols,
K.R. 1999, Determination of uptake kinetics (sampling rates) by lipid-containing semipermeable
membrane devices (SPMDs) for polycyclic aromatic hydrocarbons (PAHs) in water. Environ. Sci.
Technol. 33: 3918–3923.
Landrum, P.F.; Hayton, W.L.; Lee, H., II; McCarty, L.S.; Mackay, D.; McKim, J.M. 1994, Synopsis
of Discussion Session on the Kinetics behind Environmental Bioavailability. In Bioavailability:
Physical, Chemical and Biological Interactions; Hamelink, J.L., Landrum, P.F., Bergman, H.L.,
and Benson W., Eds.; CRC Press: Boca Raton, FL.; pp. 203–219.
McCarthy, K.A. and Gale, R.W. 2001, Evaluation of persistent hydrophobic organic compounds in the
Columbia River Basin using semipermeable membrane devices. Hydrol. Process 15: 1271–1283.
Meadows, J.C.; Echols, K.R.; Huckins, J.N.; Borsuk, F.A.; Carline, R.F.; Tillitt, D.E. 1998, Estimation
of uptake rate constants for PCB congeners accumulated by semipermeable membrane devices
and brown trout (Salmo trutta). Environ. Sci. Technol. 32: 1847–1852.
Opperhuizen, A.; van der Velde, E.W.; Gobas, F.A.P.C.; Liem, A.K.D.; van der Steen, J.M.D.; Hutzinger,
O.; 1985, Relationship between bioconcentration in fish and steric factors of hydrophobic chem-
icals. Chemosphere 14: 1871–1896.
Petty, J.D. and Orazio, C.E. 1996, Application of Semipermeable Membrane Devices (SPMDs) As
Passive Monitors of the Environment of Antarctica. USGS Midwest Science Center, Columbia,
MO; Unpublished report to National Science Foundation, Washington, DC.
Rogers, C.E. 1985, Permeation of Gases Vapours in Polymers. In Polymer Permeability, Comyn, J.,
Ed.; Elsevier Applied Science LTD: London, England; pp. 11–73.
Chapter 3
Theory and Modeling
3.1. UPTAKE MODELS
In this section we focus on solute exchange between SPMDs and water, but
the general scheme can be easily extended to vapor exchange between SPMDs
and air. Before going into details on the mathematical formulation of the exchange
kinetics between SPMDs and the sample medium, it is instructive to have a look
at the successive transport resistances that dissolved contaminants must overcome
to be absorbed by exposed SPMDs. A few mm from the SPMD surface, all trans-
port is dominated by eddy diffusion and convection (Figure 3.1). When solutes
approach the SPMD surface, molecular diffusion becomes increasingly important
relative to eddy diffusion and convection. The region where molecular diffusion
dominates the transport is commonly referred to as the “aqueous diffusion layer”,
but other names are used as well, such as “water boundary layer” (WBL), “diffusive
sub-layer”, and “diffusive boundary layer”. After crossing the WBL, contaminant
molecules may encounter a potentially complex layer of periphyton, macrofauna,
imbibed and surficial particulates, and in some cases, calcareous precipitates. We
refer to this layer or coating on the external surface of the SPMD membrane as the
“biofilm”, bearing in mind that material composition in this layer may be organic
and/or inorganic depending on the exposure site and on the season. Also, the biofilm
thickness, composition, and membrane coverage generally varies and may even
be absent. Contaminant transport through the biofilm takes place via molecular
diffusion, or in some cases by internal ventilation by the organisms present. After
crossing the biofilm, the solutes arrive at the membrane surface, diffuse through
and concentrate in the LDPE phase, and finally are concentrated in the triolein.
45
46 Chapter 3
FIGURE 3.1 Schematic overview of the concentration distribution inside and outside SPMDs. The
deltas (δ) represent the effective thickness of each region and the associated subscripts denote
the region.
Two types of models are used to quantitatively describe the SPMD-water

exchange of hydrophobic organic chemical (HOC) solutes. These models can
easily be inter-converted, and differ only in the definition of the rate constants
used. The fact that various authors have used the same symbols to denote different
kinds of rate constants is a complicating factor that requires a careful inspection
by the reader to delineate the meaning and assumptions of the respective authors.
3.1.1. Chemical Reaction Kinetics (CRK) Model

The first type of model is based on models describing bioconcentration, which
in turn are based on an analogy with chemical reaction kinetics. For this reason
we will refer to this type of model as the chemical-reaction kinetics (CRK) model.
The exchange process is thought of as the net result of a forward (uptake) and a
backward (release) reaction that is first order with respect to reactant concentration.
Hence, the rate of change of the solute concentration in the SPMD (Cs ) is given by
dCs
= ku C w − ke C s (3.1)
dt
where Cw is the aqueous concentration, ku is the rate constant for the uptake
process, and ke is the rate constant for the release process (ku and ke are often
named k1 and k2 when describing bioconcentration). Solving Eq. 3.1 for the initial
condition Cs = 0 at t = 0, gives
Cs = K sw Cw [1 − exp(−ke t)] (3.2)
Theory and Modeling 47
where the SPMD-water partition coefficient (K sw ) equals the ratio of the two rate
constants (ku /ke ). The only assumption underlying this CRK model is that both
the uptake and the release are first order processes. Equation 3.2 is useful for
comparing the kinetics of SPMD uptake and bioconcentration, and can be used to
determine the extent to which SPMDs are biomimetic (i.e., similar to contaminant
uptake by biota) for a particular exposure scenario (Huckins et al., 2004). This
model is highly empirical, and other models are needed to examine the uptake
and release rate constants in terms of fundamental processes.
3.1.2. Mass Transfer Coefficient (MTC) Model

The second type of model is a mass-transfer coefficient (MTC) model, which
is based on a mathematical description of solute-mass transfer through sequential
but distinct physical environments (phases or regions). In the case of SPMD ex-
posures, these environments include the WBL, the biofilm (not always present),
the membrane, and the triolein. The MTC (k) can be viewed as the velocity (e.g.,
cm s−1 ) of a solute moving through a region, and k is equivalent to D/δ, where
D is the diffusion coefficient and δ is the effective thickness of a particular phase.
With the MTC model, it is assumed that the fluxes through the successive regions
are linearly proportional to the concentration difference between the end points of
the phases (Figure 3.1).
The flux ( jw : mass per unit surface area per unit time) through the WBL is
assumed to be proportional to the concentration difference between bulk water
(Cw ) and the water side of the biofilm-water interface (Cw− )
jw = kw (Cw − Cw− ) (3.3)
where kw is the MTC for the WBL. Similarly, the flux through the biofilm ( jb )
equals
jb = kb (Cb+ − Cb− ) (3.4)
where Cb+ , Cb− are the concentrations in the biofilm at the water side and at the
membrane side, respectively, and kb is the MTC for the biofilm. Finally, the flux
through the membrane ( jm ) is given by
jm = km (Cm+ − Cm− ) (3.5)
where Cm+ , Cm− are the concentrations in the membrane at the biofilm side and
at the triolein side, respectively, and km is the MTC for the membrane. To use
the MTC model, it is assumed that sorption equilibrium exists at all interfaces,
that the fluxes through the various phases are equal ( jw = jb = jm ), and that the
triolein phase is well mixed as shown in Figure 3.1. The assumptions of interfacial
equilibrium and equality of the fluxes through the phases can be used to eliminate
the contaminant concentrations at the interfaces (Flynn and Yalkowsky, 1972),
48 Chapter 3
yielding an expression for the solute flux ( j) between the phases

CL
j = ko C w − (3.6)
K Lw
where CL is the concentration in the lipid phase, K Lw is the lipid-water partition
coefficient, and ko is the overall MTC. The ko is also known as the overall conduc-
tivity, because it is the proportionality constant between the flux and the effective
concentration or fugacity difference that drives this flux. Similarly, 1/ko is known
as the overall transport resistance (or impedance, Io ), which is the sum of the
transport resistances for the successive phases (Iw , Ib , Im ).
Io = Iw + Ib + Im (3.7)
1 1 1 1
= + + (3.8)
ko kw kb K bw km K mw
From Eq. 3.8, it follows that transport resistances are additive and that high partition
coefficients (K bw and K mw ) reduce the resistance to mass transfer in the respective
phases. As discussed earlier, MTCs can be written as D/δ (Figure 3.1 illustrates
the hypothetical δ of each SPMD related barrier to mass transfer). The importance
of barrier thickness in resistance to mass transfer is more explicit in the following
version of Eq. 3.8.
1 δw δb δm
= + + (3.9)
ko Dw Db K bw Dm K mw
From Eq. 3.9, the greater the thickness (δ) of each phase the larger the resistance
to solute transfer. Although the use of δ w is convenient for modeling and con-
ceptualizing WBL resistance, it is largely fictitious, as complex hydrodynamics
control resistance to mass transfer across the WBL in environmental exposures
(see Section 3.6.5. for a more in- depth discussion on this phenomenon).
It is now common practice to analyze HOC residues in the membrane and
lipid phases together, as opposed to analyzing the lipid phase alone. Assuming
that the concentrations in the lipid and in the membrane are close to equilibrium,
Eq. 3.6 can be written as

Cs
j = ko C w − (3.10)
K sw
where Cs is the concentration in the whole SPMD, and K sw is
Vm K mw + VL K Lw
K sw = (3.11)
Vm + VL
which is the volume weighted average of the membrane-water and the lipid-water
partition coefficients. From Eq. 3.10, the rate of change of the solute concentration
in the SPMD (Cs ) is given by

dCs Aj Ako Cs
= = Cw − (3.12)
dt Vs Vs K sw
where A is the SPMD surface area, and Vs its volume. With the initial condition
Cs = 0 at t = 0, the solution to Eq. 3.12 is
Cs = K sw Cw [1 − exp(−ke t)] (3.13)
which is identical to Eq. 3.2. The exchange rate constant ke is given by

A ko
ke = (3.14)
Vs K sw
The advantage of the MTC model, as opposed to the CRK model, is that the
exchange rate constant ke is no longer an empirical constant, but is now defined in
terms of more fundamental processes that can be separately modeled. Equations
3.1 and 3.12 are linked via the equality
ko A
ku = (3.15)
Vs
Caution is needed regarding the units of ku . Eq. 3.15 gives ku in units of reciprocal
time (e.g., d−1 ). This is a result of expressing Cs on a volume basis. In the case
that Cs is expressed on an SPMD mass basis, ku has units of volume per mass
per time (e.g., mL g−1 d−1 ). An uptake rate constant given in these units can be
converted to ku in d−1 by multiplying with the SPMD density (0.91 g mL−1 ), as
further discussed in Appendix A.
A number of authors have presented models of somewhat higher mathematical
complexity. Hofmans (1998) presented an uptake model based on Fick’s second
law. Using the results obtained after numerical integration of the partial differential
equations, she confirmed the validity of the assumptions that the triolein phase
is well mixed, and that the CL /Cm ratio equals the triolein-membrane partition
coefficient after 1 to 13 days for phenanthrene and polychlorinated biphenyl (PCB)
congener 209, respectively. Gale (1998) presented a pseudo-steady state numerical
model that allowed for studying the effects of transient exposure concentrations,
changes in analyte physicochemical properties, and modifications in SPMD design.
Booij et al. (2003a) have given an analytical solution of the uptake equation for
the case of aqueous concentrations that vary linearly with time.
3.2. KINETIC AND EQUILIBRIUM SAMPLING
Equation 3.13 shows that the concentration in SPMDs gradually increases

with time until equilibrium is attained. When ke t 1, Cs attains its equilibrium
value
Cs = K sw Cw (3.16)
In this case, the exposure mode is called “equilibrium sampling”. When ke t 1

(short exposure times and/or highly hydrophobic compounds), the group between
50 Chapter 3
square brackets in Eq. 3.13 is approximately equal to ke t, and Cs is given by

Cs = K sw Cw ke t (3.17)
Because Cs increases linearly with time, this phase of an exposure is called “kinetic
sampling”, or the “linear uptake mode”, and sampling is time-integrative. The
amount (N ) absorbed by SPMDs during kinetic sampling equals
N = Vs K sw ke Cw t (3.18)
N = Rs C w t (3.19)
which defines the apparent water sampling rate (Rs ) for kinetic sampling.
Rs = Vs K sw ke (3.20)
The sampling rate provides a conceptual link between classical batch extraction
techniques and passive sampling with SPMDs, because Rs t equals the equivalent
extracted water volume. For distinguishing between the kinetic- and equilibrium-
sampling modes, limits must be set that are (unavoidably) arbitrary in nature.
The errors involved depend on the degree of equilibrium attained during sam-
pling. However, there is no reason to use either approximation. By substituting
the definition of Rs (Eq. 3.20) into the full model (Eq. 3.13), one establishes the
links among the calibration data (Rs s and K sw s), the absorbed amounts after the
exposure, and the aqueous concentration.

Rs t
N = Vs K sw Cw 1 − exp − (3.21)
Vs K sw
The aqueous concentration can therefore be calculated from the absorbed amounts
by
N
Cw = (3.22)
Rs t
Vs K sw 1 − exp −
Vs K sw
In the limiting cases, where t → 0 or t → ∞, Eq. 3.22 reduces to the special cases
of sampling during the linear uptake kinetic phase and equilibrium-sampling phase
(Eqs. 3.16 and 3.19), respectively. The advantage of Eq. 3.22 is that no arbitrary
limits have to be set for the two sampling modes. Thus, slightly more complex
mathematics is the price paid for avoiding this ambiguity.
3.3. DISSIPATION OF PERFORMANCE REFERENCE

COMPOUNDs (PRCs)
In order to assess an analyte’s in situ SPMD-water exchange kinetics, per-

formance reference compounds (PRCs; as described in Chapter 1) are added to
SPMD triolein prior to an exposure (Ellis et al., 1995; Booij et al., 1998; Ockenden
et al., 2001; Huckins et al., 2002a). To use this approach, the investigator must be
sure that the PRCs do not occur in the environment (e.g., certain non-labeled or
native compounds such as PCB congeners 14, 29, 50, and a variety of compounds
labeled with deuterium, 13 C, or 14 C generally can be used). PRC dissipation is
governed by
N = N0 exp (−ke t) (3.23)
where N0 is the amount present at t = 0. If N and N0 are measured, the PRC
release rate constant (ke ) can be estimated using
ln (N /N0 )
ke = − (3.24)
t
When the ke and SPMD-water partition coefficient of the PRC are known, its
Rs can be calculated from Eq. 3.20. More precisely, we assume that the PRC Rs
is representative of the in situ sampling rates of target compounds with similar
physicochemical properties as the PRC. Various approaches have been used to
estimate sampling rates for all analytes from the PRC-derived sampling rates (see
Section 3.6.).
3.4. POTENTIAL EFFECTS OF DISSOLVED ORGANIC CARBON

(DOC ) ON SPMD CALIBRATION DATA
Because SPMD sampling rates are independent of chemical concentration,

the attenuation of dissolved phase concentrations by DOC (including colloids) or
particulate organic carbon (POC: particle size >0.45 µm) should have no mea-
surable effect on actual analyte Rs values. In practice, the fractional amount of
chemical sorbed to DOC can affect the accuracy of SPMD calibration exper-
iments, which in turn can affect SPMD-derived estimates of dissolved environ-
mental concentrations. Even in the laboratory, measurements of the true dissolved-
phase concentrations of test chemicals are especially difficult for compounds with
high octanol-water (K ow s) partition coefficients. When batch extraction methods
(i.e., liquid-liquid extraction) are used for sampling water for SPMD calibration,
the fractional amounts of a compound sorbed to POC and DOC are recovered
along with the dissolved fraction. This results in an overestimation of dissolved-
phase concentrations. Pre-filtration of water samples with glass fiber filters (GFFs)
is effective in removing POC, but this step may result in attenuation of dissolved
phase concentration due to adsorption of solutes on the GFF. To circumvent this
problem, cartridges or columns of solid phase sorbents (SPE) equipped with GFF
pre-filters have been used to measure dissolved-phase water concentrations. This
approach is largely based on the results of one study, which showed that DOC
associated chemicals are not extracted during SPE extraction of dissolved residues
(Landrum et al., 1984). However, Mackay (1994) has suggested that desorption
times for very small particles (i.e., <1 µm) are on the order of a few seconds for
compounds with particle-water partition coefficients of about 104 . Based on this
52 Chapter 3
study, SPE sampling of DOC-associated residues cannot be ruled out. Also, no

data are available on the potential of the surficial retention of colloids or DOC by
the SPE sorbent. These factors would lead to an overestimation of dissolved con-
centrations. Other sampling equipment may also yield inaccurate measurements
of dissolved concentrations. Pre-equilibration of laboratory exposure systems is a
separate issue related to attainment of steady-state water concentrations in the ex-
posure system prior to initiation of the test. Clearly, the issue of DOC or colloidally
sorbed contaminants deserves additional discussion.
Recently, Burkhard (2000) reviewed contaminant sorption by dissolved or-
ganic matter. Using several hundreds of DOC-water partition coefficients (K DOC )
reported in these studies, he found that DOC-water partition coefficients for natu-
rally occurring DOC (humic and fulvic acids, sediment pore water, soil pore water,
groundwater, and surface water) was best described by
log K DOC = log K ow − 1.11 s = 0.66, n = 127 (3.25)
The 95% confidence interval amounted to 1.3 log units, which corresponds to
a scatter in the K DOC values by a factor of 20. Burkhard (2000) argues that the
uncertainty in Eq. 3.25 originates partly from inter-laboratory variation, and partly
from differences in DOC quality. In view of the wide range of DOC sources
included in Eq. 3.25, it does not seem likely that the uncertainties in any K DOC
encountered would be more than 1.3 log units. Hence, the worst-case estimate
(maximum sorption of dissolved phase residues by DOC) can be described by
log K DOC = log K ow + 0.20 (3.26)
When DOC-bound residues are extracted along with the dissolved phase (i.e., the
total concentrations are measured), then the ratio of the truly dissolved concentra-
tions (Cw ) to the total concentration (Ct ) of an HOC can be estimated by
Cw 1
= (3.27)
Ct 1 + [DOC]K DOC
where the brackets represent the concentration of DOC. Only few authors report
DOC concentrations in their SPMD calibration studies. Meadows et al. (1998)
and Huckins et al. (1999; 2002b) found DOC values to be below detection limits
of <1 mg L−1 . In a later study, the DOC content of deep-well water used by
Huckins et al. (1999) was quantified as 0.26 mg L−1 (McCarthy, 2004). Luellen
and Shea (2002) report values of 0.3 and 1.3 mg L−1 in two separate experiments.
Rantalainen et al. (2000) argued that DOC-bound contaminants did not contribute
to their Cw estimates because the fine fraction of the sediment (<63 µm) in their
exposure system was sieved out, and because the XAD-2 resin used to extract the
water would not retain DOC-bound contaminants, but no experimental values of
DOC concentrations are available for this study. Booij et al. (2003a) used ultra pure
Milli-Q water with a specified DOC content <0.01 mg L−1 , but did not measure
the actual DOC concentrations in their system.
1
Cw/Ctot
0.1
4 5 6 7 8
log Kow
FIGURE 3.2 Reduction of concentrations of dissolved compounds due to sorption to DOC, for
DOC levels of 0.1 (solid line) and 1 mg L−1 (dashed line).
The effect of binding to DOC on the freely dissolved contaminant concen-

tration is shown for DOC concentrations of 0.1 and 1 mg L−1 in Figure 3.2, using
Eqs. 3.25 and 3.27. At DOC concentrations of 1 mg L−1 , a two-fold reduction
in Cw can be expected at log K ow = 7.3. Because of the uncertainties associated
with the Burkhard equation (1.3 log units), this reduction could occur in the range
of 6 < log K ow < 8.6 at this DOC level.
3.5. SPMD-WATER PARTITION COEFFICIENTS
Because both LDPE and triolein are non-polar organic phases, it can be
expected that SPMD-water partitioning is driven by hydrophobic interactions, and
that strong correlations will exist between K sw values and K ow s. Experimental
values of K sw are available for polycyclic aromatic hydrocarbons (PAHs), PCBs,
chlorobenzenes, and nonpolar and moderately polar pesticides (Figure 3.3). The
experimental evidence suggests that K sw values are not temperature dependent in
the 2 to 30 ◦ C temperature range. Huckins et al. (2002a) reported that phenanthrene
deviates from this general rule (0.4 log units decrease between 8 to 30 ◦ C), whereas
54 Chapter 3
FIGURE 3.3 SPMD-water partition coefficients (mL mL−1 units) as a function of log Kow for
PAHs—filled circles: Huckins et al. (1999), filled triangles: Huckins et al. (2004); phenanthrene,
PCB 52, and p,p -DDE—open triangles: Huckins et al. (2002a); chlorobenzenes, PAHs, and
PCBs—squares: Booij et al. (2003a); pesticides—filled diamonds: Sabaliunas and Södergren
(1997); and HCHs—open diamonds: Vrana and Schüürmann (2002). Additional Ksw data
were calculated for PCBs (asterisks) and alkylated benzenes (crosses) using the Kmw data from
Lefkovitz et al. (1996) and Reynolds et al. (1990), and the KLw data from Chiou (1985).
PCB 52 and p,p’-DDE did not. However, Booij et al. (2003a) reported the log K sw
value for phenanthrene to be constant within 0.17 log units in the temperature
range 2 to 30 ◦ C. Additional K sw values for mono- to hexachlorobiphenyls can be
determined by combining the LDPE-water partition coefficients of PCB congeners
1, 5, 29, 47, 98, and 154 (Lefkovitz et al., 1996) with the triolein-water partition
coefficients of PCBs with an equal number of chlorine atoms (Chiou, 1985), and
adopting a triolein mass fraction of 0.20 in standard SPMDs. All log K sw data are
plotted as a function of log K ow in Figure 3.3. When applicable, the average log
K sw values over different temperatures (within a single study) are plotted. Log K ow
values were adopted as specified in Appendix A (Tables A.1 to A.3). The log K sw
data could best be described by a quadratic equation in log K ow . The inclusion of
a different intercept for moderately polar and nonpolar compounds resulted in a
significantly better fit ( p < 0.001).
log K sw = a0 + 2.321 log K ow − 0.1618 (log K ow )2 (3.28)
PCBs, PAHs, 4,4 -DDE : a0 = −2.61
polar pesticides : a0 = −3.20
n = 45, s = 0.25, r = 0.97
The inclusion of a third-order term did not yield a significantly better fit ( p > 0.6).
There are no indications that PAHs and nonpolar chlorinated hydrocarbons have
a different K sw − K ow dependence. By contrast, the moderately polar pesticides
such as hexachlorocyclohexanes (HCHs), dieldrin, chlorpyrifos, heptachlor, and
trifluralin (open and closed diamonds in Figure 3.3) have K sw values that are 0.6
log units lower than PAHs and PCBs with similar K ow values ( p < 0.001).
Figure 3.3 shows that the log K ow -log K sw plot for compounds with
log K ow > 5 deviates from linearity. This phenomenon is also observed for plots of
log bioconcentration factor (BCF) versus log K ow (Connell, 1990). Chiou (1985)
has shown that a similar deviation occurs in a triolein-water system alone, at log
K ow > 5.5, as a result of solute-triolein incompatibility. Similarly, Banerjee and
Baughman (1991) argued that BCFs of large molecules are smaller than expected
on the basis of their hydrophobicity as a result of their disrupting effect on the struc-
ture of the lipid phase. The curvilinearity is not likely to be caused by sorption to
DOC in the experimental systems (Section 3.4.), because deviations from linearity
already occur at log K ow ≈ 5, whereas the effect of DOC-bound contaminants can
only be expected at log K ow values > 7.
3.6. WATER SAMPLING RATES
Sampling rates have been determined for a large number of compounds,

representing several compound classes, and include PAHs (Huckins et al., 1993;
1999, 2004; Luellen and Shea, 2002; Booij et al., 2003a), PCBs (Huckins et al.,
1993; Meadows et al., 1998; Booij et al., 2003a), chlorobenzenes (Vrana and
Schüürmann, 2002; Booij et al., 2003a), PCDDs/PCDFs (Rantalainen et al., 2000),
and a number of polar pesticides (Sabaliūnas and Södergren, 1997; Vrana and
Schüürmann, 2002), both in the lab and in the field. A comparison of experimental
sampling rates is hindered by the differences in experimental conditions, such
as temperature, flow rates, and geometry of the mounting cages. The sorption of
analytes to DOC may result in an overestimation of the dissolved concentrations,
and hence to an underestimation of the sampling rates.
3.6.1. Temperature Effects

Sampling rates at different temperatures have been determined by Huckins et
al. (1999) for PAHs at 10, 18, and 26 ◦ C, by Rantalainen et al. (2000) for PCDDs,
PCDFs, and non-ortho chlorine substituted PCBs at 11 and 19 ◦ C, and by Booij
et al. (2003a) for chlorobenzenes, PCBs, and PAHs at 2, 13 and 30 ◦ C. The effect of
temperature on the sampling rates can be quantified in terms of activation energies
(E a ) for mass transfer, as modeled by the Arrhenius equation

E a
Rs = Rs,∞ exp − (3.29)
RT
56 Chapter 3
FIGURE 3.4 Activation energies of contaminant sampling by SPMDs as a function of molar volume
(LeBas method). PAHs—filled circles: Huckins et al. (1999); filled squares: Booij et al. (2003a);
pesticides—open circles: Huckins et al. (2002b); chlorobenzenes and PCBs—open squares:
Booij et al. (2003b); planar PCBs—triangles: Rantalainen et al. (2000); and PCDD/Fs—asterisks:
Rantalainen et al. (2000).
where Rs,∞ is the sampling rate at the hypothetical upper limit where temperature
is infinite, R is the gas constant, and T is the absolute temperature.
Values of E a can be determined by plotting the natural logarithm of Rs
(ln Rs ) versus the reciprocal absolute temperature (1/T ). The activation energy
can then be calculated by multiplying the slope of the regression line with the gas
constant. Figure 3.4 shows a plot of E a as a function of molar volume, estimated
using the LeBas method (Reid et al., 1987; Mackay et al., 1992a, 1992b). Activa-
tion energies range between −3 and 103 kJ mol−1 . For PAHs, good correspondence
exist for the E a values estimated from the data reported by Huckins et al. (1999)
and Booij et al. (2003a). The activation energies reported by Rantalainen et al.
(2000) for tetrachlorobiphenyls with non-ortho Cl atoms show good correspon-
dence with the values for PCBs reported by Booij et al. (2003a). For penta- and
hexachlorobiphenyls, the E a values reported in both studies differ by 40 to 60 kJ
mol−1 . This difference may be related to differences in planarity for the PCBs
used in both studies. A number of non-polar and moderately polar pesticides
(HCB, DDT-like compounds, HCHs, dieldrin, endrin, chlordanes) have activation
energies that are similar to those of PAHs and PCBs. Taking all data together,
E a values seem to increase slightly from about 20 kJ mol−1 at Vm = 180 cm3
mol−1 (log K ow ≈ 4) to about 40 kJ mol−1 at Vm = 350 cm3 mol−1 (log K ow ≈ 7),
with the exception of the data on PCDDs/PCDFs and non-ortho substituted PCBs
reported by Rantalainen et al. (2000), which are in the range 40 to 100 kJ mol−1 . The
activation energies of 23-64 kJ mol−1 obtained from the dissipation of trichloroben-
zene, PCB congeners 4 and 29 in the field are in line with the values shown
in Figure 3.4 (Booij and van Drooge, 2001). The average of all E a values is
37 kJ mol−1 with a standard deviation of 21 kJ mol−1 . This means that a temper-
ature increase from 10 to 20 ◦ C causes an increase in sampling rate by a factor
of about 1.7. The highest (103 kJ mol−1 ) and lowest (−3 kJ mol−1 ) E a value
observed would correspond to a change in sampling rate by a factor of 4.5 and 1
over this temperature range, respectively. Therefore, the effect of temperature on
SPMD sampling rates in the field appears to be modest, unless large temperature
differences exist between exposure sites or exposure periods.
3.6.2. R s Calibration Data

A selection of published Rs values is shown as a function of log K ow in
Figure 3.5. Only data from calibration studies conducted at 15 ± 4 ◦ C were con-
sidered, in order to eliminate the temperature effect as much as possible. When
FIGURE 3.5 Sampling rates at 15 ± 4 ◦ C as a function of log Kow for PCBs—open circles:
Meadows et al. (1998), PAHs—closed circles: Huckins et al. (1999), closed triangles: Huckins
et al. (2004); chlorobenzenes/PCBs/PAHs—open triangles: Booij et al. (2003a); HCHs/HCB—
diamonds: Vrana and Schüürmann (2002); non-ortho PCBs/PCDDs/PCDFs—asterisks: Ranta-
lainen et al., 2000); and pesticides—squares: Huckins et al. (2002b).
58 Chapter 3
FIGURE 3.6 Sampling rates of compounds in the log Kow range 6 to 7 as a function of water flow
velocity. Connected data points represent measurements within single studies (dashed line:
Luellen and Shea, 2002; solid line: Vrana and Schüürmann, 2002).
calibration data at multiple temperatures existed, the temperature-interpolated val-

ues at 15 ◦ C were calculated from Eq. 3.29. For compounds that are low to mod-
erately hydrophobic (log K ow = 3 to 5), log Rs values increase linearly with log
K ow . A maximum in Rs values is reached for compounds with log K ow = 5 to 6,
and thereafter, compound Rs values decrease with increasing log K ow . Although
the shape of the log Rs versus log K ow curves is similar for all studies, the sampling
rates among studies appear to be shifted by a relatively constant factor.
In order to check if this shift is related to flow velocity, the geometric mean
of Rs values for compounds with log K ow = 6 to 7 was plotted versus the reported
linear flow velocity (Figure 3.6). The linear flow velocity seems to be a poor pre-
dictor of sampling rates. At flow velocities below 10 cm s−1 , three values cluster
around 10 ± 3 L d−1 and 6 values fall in the range 3 ± 1 L d−1 . The highest
sampling rates (≈100 L d−1 ) are reported by Booij et al. (2003a) at a flow veloc-
ity of about 90 cm s−1 . The lack of correlation between linear flow velocity and
Rs values is not surprising. The linear flow velocity is typically calculated from
the volumetric rate of flow and the cross-sectional area of the exposure cham-
ber, which is only a crude measure of the hydrodynamical conditions prevailing
at the SPMD-water interface. At low flow velocities, inertial currents originat-
ing from the inflow-orifices may be much larger than calculated linear velocities.
Moreover, it is the flow at the membrane-water interface that controls the ex-
change rates in the boundary layer, rather than the flow at larger distance from the
interface. For fully developed linear flow that is parallel to semi-infinite flat plates
these two flow characteristics are closely related (Bird et al., 1960; Levich, 1962),
but this is seldom the case in relatively small exposure chambers, with water en-
tering at discrete spots, and SPMDs that are mounted in widely different exposure
configurations.
A general conclusion that can be drawn from Figure 3.6 is that sampling
rates of compounds with log K ow values between 6 and 7 are in the range 2 to
12 L d−1 at flow velocities below 10 cm s−1 , with a geometric mean of 4.2 L d−1 .
These data underscore the importance of using PRCs for a site- and SPMD-specific
assessment of the effects of exposure conditions.
3.6.3. Empirical Uptake Rate Model

Booij et al. (1998) and Huckins et al. (2002a) showed that differences in
exposure conditions cause sampling rates to be shifted by a constant factor for all
compounds. Building upon this observation, and acknowledging that the log Rs -
log K ow plots in Figure 3.5 have highly similar shapes, the sampling rate of a
particular compound (i) in an exposure (j) can be written as
Ri,j = Rs,ref αi βj (3.30)
where Rs,ref is the sampling rate of a model or reference compound exposed under
standard conditions, α i is a unitless compound-specific effect and β j is a unitless
exposure-specific effect. It is of little concern which standard compound is selected
and how “standard exposure conditions” are defined, as explained below.
The relative sampling rate of a single compound (i) that is exposed under
different site conditions (j = 1, 2) equals
Ri,2 β2
= = EAF (3.31)
Ri,1 β1
where EAF is the exposure adjustment factor introduced by Huckins et al. (2002a).
In this case, both the reference sampling rate (Rs,ref ) and the compound-specific
effect α i are divided out. Huckins et al. (2002a) showed for a number of cases that
EAFs are relatively independent of the physicochemical properties of the analytes.
They therefore proposed to estimate the EAF for a specific exposure from the
ratio of PRC-based sampling rates in the field and in the laboratory. A problem
with the original EAF method is that sampling rate calibration data for PRCs and
many other analytes often are not available. As a result, the sampling rates for
these compounds have to be estimated from the listed values for compounds with
similar properties. The fact that some compounds were included in more than one
calibration study, introduces an additional ambiguity, because SPMD users may
choose one or the other calibration study as a basis for estimating the sampling
rates in the field. These ambiguities can be removed if Eq. 3.30 can be applied. In
60 Chapter 3
this case, the in situ sampling rate of a PRC is given by

RPRC,j = Rs,ref αPRC β j (3.32)
Dividing Eq. 3.30 by Eq. 3.32 gives the sampling rate of compound i in exposure j
αi
Ri,j = RPRC,j (3.33)
αPRC
The reference sampling rate (Rs,ref ) as well as the exposure-specific effect β j
are divided out. For practical applications, it therefore suffices to know how the
compound-specific effect depends on the properties of the analytes. Observing
that the experimental sampling rates have a similar dependence on log K ow , but
show a varying offset for the different studies, the log-transformed sampling rates
observed in 19 calibration experiments in 9 studies were fitted as a third order
polynomial in log K ow :
log Ri,j = 0.0130 log K ow,i
3
− 0.3173 log K ow,i
2
+ 2.244 log K ow,i + a0,j (3.34)
n = 412, s = 0.17, r = 0.91
where a0,j is the intercept that is observed for calibration experiment j. Comparing
Eq. 3.34 with the log-transformed Eq. 3.30 shows that the compound-specific
effect can be modeled by
log αi = 0.0130 log K ow
3
− 0.3173 log K ow
2
+ 2.244 log K ow (3.35)
A plot of log α i (= log Rij − a0j ) is shown in Figure 3.7. The standard devia-
tion of the fit (0.17 log units) corresponds to an uncertainty factor of about 1.5,
which is quite good considering the large differences in exposure conditions in-
volved (flow rates between 0.004 and 90 cm s−1 , temperature range 2 to 30 ◦ C).
Thus, when PRC-derived sampling rates are available, Eq. 3.35 can be used to
calculate reasonably accurate relative sampling rates for all the analytes plotted in
Figure 3.7. Box 3.1 shows an example of how to calculate in situ Rs values of a target
compound.
A brief discussion is warranted on the physical interpretation of the shape of
the regression curve in Figure 3.7. For compounds with log K ow < 4.5, sampling
rates rise with K ow , indicating that a significant portion of the overall resistance
to mass transfer lies in the SPMD membrane under the conditions of the tests
(see discussion on membrane control in Section 3.6.4.). As can be deduced from
Eqs. 3.7 and 3.8, an increase in K mw causes a decrease of the resistance to mass
transfer across the membrane, resulting in a concomitant increase in sampling
rate, until rate control switches to the WBL. Under WBL control, Rs values of
compounds with increasing K ow values are expected to gradually decline due to
small reductions in their Dw (see Section 3.6.5.). However, the apparent steepness
in the decline of Rs values is more than predicted (Huckins et al., 1998). Several
potential factors for this decline in sampling rate of high K ow compounds are
subsequently discussed.
FIGURE 3.7 Compound-specific effect (α) on the sampling rate as a function of log Kow for PCBs—
open squares: Meadows et al. (1998); PAHs—filled circles: Huckins et al. (1999), open triangles:
Luellen and Shea (2002), filled squares: Huckins et al. (2004); HCHs/chlorobenzenes—open
diamonds: Vrana and Schüürmann (2002); chlorobenzenes/PAHs/ PCBs—filled triangles: Booij
et al. (2003a); PCDDs/PCDFs/PCBs—asterisks: Rantalainen et al. (2000); and pesticides—filled
diamonds: Sabaliunas and Södergren (1997), crosses: Huckins et al. (2002b).
3.6.4. Theoretical Uptake Model: Membrane-Controlled Uptake

In order to properly interpret differences in sampling rates among compounds
and among exposure conditions, it is important to discriminate between membrane-
controlled uptake and WBL-controlled uptake. From Eq. 3.8 (in the absence of
biofouling), contaminant uptake is rate-limited by the membrane when
km K mw kw (3.36)
Because neither km nor kw are strong functions of the physical-chemical properties
of the analytes (see below), the issue of which phase controls the uptake rate is
primarily governed by the membrane-water partition coefficient, which varies be-
tween compounds by many orders of magnitude (Reynolds et al., 1990; Lefkovitz
et al., 1996; Booij et al., 2003a). With increasing log K ow , there always will be a
critical log K ow value where the uptake rates will be controlled by the WBL instead
of by the membrane. Next to K mw (which is a compound specific property) it is
important to note that rate control is also dependent on the magnitude of kw , which
is determined by the hydrodynamical conditions prevailing at the membrane-water
BOX 3.1 Example of the Calculation of Sampling Rates from a PRC-Derived
Sampling Rate, Using the Empirical Uptake Model (Eqs. 3.33 and 3.35)
Input data
Vs = 4.95 cm3
exposure time = 42 d
chrysene-d12 amounts per SPMD:
118 ng at t = 0
84 ng at t = 42 d
log K ow values:
chrysene-d12 : 5.8
pyrene : 5.2
PCB 153 : 6.9
accumulated amounts:
pyrene : 264 ng
PCB153 : 13 ng
Step 1. Calculate the Rs of the PRC
from Eq. 3.24: ke = −ln (84/118)/42 = 0.0081 d−1
from Eq. 3.28: log K sw,chrysene-d12 = 5.41
from Eq. 3.20: Rs,chrysene-d12 = 4.95 · 105.41 · 0.0081 = 10306 cm3 d−1 ≈
10.3 L d−1
Step 2. Calculate the relative Rs of analyte and PRC
from Eq. 3.35:
chrysene-d12 : log α chrysene-d12 = 4.88 ⇒ α chrysene-d12 = 75454
pyrene : log α pyrene = 4.92 ⇒ α pyrene = 82587
PCB153 : log α PCB153 = 4.65 ⇒ α PCB153 = 44419
Step 3. Calculate Rs of the analyte
from Eq. 3.33:
Rs,pyrene = Rs,chrysene-d12 · (αpyrene /αchrysene-d12 )=10.3 × 82587 ÷ 75454 = 11.3 L d−1
Rs,PCB153 = Rs,chrysene-d12 ·(αPCB153 /αchrysene-d12 ) = 10.3×44419 ÷ 75454 = 6.1 L d−1
Step 4. Calculate the aqueous concentrations
from Eq. 3.28:
log K sw,pyrene = 5.08
log K sw,PCB153 = 5.70
264
Cw,pyrene = = 0.81 ng L−1
11300 · 42
4.95 · 105.08 1 − exp −
4.95 · 105.08
13
Cw,PCB153 = = 0.053 ng L−1
6100 · 42
4.95 · 105.70 1 − exp −
4.95 · 105.70
interface. For a given log K ow range, all compounds could be membrane-controlled

at very high flow rates. The same set of compounds could also be boundary-layer
controlled at very low flow rates (e.g., quiescent conditions). Hence, specification
of the exposure conditions must be included if an investigator wants to pin-point
the critical log K ow value where rate control changes from one phase to the other.
For example, under relatively low flow-turbulence conditions (i.e., <1 cm s−1
flow velocity), rate control likely switches from the membrane to the WBL for
compounds with log K ow values in the range 4 to 4.5.
Membrane-controlled sampling rates can be modeled by
Rs = ko A ≈ km K mw A (3.37)
Because δ m is well defined, km can be written as
km = Dm /δm (3.38)
where Dm is the diffusion coefficient in the LDPE phase. On the polymer side,
diffusion rates increase with increasing segmental motility of the polymer chains
and free volume (Moisan, 1981; Lloyd and Meluch, 1985; Asfour et al., 1989;
Saleem et al., 1989; Reynolds et al., 1990). Both of these related factors control
the availability of cavities that are needed for diffusional jumps. Diffusion coeffi-
cients are inversely related to molecular volume and molecular weight (Lieb and
Stein, 1969; Möller and Gevert, 1994), but it is recognized that for a given size
class, more elongated diffusants with greater conformational freedom have larger
diffusion coefficients than rigid molecules with large cross sectional diameters
(Lloyd and Meluch, 1985; Asfour et al., 1989; Saleem et al., 1989; Reynolds
et al., 1990). Hofmans (1998) used molecular weight (M) to correlate diffu-
sion coefficients (m2 s−1 units) in LDPE obtained from five literature sources
(Figure 3.8).
log Dm = −7.47 − 2.33 log M (3.39)

n = 42, s = 0.44, 70 < M < 655
Distinguishing between rigid molecules and those with a high degree of confor-
mational freedom within this data set, showed that the more flexible compounds
had diffusion coefficients that were higher by a factor of 1.7, but this difference
was not statistically significant. Because molecular weight and log K ow are closely
related quantities for non-polar analytes, Booij et al. (2003a) argued that sampling
rates for membrane-controlled uptake can be modeled by
Rs = Akm = ABm K ow
0.682
(3.40)
where Bm is a temperature-dependent proportionality constant that is inversely

proportional to the membrane thickness. For SPMDs with a 70 µm LDPE wall
thickness these authors arrived at Bm estimates of 14 nm s−1 at 2 ◦ C and 50 nm
s−1 at 30 ◦ C.
64 Chapter 3
FIGURE 3.8 Diffusion coefficients in LDPE at 22–25 ◦ C, as a function of molecular weight. Filled
circles: Saleem et al. (1989), triangles: Reynolds et al. (1990), open circles: Möller and Gevert
(1994), asterisks: Moisan (1981).
3.6.5. Theoretical Uptake Model: WBL-Controlled Uptake

Sampling rates that are controlled by the WBL can be modeled by
Rs = k w A (3.41)
where kw is the mass transfer coefficient of the WBL. In general, kw increases when
flow rates and turbulence intensities increase. However, because of the complexity
of the flow around exposed SPMDs, it is quite difficult to estimate the magnitude
of kw from first principles, except for some simple configurations, such as SPMDs
that are mounted parallel to a fully developed unidirectional flow (Bird et al.,
1960; Levich, 1962). In a typical SPMD exposure study, however, the samplers
are mounted in a zigzag or twisted fashion with a random orientation relative to the
main water flow and placed in exposure cages that contain sharp edges and orifices.
Similar problems are often encountered by chemical engineers, when ill-defined
flows characterize the system of interest (e.g., in the case of mass transfer in reactor
beds that are packed with irregularly shaped packing material). Chemical engineers
have worked their way around the problem of determining precise hydrodynamical
conditions at solid-water interfaces by establishing semi-empirical correlations
between three dimensionless groups of variables, the Sherwood (Sh), Reynolds
(Re) and Schmidt (Sc) numbers
Sh = k d/Dw (3.42)
Re = u d/v (3.43)
Sc = ν/Dw (3.44)
where k is the mass transfer coefficient at the surface, d is a characteristic length

scale (e.g. a particle diameter), Dw was defined earlier, u is a characteristic velocity
(e.g. the interstitial flow rate, or the flow velocity at infinite distance from the
surface) and ν is the kinematic viscosity (ratio of a fluid’s viscosity to its density) of
the fluid. Typically, the Sherwood number is proportional to Sc1/3 (Bird et al., 1960;
Levich, 1962; Boudreau and Guinasso, 1982; Worch, 1993). However, Levich
(1962) argues that there are reasons to believe that Sh ∼ Sc1/4 . We find that the 1/3
power dependence occurs more frequently in the engineering literature than the 1/4
power, and will leave the discussion on this issue to the hydrodynamicists. The
typical relation between kw and the diffusion coefficient can then be summarized as
kw ∼ Dw2/3 (3.45)
Diffusion coefficients may be estimated using the Wilke-Chang equation

(Danckwerts, 1970), the Sutherland-Einstein equation (Gobas et al., 1986), or
the Hayduk-Laudie equation (Tucker and Nelken, 1982), which state that Dw
values decrease with the molar volume (Vm ) to the power 0.3 to 0.6. Alternatively,
the semi-empirical Worch relation may be used (Worch, 1993), which predicts
diffusion coefficients to decrease with increasing molar mass to the power of 0.53.
These four equations yield very similar Dw estimates (factor of 1.2 difference).
Using the Dw estimates from the most commonly used Hayduk-Laudie equation
(Dw ∼ Vm−0.589 ), and combining Eqs. 3.41 and 3.45, SPMD sampling rates for
WBL-controlled uptake would vary according to
Rs ∼ Vm−0.39 (3.46)
where Vm is the LeBas molar volume. Over the molar volume range 199 cm3
mol−1 (phenanthrene) to 480 cm3 mol−1 (fenvalerate), Eq. 3.46 predicts a
reduction in sampling rates by a factor of 1.4.
Because sampling rates are commonly given as a function of log K ow , Booij
et al. (2003a) expressed log Dw for PCBs, PAHs and chlorobenzenes as a function
of log K ow , and obtained
−0.044
Rs = ABw K ow (3.47)
where Bw is a constant for a given exposure, but may vary among exposures
according to differences in hydrodynamical conditions, and sampler geometry.
3.6.6. Combined Theoretical Uptake Model

Combining the models for membrane-controlled and boundary layer-
controlled uptake (Eqs. 3.40 and 3.47) for nonpolar compounds yields (Booij
et al., 2003a)
1
Rs = k o A = 1 1
(3.48)
0.682
+ −0.044
AB m K ow ABw K ow
66 Chapter 3
A qualitative comparison of Eq. 3.48 with the measured uptake rates (Figure 3.5)
reveals a number of features that are worth noting. First, one would expect the
sampling rates for membrane-controlled uptake to fall on the same straight line.
This expectation is met for all studies, except for the sampling rates reported by
Booij et al. (2003a). However, the slope of the log Rs versus log K ow lines attains
similar values for all studies. Second, one would also expect the sampling rates
to weakly decrease with increasing log K ow in the high log K ow range. Based on
Eq. 3.47, sampling rates at log K ow = 8 can be expected to be about 80% of the
sampling rates at log K ow = 6. This weak log K ow -dependence of the sampling
rates was observed in some studies (Huckins et al., 2002b, 2004; Booij et al.,
2003a), but in several other studies a decrease in sampling rates over this log K ow
range by a factor of 3 to 8 was observed (Meadows et al., 1998; Huckins et al.,
1999; Rantalainen et al., 2000; Luellen and Shea, 2002). A possible reason for
a large drop in measured sampling rates of very hydrophobic compounds may
be the overestimation of aqueous concentrations due to sorption to DOC. This
was first acknowledged by Meadows et al. (1998), who adopted a hypothetical
DOC concentration of 0.5 mg L−1 and assumed that the contaminant affinity for
DOC was similar to that for octanol, i.e., K DOC = K ow . Luellen and Shea (2002)
measured the DOC concentrations and applied the Burkhard equation (Burkhard,
2000) to predict DOC sorption affinities. Adopting Eq. 3.48 for the sampling rate
of truly dissolved analytes, and the Burkhard relationship for sorption to DOC, the
apparent sampling rate (Rs,app ) is given by
Rs 1
Rs,app = =
1 + [DOC]K DOC 1 1
0.682
+ −0.044
(1 + [DOC]Q K ow )
ABm K ow ABw K ow
(3.49)
where Q is dependent on DOC quality (Q = 10−1.11 ≈ 0.078 for DOC of average

quality, see Eq. 3.25). In order to check if Eq. 3.49 sufficiently explains the observed
sampling rates, this model was fitted to the calibration data from 19 experiments
in 9 studies, using log ABm , log ABw , and log Q[DOC] as adjustable parameters,
and assuming a log normal distribution of errors.
The results are summarized in Table 3.1. Inclusion of a DOC-sorption term
in the model significantly improved the Rs − K ow fit for the studies by Huckins
et al. (1999, 2004), Rantalainen et al. (2000), Meadows et al. (1998), and Luellen
and Shea (2002). The non-significance of DOC sorption in the studies by Vrana
and Schüürmann (2002) and by Sabaliūnas and Södergren (1997) may be related
to the fact that highly hydrophobic compounds were not included in these studies.
The sampling rate data for organochlorine pesticides (OCPs) by Huckins et al.
(2002b) gave no indication of sorption to DOC, even though the log K ow range for
this data set was similar to the range covered by PAHs studied by these authors
(Huckins et al., 1999). However, it should be stressed that the DOC concentra-
tion of 0.26 mg L−1 was determined in a separate sample taken from the same
TABLE 3.1 Parameter Estimates Obtained by Fitting Experimental Sampling Rates to the Membrane-WBL-DOC Model (Eq. 3.49)
Flow Rate Temperature log K ow log ABm log ABw log Q [DOC] log Q
Sources (cm s−1 ) ◦C n range (L d−1 ) (L d−1 ) [DOC] (mg L−1 ) (cm3 g−1 )
Theory and Modeling
Huckins et al. (1999) 0.004 10 16 3.5–6.9 −1.91 ± 0.10a 0.82 ± 0.04 −7.10 ± 0.19 0.26 −0.5
” 0.004 18 15 3.5–6.9 −2.36 ± 0.11 0.97 ± 0.08 −6.90 ± 0.27 0.26 −0.3
” 0.004 26 15 3.5–6.9 −2.29 ± 0.12 1.15 ± 0.11 −6.71 ± 0.28 0.26 −0.1
Vrana and Schüürmann (2002) 0.06 19 6 3.7–5.5 nsb 0.60 ± 0.14 ns — —
” 0.28 19 6 3.7–5.5 −2.17 ± 0.03 1.52 ± 0.07 ns — —
” 1.14 19 6 3.7–5.5 −2.52 ± 0.13 ns ns — —
Booij et al. (2003a) 90 2 16 3.9–7.4 −1.26 ± 0.10 2.06 ± 0.04 ns — —
” 90 13 14 3.9–7.4 −0.77 ± 0.05 2.30 ± 0.01 −8.07 ± 0.20 — —
” 90 30 17 3.9–7.4 −0.70 ± 0.05 2.49 ± 0.03 ns — —
Rantalainen et al. (2000) 8 11 23 6.4–8.2 ns 0.60 ± 0.06 −7.86 ± 0.22 — —
” 8 19 23 6.4–8.2 ns 0.83 ± 0.03 −8.49 ± 0.23 — —
Huckins et al. (2004) 0.1 16 32 3.1–6.8 −2.06 ± 0.08 0.63 ± 0.03 −6.94 ± 0.21 0.26 −0.4
Meadows et al. (1998) 0.004 12 73 5.0–7.8 ns 1.08 ± 0.02 −7.09 ± 0.08 <1 > −1.1
Sabaliūnas and Södergren (1997) 0.006 ? 6 5.0–6.2 0.85 ± 0.16 ns — —
Luellen and Shea (2002) 50 25 47 3.4–7.6 −1.27 ± 0.20 0.92 ± 0.02 −6.88 ± 0.09 0.35 −0.4
” 0.01 25 32 3.4–7.6 ns 0.86 ± 0.04 −6.32 ± 0.14 1.29 −0.4
Huckins et al. (2002b) 0.004 10 22 3.7–6.9 −2.69 ± 0.06 0.71 ± 0.03 ns 0.26 —
” 0.004 18 20 3.7–6.9 −2.24 ± 0.11 0.66 ± 0.03 ns 0.26 —
” 0.004 26 23 3.7–6.9 −2.19 ± 0.08 1.07 ± 0.04 ns 0.26 —
a
Standard error of the estimated parameter.
b
Parameters that did not significantly ( p < 0.05) improve the quality of the fit are listed as “ns”.
67
68 Chapter 3
water source (a deep well), after the exposure experiments had finished. The DOC
quality can be estimated for those studies where DOC concentrations are reported.
Values of log Q fall within the 95% confidence range of log Q values (−2.4 to
+0.2) reported by Burkhard (2000). Although the evidence is indirect, sorption
to DOC may have caused an underestimation of the sampling rates of highly
hydrophobic compounds in many studies. An underestimation of the sampling rate
by a factor of 2 occurs when Q [DOC] K ow = 1. Inspection of Table 3.1 shows
that sampling rates of compounds with log K ow values > 6.3 to 8.5 may have been
underestimated, depending on the experiment. Thus, investigators should consider
including methods for independently measuring the extent of sorption to DOC
in future calibration experiments. A potential approach would be to measure the
water solubility enhancement of the most hydrophobic analyte of the calibration
set, relative to that of ultra-pure water.
An alternative explanation for the lower sampling rates of very hydrophobic
compounds is that the membrane may become rate limiting again for compounds
with large molar volumes and low conformational freedom, due to the fact that
molecular size may be too large to fit into the transient cavities in the LDPE.
Combining Eqs. 3.36 and 3.38 gives the condition for membrane controlled uptake
Dm K mw
km K mw = kw (3.50)
δm
Diffusion coefficients in LDPE steadily decrease with increasing molecular weight
(Figure 3.8), and LDPE-water partition coefficients increase with increasing
molecular weight. In general, the increase in K mw with molecular size is much
larger than the decrease in Dm . In the case of acenaphthene and benzo[a]pyrene,
for example, Dm can be expected to decrease by a factor of about 3 (Eq. 3.39),
but this decrease is more than offset by a 350 fold increase in K mw (Booij et al.,
2003a). Although membrane-controlled uptake cannot be excluded for very large
hydrophobic molecules, the drop in Dm and/or K mw for these compounds would
have to be very sharp indeed.
3.6.7. Implications of DOC Sorption in the Lab for

In Situ Sampling Rate Estimates
The underestimation of laboratory-derived sampling rates due to analyte sorp-
tion to DOC has no counterpart in PRC-derived sampling rates in the field, since
aqueous PRC concentrations are zero anyway, and sorption to DOC would not
change that. This implies that PRC-derived sampling rates in the field may be
more accurate than laboratory-derived sampling rates for highly hydrophobic com-
pounds (log K ow > 7). For typical SPMD deployments without intermediate sam-
pling, PRC-derived sampling rates can only be estimated for a limited number of
PRCs (log K ow ≈ 5 to 6, for a typical exposure period of 3 to 6 weeks). PRCs with
low K ow values will have dissipated to below the detection level. For highly hy-
drophobic PRCs, the dissipated amount is usually too small to quantify. This raises
the question of how PRC-based sampling rates should be extrapolated into the high
and low log K ow range. Verweij et al. (2004) argued that for WBL-controlled up-
take, the decrease in sampling rates with molecular size is small enough to be
neglected. Booij et al. (2003a, 2003b) proposed to calculate Rs values from the
PRC-derived sampling rate using an equation similar to Eq. 3.46 when the ex-
change kinetics of PRCs and analytes are WBL-controlled. This procedure relies
on theoretical considerations and on the observation that accounting for sorption
to DOC may bridge the gap between theory and experimental sampling rates, as
indicated above. Therefore, the sampling rate of a WBL-controlled PRC may be
used to calculate Rs values for more hydrophobic analytes, by adopting Eq. 3.46.

VPRC 0.39
Rs = Rs,PRC (3.51)
V
Subsequent application of these estimated Rs values in Eq. 3.22 then allows for
estimating the aqueous concentrations (Box 3.2).
Sampling rate extrapolation into the low log K ow range is more difficult, be-
cause of the increasing resistance of the membrane, which causes the sampling
rates to fall below the values that are predicted by Eq. 3.51. Fortunately, this extrap-
olation is less critical, because compounds that are less hydrophobic than the PRCs
typically have attained a substantial degree of equilibrium. As a result, aqueous
concentration estimates for these compounds are quite insensitive to uncertainties
in the sampling rates. Alternatively, when the log K ow interval between successive
PRCs is small, the degree of equilibrium attained by analytes with intermediate
log K ow values may be obtained by interpolation. The aqueous concentrations may
be subsequently calculated from the partition coefficients and corrected for partial
equilibrium attainment.
3.6.8. Biofouling
The thickness of the biofilm on exposed SPMDs varies not only from ex-
posure to exposure, but varies from spot to spot on the same membrane as well.
Biofilms as thick as about 1 mm have been observed on SPMD membranes in
extended (>30 d) warm water exposures. The composition of biofilms can vary
significantly, depending on the aquatic system. For example, the biofilm may not
only consist of periphytic communities, but may contain imbibed particles and
mineral precipitates as well. Huckins et al. (1997) determined PAH sampling rates
using SPMDs that were heavily biofouled in a control pond before use, relative
to unfouled SPMDs. Sampling rate ratios (fouled divided by unfouled) ranged
between 0.7 at log K ow = 4 to 0.3 at log K ow = 7 (Figure 3.9). As far as organic
contaminant transport is concerned, a biofilm may be viewed as a layer of im-
mobilized water with dispersed organic carbon sorption sites. From Eq. 3.8, the
conductivity of the biofilm is given by
1/Ib = kb K bw (3.52)
70 Chapter 3
BOX 3.2 Example of the Calculation of Sampling Rates from a PRC-Derived

Sampling Rate, Using the WBL-Model (Eq. 3.51)
Input data
Vs = 4.95 cm3
chrysene-d12 amounts per SPMD:
118 ng at t = 0
84 ng at t = 42 d
log K ow values:
chrysene-d12 : 5.8
pyrene : 5.2
PCB 153 : 6.9
LeBas volume (cm mol−1 ):
3
chrysene-d12 : 251
pyrene : 214
PCB153 : 310
accumulated amounts:
PCB 153 = 13 ng
pyrene = 264 ng
Step 1. Calculate the Rs of the PRC
from Eq. 3.24: ke,chrysene-d12 = − ln(84/118)/42 = 0.0081 d−1
from Eq. 3.28: log K sw,chrysene-d12 = 5.41
from Eq. 3.20: Rs,chrysene-d12 = 4.95 · 105.41 · 0.0081 = 10306 cm3 d−1 ≈
10.3 L d−1
Step 2. Calculate the Rs of the analytes
from Eq. 3.51:
Rs, pyrene = 10.3(251 ÷ 214)0.39 = 11.0 L d−1
Rs, PCB153 = 10.3(251 ÷ 310)0.39 = 9.5 L d−1
Step 3. Calculate the aqueous concentrations
from Eq. 3.28:
log K sw, pyrene = 5.08
log K sw, PCB153 = 5.70
264
Cw,pyrene = = 0.82 ng L−1
11000 · 42
4.95 · 105.08 1 − exp −
4.95 · 105.08
13
Cw,PCB153 = = 0.035 ng L−1
9500 · 42
4.95 · 105.70 1 − exp −
4.95 · 105.70
FIGURE 3.9 Ratio of sampling rates with biofouled and non-biofouled SPMDs as a function of
log Kow .
Because of the similarity of transport in biofilms and in stagnant sediments,

information on the parameters that control the conductivity of the biofilm can
be obtained from diagenetic models for contaminant diffusion in pore waters.
Assuming that molecular diffusion is the dominant transport mechanism, and that
instantaneous sorption equilibrium exists between dissolved and particle-bound
solutes, the vertical flux ( j) through a stagnant sediment is given by (Berner, 1980)
Dw dC
j =φ (3.53)
θ (1 + K ) dz
where φ is the porosity, θ is the tortuosity of the sediment diffusional pathways,
Dw is the molecular diffusion coefficient in particle-free water and K is the
bulk sediment-water partition coefficient, defined as the concentration ratio of
sorbed and dissolved solutes (both on a bulk sediment volume basis). Since K bw
is defined as the concentration ratio of sorbed plus dissolved contaminants (on
a biofilm volume basis) and dissolved contaminants (on a particle-free water
volume basis), the factor 1 + K is given by
1 + K = K bw /φ (3.54)
Combining Eqs. 3.53 and 3.54, and assuming a linear concentration gradient over
a biofilm with thickness δ b , the flux through the biofilm can be written as
φ 2 Dw
j= C = kb C (3.55)
θ K bw δb
72 Chapter 3
From Eq. 3.55, the conductivity of the biofilm is given by

1 φ 2 Dw
= (3.56)
Ib θ δb
Equation 3.56 indicates that the biofilm essentially behaves like an immobilized
water layer, with a resistance that is independent of the biofilm-water partition
coefficient. Evidently, when the growth rate of the biofilm and the diffusion rate
of the contaminants are of similar magnitude, this highly idealized model breaks
down, and it can be expected in those cases that highly hydrophobic compounds
will have more difficulty in reaching the membrane than less hydrophobic (more
mobile) compounds. Also, Eq. 3.56 will likely fail to predict solute transport in
biofilms with sizable populations of invertebrates because of bioturbation.
The sampling rate reduction that is shown in Figure 3.9 is relatively inde-
pendent of the hydrophobicity for compounds with log K ow values between 4.5
and 7, in accordance with Eq. 3.56. The initial rapid fall in the magnitude of the
Rs ratios of PAHs up to log K ow = 4.5 is indicative of the reduced importance of
membrane-controlled uptake and an increasing dependence on WBL and biofilm
control. The 3-fold reduction in uptake rates due to biofouling clearly is impor-
tant. It can be expected that the dissipation of PRCs is similarly reduced, although
only very limited experimental evidence is available. Huckins et al. (1994, 2002b)
showed for biofouled SPMDs that PRC-based sampling rates, obtained from the
dissipation of perdeuterated acenaphthene, phenanthrene, and pyrene, were within
1.5 fold of the Rs values of the corresponding native compounds.
3.7. PORE WATER SAMPLING
Investigators have used SPMDs buried in fresh water sediments and shore
soils to determine the relative contamination by chlorinated hydrocarbon contam-
inants (Rantalainen et al., 1998, 2000) and PAHs (Williamson et al., 2002). Ranta-
lainen et al. (2000) have suggested that, at the points of contact between the SPMD
membrane and sediment particles, the aqueous film thickness is very small, giving
rise to high initial uptake rates. Huckins et al. (1996) and Rantalainen et al.(2000)
reported that the initial rapid rise in SPMD concentrations (<1 week), is followed
by a less rapid linear uptake of hydrophobic chemicals by buried SPMDs. When
mixing of particles and pore water by waves, currents, and biota can be neglected,
transport of solutes from the sediment to the SPMD surface takes place by molec-
ular diffusion only. In the initial stages of the uptake, the contaminant distribution
in sediment and pore water is relatively homogeneous. During the uptake process,
dissolved contaminants are removed from the pore water in the immediate vicin-
ity of the SPMD surface. The resulting concentration gradients induce desorption
from contiguous particles, and trigger diffusional fluxes from the pore waters at
larger distances from the SPMD surface (Booij et al., 2003b). These processes
result in a relatively slow but progressive contaminant depletion of both the pore
water and the particulate phase. This causes the contaminants to be transported
to the membrane over increasingly large distances, preventing the establishment
of a steady-state flux. As a result, the uptake rates decrease with time, even when
the concentrations in the SPMD are far below their equilibrium value. The growth
of the effective thickness of the WBL with time is counteracted by contaminant
desorption from the particulate phase. Sediments with high sorption capacities
will be able to efficiently replenish the pore water, whereas sediments with low
sorption coefficients will quickly be depleted themselves. A complicating factor
in this respect is that desorption rates may become rate limiting once the quickly
equilibrating fractions of the sediment have been depleted (Cornelissen et al., 1998,
2000, 2001; ten Hulscher et al., 1999).
Booij et al. (2003b) made an effort to model contaminant uptake by buried
passive samplers. The major assumptions underlying this model are that the sam-
pler can be regarded as an infinite sink for target contaminants, that the depletion
of the bulk sediment phase is insignificant, and that the contaminant desorption
kinetics are not rate-limiting.
√
2φ ACw Dw (1 + K ) √
N= √ t (3.57)
πθ
where A is the SPMD surface area, φ is the porosity, θ was defined earlier, Dw is
the molecular diffusion coefficient in particle-free water and K is defined as the
concentration ratio of sorbed and dissolved solutes (both on a bulk sediment volume
basis). This model predicts that the absorbed amounts increase with the square
root of time, and with the square root of K . Thus, aqueous concentrations can in
principle be calculated from the amount absorbed by SPMDs when the porosity,
torturosity, and molecular diffusion coefficients are known. Unfortunately, the
calculations require a number of estimates or measurements, and are rather tedious.
In addition, the requirement that the concentrations in SPMDs are far below their
equilibrium values is not always satisfied for compounds with low and intermediate
log K ow values. For these reasons, Booij et al. (2003b) recommended that SPMDs
be incubated in slowly stirred sediment slurries. In this case, the modeling of
contaminant uptake is the same as for uptake from the water phase (Eq. 3.13),
although the reported sampling rates are much higher than with exposures to
particle-free water (≈30 − 300 L d−1 for 460 cm2 samplers).
The assumptions that the depletion of the sediment phase is insignificant,
and that the contaminant desorption kinetics are not rate-limiting for exposures to
sediment slurries, are only valid if some critical conditions regarding experimental
design are met. For compounds that attain equilibrium, the total amount in the
SPMDs should be much smaller than the total amount in the sediment phase. This
condition can be expressed as
Vs Cs m oc Coc f r (3.58)
where m oc is organic carbon mass of sediment present in the incubation vessel,
Coc is the contaminant concentration in the sediment phase on an organic carbon
74 Chapter 3
basis, and f r is the contaminant fraction that quickly (i.e. on the time scale of the
experiment) equilibrates with the pore water. Introducing the partition coefficients
of sediment organic carbon-water (K oc ) and K sw , and rearranging gives a criterion
for the ratio of SPMD volume to organic carbon mass
Vs K oc
fr (3.59)
m oc K sw
Since the right hand side of Eq. 3.59 would typically approximate 1, a Vs /m oc
ratio of 0.05 mL g−1 seems to be a safe choice.
For compounds that do not reach a significant degree of equilibrium during
the exposure, the absorption rate by the SPMD should be much smaller than the
desorption rate by the sediment (Booij et al., 2003b)
Rs m oc k2 K oc (3.60)
where k2 is the first-order desorption rate constant of the sediment phase, which
typically attain values of 0.1–1 h−1 and 10−3 h−1 for the rapidly and the slowly
equilibration sediment fraction, respectively (Cornelissen et al., 1998, 2000, 2001;
ten Hulscher et al., 1999). Based on a number of model calculations, using literature
values for k2 and K oc , Booij et al. (2003b) recommend that sampler volume to
organic carbon mass ratios of 0.05 mL g−1 would be a safe choice in this case as
well.
3.8. GROUNDWATER SAMPLING
Considerations for the deployment of SPMDs in groundwater wells have been

described by Gustavson and Harkin (2000), who also presented a comparison of
SPMD-derived and batch extraction-derived PAH concentrations in groundwater.
The sampling of water in subterranean strata with SPMDs is generally straightfor-
ward, as environmental conditions are usually more constant than surface water,
and biological growths on the SPMD membrane surface are minimal. However,
permeability in fine-grained strata can be very low, which may result in the de-
pletion of target solutes at the membrane surface in a similar fashion as with the
exposure of SPMDs in stagnant sediments. In this case, the uptake rate is limited
by the groundwater flow. Groundwater fluxes can be estimated from Darcy’s Law,
which links well flow rates (Fw ; volume per time) or recharge rates to the local
pressure gradients and the hydraulic permeability/conductivity (Phc ; distance per
time) of the strata (Gustavson and Harkin, 2000).
Fw = Phc Ah/L f (3.61)
where h is the difference in the hydraulic head over the water bearing strata,
and L f is the length of the strata. When SPMDs are placed in wells, Fw should
be compared to the sampling rate of the analyte of interest. If Fw Rs , then the
methods described for SPMDs exposed to HOCs in surface waters can be applied.
When Fw Rs, groundwater sampling is essentially equivalent to the exposure

of an SPMD in a small volume of well water in a closed system (Vgw ). Assuming
that residue contributions from suspended particulates or colloids and dissolution
of non-aqueous phase liquids is insignificant, the aqueous concentrations may be
calculated from
Cw = N /Vgw (3.62)
When Fw is not known, using both Eqs. 3.22 and 3.62 to compute groundwater
concentrations of analytes, provides an investigator with defined limits of the range
of potential analyte concentrations.
3.9. AIR SAMPLING
Because processes such as diffusion and partitioning are fundamentally the

same in water and air, equations that describe uptake from air can be obtained by
replacing the subscripts in equations for water with those appropriate for air

Rs t
Cs = K sa Ca 1 − exp − (3.63)
Vs K sa
Rs = Vs K sa ke (3.64)
Ako
ke = (3.65)
Vs K sa
1 1 1 1
= + + (3.66)
ko ka km K ma kLd K Lda
where all parameters have the same meaning as described earlier, and the subscript
“a” and “Ld” refer to the air phase and the lipid derived film on the exterior surface
of SPMDs exposed for extended periods, respectively. Physical chemical properties
of organic contaminants differ widely for water and air. Diffusion coefficients in
air are larger than in water by about four orders of magnitude, and SPMD partition
coefficients for air are two to three orders of magnitude higher than for water.
Flow velocities in air (1–20 m s−1 ) are typically much higher than in water (0.001–
1 m s−1 ), and air and water viscosities differ by a factor of about 60. Because of these
differences, the general lack of biofouling, and the common presence of a surficial
lipid-derived film, a separate discussion on air sampling by SPMDs is required.
3.9.1. SPMD-Air Partition Coefficients

No published values of SPMD-air partition coefficients (K sa ) exist. These
values therefore have to be calculated from published values of K sw and Henry’s
law constants (H ) using
K sw RT
K sa = (3.67)
H
76 Chapter 3
where R is the gas law constant, and T is the absolute temperature. Henry’s law
constants (Pa m3 mol−1 ) are available for a wide range of contaminants at various
temperatures (ten Hulscher et al., 1992; Alaee et al., 1996; de Maagd et al., 1998;
Paasivirta et al., 1999; Bamford et al., 1999, 2002; Shiu and Ma, 2000; Staudinger
and Roberts, 2001; Sander, 2003; US EPA, 2003), including a large number of polar
pesticides. Sometimes, Henry’s law constants are given in the dimensionless form
(H/RT ), which is the air-water partition coefficient. The temperature dependence
of H is well-documented for nonpolar compounds like chlorobenzenes, PCBs,
PAHs, PCDDs, PCDFs, and the classical OCPs like DDTs, chlordanes, and HCHs.
Typically, values of H increase by 0.3 log units (range 0.06 to 0.6 log units), when
the temperature increases from 10 to 20 ◦ C. Bearing in mind that K sw is virtually
temperature-independent, a decrease in K sa values by a factor of about 2 (range 1.1
to 4) can be expected for each 10 ◦ C temperature increase, due to the temperature
dependence of H .
3.9.2. Air Sampling Rates

In a number of field studies, sampling rates were obtained by parallel deploy-
ment of SPMDs and high-volume active air samplers (HiVols) for PCBs (Petty
et al., 1993; Prest et al., 1995; Ockenden et al., 1998; Shoeib and Harner, 2002),
and polychlorinated naphthalenes (Shoeib and Harner, 2002). Lohman et al. (2001)
reported sampling rates of PAHs and PCDDs/PCDFs that were based on SPMD
and HiVol data obtained during different periods of the year. Laboratory-derived
sampling rates have been reported for PCBs (Huckins et al., 1994), and for PAHs
and a number of moderately polar pesticides (Robertson, 2004). In a similar labo-
ratory experiment, Huckins et al. (2001) followed the dissipation of phenanthrene,
diazinon and chlorpyrifos at two flow velocities and two temperatures. In addition,
Ockenden et al. (2001) compared the kinetics of SPMD-air equilibration of native
PCBs and the dissipation rate of 13 C-labeled PCBs, but no sampling rates were
reported due to the absence of HiVol data. A summary of SPMD sampling rates is
shown as a function of the octanol-air partition coefficient in Figure. 3.10. Log K oa
values were adopted as specified in Appendix A (Tables A.13–A.16). Sampling
rates could be described by
log Rs = 0.154 log K oa − 0.80 (3.68)

n = 75, s = 0.20, r = 0.64
No large variation in sampling rates is observed among the different studies, despite
differences in exposure conditions, such as wind speeds, temperature, and SPMD
mounting layout. It should be noted, however, that the effect of temperature is
partially accounted for by our use of temperature-corrected log K oa values. An
example of the application of Eq. 3.68 for calculating atmospheric concentrations
is given in Box 3.3.
FIGURE 3.10 Experimental air sampling rates as a function of log Koa for PCBs—open circles:
Ockenden et al. (1998), diamond: Petty et al. (1993), closed circles: Huckins et al. (1994), closed
triangles: Shoeib and Harner (2002); polychlorinated naphthalenes—open triangles: Shoeib
and Harner (2002); PAHs—closed squares: Robertson (2004); and pesticides—open squares:
Robertson (2004).
The small value of the log Rs versus log K oa slope indicates that the uptake is
partially membrane controlled and partially air boundary layer (ABL) controlled.
Applying the same line of reasoning as for SPMD-water exchange, membrane-
controlled uptake would result in log Rs − log K oa slopes close to (but smaller than)
one. On the other hand, ABL controlled sampling rates would be proportional to the
compound’s diffusion coefficient to the power of 2/3. Since diffusion coefficients
in air are only weak functions of molecular size (Tucker and Nelken, 1982; Shoeib
and Harner, 2002), a log Rs − log K oa slope of about 0 would be expected in this
case. Because the slope obtained in Eq. 3.68 is between these two limiting slopes,
we suggest that the uptake generally is partially controlled by the membrane, and
partially by the ABL. However, the fine structure in the individual data series is
worth noting. Sampling rates reported by Ockenden et al. (1998) seem to level
off to a constant value for log K oa > 9.5 in the 14 ◦ C, 18 ◦ C series, and for
log K oa > 11 in the 4 ◦ C series. For the data reported by Huckins et al. (1994)
a less steep slope is observed for log K oa > 9. A similar trend was observed for
sampling rates of PAHs and moderately polar pesticides, which level off to a nearly
constant Rs value for compounds with log K oa > 9, albeit with a relatively large
78 Chapter 3
BOX 3.3 Example of the Calculation of Atmospheric Concentrations of Vapor-

Phase Compounds Using Eq. 3.68
Input data
Vs = 4.95 cm3
exposure temperature 4 to 16 ◦ C
log K ow = 5.5
log K oa = −6.3 + 3928T−1 (Harner and Mackay, 1995)
Henry’s law coefficient at 20 ◦ C 41 Pa m3 mol−1 (ten Hulscher et al., 1992)
water-air transfer enthalpy: 49 kJ mol−1 (ten Hulscher et al., 1992)
HCB accumulated per SPMD = 107 ng
Step 1. Calculate Rs
estimate log K oa at the average exposure temperature of 10 ◦ C, using the data
given by Harner and Mackay (1995). log K oa (283 K) = 7.6
from Eq. 3.68: log Rs = 7.6 * 0.154 − 0.80 = 0.37 ⇒ Rs = 2.3 m3 d−1
Step 2. Calculate K sa
from Eq. 3.28: log K sw = 5.3
from ten Hulscher et al. (1992):
ln H (283 K) = ln H (293 K) − 49000/8.314 (1/283 − 1/293) = 3.0
H = 20 Pa m3 mol−1
from Eq. 3.67 : log K sa = 7.4
Step 3. Calculate Cair
107
Ca,HCB =
2.3 · 106 50
4.95 · 10 7.4
1 − exp −
4.95 · 107.4
= 1.4 · 10 ng cm = 1.4 ng m−3
−6 −3
The exponential factor (−0.92) indicates that HCB has attained [1−
exp(−0.92)] · 100% = 60% of its equilibrium value.
scatter (Robertson, 2004). Sampling rates reported by Shoeib and Harner (2002)
are relatively constant in the range of 8 < log K oa < 10, and show a modest increase
at log K oa > 10. Again, the constancy of sampling rates with log K oa is indicative of
ABL controlled uptake. Huckins et al. (2001) showed that the ke of phenanthrene
and diazinon increased by a factor of 2.2 and 1.5, respectively, when the air-flow
rate increased from <15 cm s−1 to about 60 cm s−1 . The suggestion that sampling
rates are under partial membrane control and boundary layer control is supported
by the Ockenden et al. (2001) observation that PCB amounts sampled by shielded
SPMDs was slightly less than for fully exposed SPMDs (smaller than a factor of
1.5, depending on the compound). These authors conclude that the effect of wind
speed on the sampling rates is insufficient to explain the earlier observation by
Ockenden et al. (1998) that sampling rates were higher in winter than in summer,
and that the increase of sampling rates at lower temperature implies membrane-
controlled uptake.
The temperature dependence of Rs yields additional information on the issue
of whether the membrane or the ABL controls the uptake rates. In the case of
membrane-controlled uptake, the sampling rates are given by
Akm K mw RT
Rs ≈ Akm K ma ≈ (3.69)
H
Activation energies for membrane-controlled uptake rates from water attain values
of about 20 kJ mol−1 (Figure 3.4). Because membrane-controlled water sampling
rates are proportional to km K mw , and because the temperature dependence of the
LDPE-water partition coefficient (K mw ) is quite weak (≈−0.3 kJ mol−1 , Booij
et al., 2003a), these activation energies are almost entirely related to km . For
that reason, the value of km derived from water sampling rates can be adopted
for km when analyzing air sampling rate data. This logic indicates that km val-
ues increase by a factor of about 1.3 when the temperature increases from 10 to
20 ◦ C in air exposures. This increase in km is offset by a decrease of K ma , caused
by the increase of Henry’s law constant, as illustrated for PCB congeners 28 and
153. These compounds have water-air transfer enthalpies of 41 and 27 kJ mol−1 ,
respectively (Bamford et al., 2002), indicating that a temperature increase from
10 to 20 ◦ C causes H to increase by about 1.8 fold for PCB 28 and by about 1.5
fold for PCB 153. A 10 ◦ C temperature increase therefore causes both km and
H to increase. As a result, sampling rates for PCB congeners 28 and 153 can be
expected to change by a factor of 1.3/1.8 = 0.7 and 1.3/1.5 = 0.9, respectively, if
the uptake would be membrane-controlled for these compounds.
Sampling rates for the case of total boundary layer-control can be expected to
be nearly independent of temperature, since both the diffusion coefficients in air,
and the kinematic viscosity of air are only weak functions of temperature (Shoeib
and Harner, 2002). This leaves the air-flow velocity as the major factor that can
be responsible for the seasonal differences among sampling rates observed by
Ockenden et al. (1998). The absence of large Rs differences between indoor and
outdoor exposures may be indicative of membrane-control, but it may also reflect
the efficient damping of high flow velocities by the deployment devices used for
SPMD air exposures (Ockenden et al., 2001).
Summarizing, some of the evidence indicates membrane-controlled uptake.
Other evidence suggests ABL controlled uptake. However, the reasonably small
variance in the available sampling rates obtained under widely differing flow and
temperature conditions suggests that relatively accurate Rs values for PCBs and
related compounds may be estimated from Eq. 3.68.
80 Chapter 3
A more complicated situation exists for particle-associated contaminants.

Lohman et al. (2001) report that predominantly (>95%) particle-associated con-
taminants (5- and 6-ring PAHs, hepta- and octachloro dibenzo- p-dioxins) are effi-
ciently sampled by SPMDs. This observation has been supported by other research
(Bartkow, 2004). In both studies, it appears that the surficial film was not removed
prior to dialytic recovery of analytes. Particle sampling is likely mediated by the
presence of a sticky triolein related film on the exterior of SPMDs exposed to air.
The time-dependent development of this film has been documented by Petty et al.
(1993) for SPMDs that contained 95% pure triolein. After a 28 day exposure to
indoor air at 25 ◦ C, the mass of the exterior film amounted to 3% of the triolein
mass. The authors suggest that it likely consisted of lower molecular weight tri-
olein impurities, such as methyl oleate, oleic acid, and glyceryl (di-)oleate. Lebo
et al. (2004) have subsequently shown that 4% of the impurities in one lot of 95%
triolein was methyl oleate. Also, unsaturated lipids are known to oxidize over time
and reaction rates are enhanced by light (Dobarganes and Marquez-Ruiz, 1998).
The oxidation of unsaturated lipids generally produces epoxides, ketones, and in
the case of triglycerides, monomers, dimers and oligomers can be produced as
well. The reactivity of unsaturated lipids on the exterior surface of the membrane
is expected to be much higher than for membrane enclosed triolein, because sup-
plies of oxygen and water are not limited by membrane permeation. In regard to
membrane-enclosed lipids, exposure temperature is a critical parameter in that it
affects both reaction rates and the diffusivity of oxidation products through the
membrane. The availability of oxygen in the SPMD interior during air exposures
may be a limiting factor for several days, since the permeability of oxygen in LDPE
has been shown to be relatively low (Pauly, 1989; Divine and McCray, 2004). To
our knowledge, no published data exists on the potential for the autoxidation of
LDPE enclosed triolein during long-term air exposures. However, we report that
82 d air exposures (≈25 ◦ C) of standard SPMDs (A = 460 cm2 ) containing 1-mL
of high purity triolein (Lebo et al., 2004) in a dark room, resulted in only a very
slight surficial film with a residue mass of ≈13 mg ± 2.5 (n = 4). These data
suggest that autoxidation of membrane enclosed high purity triolein (99%) does
occur very slowly, but for exposures of moderate duration (<60 d) and tempera-
ture (≤25 ◦ C), and in the absence of light, oxidation products of triolein should
contribute only slightly to the development of a surficial film.
Although Rs values of high K sa compounds derived from Eq. 3.68 may have
been partly influenced by particle sampling, it is unlikely that the equation can
accurately predict the summed vapor plus particulate phase concentrations, be-
cause transport rates through the boundary layer and through the membrane are
different for the vapor-phase fraction and the particle-bound fraction, due to dif-
ferences in effective diffusion coefficients between molecules and small particles.
In addition, it will be difficult to define universally applicable calibration curves
for the sampling rate of total (particle + vapor) atmospheric contaminants. At
this stage of development, results obtained with SPMDs for particle-associated
compounds provides valuable information on source identification and temporal
trends (Lohman et al., 2001), but these results can only be considered to be quali-
tative or semi-quantitative. However, Eq. 3.68 can be used to set an upper limit to
vapor-phase concentrations.
An additional complicating factor with respect to air sampling rate calibra-
tion is the uncertainty associated with determinations of vapor-phase and particle
sorbed concentrations of analytes by HiVol sampling. These systems suffer from
artifacts such as the volatilization of particle-bound contaminants, insufficient re-
tention of small particles, and adsorption of vapor-phase contaminants on the GFFs
(Ockenden et al., 1998; Lohman et al., 2001). These artifacts may cause concen-
trations in the vapor-phase to be overestimated or underestimated, which results
in sampling rates that are too low or too high.
3.9.3. The Use of PRCs in Atmospheric Sampling

The use of PRCs in atmospheric sampling has been fairly limited. Booij and
van Drooge (2001) have used 1,3,5-trichlorobenzene, and the PCB congeners 4,
29, 155, and 204 as PRCs in SPMD exposures to coastal air for three weeks.
No measurable loss was observed for PCB congeners 155 and 204. Ockenden
et al. (2001) studied the dissipation of 13 C-labelled PCBs (congeners 28, 52, 101,
138, 153, and 180) for 120 days. The ratios of PRC-based sampling rates and the
sampling rates of native PCBs were in the range 0.8 for PCB 28 to 3.6 for PCB
153. No measurable losses were observed for 13 C-labelled PCB congeners 138
and 180. Although the calibration data available give no reason to assume that
sampling rates need a correction for differences in exposure conditions, the use of
PRCs may provide a useful check on the correctness of the proposed calibration
model. In addition, the use of PRCs may allow for correcting the results obtained
after more calibration data becomes available.
3.10. REFERENCES
Alaee, M.; Whittal, R.M.; Strachan, M.J. 1996, The effect of water temperature and composition on
Henry’s law constants for various PAH’s. Chemosphere 32: 1153–1164.
Asfour, A.-F.; Saleem, M.; De Kee, D. 1989, Diffusion of saturated hydrocarbons in low density
polyethylene (LDPE) films. J. Appl. Polym. Sci. 38: 1503–1514.
Bamford, H.A.; Poster, D.L.; Baker, J.E. 1999, Temperature dependence of Henry’s law constants of
thirteen polycyclic aromatic hydrocarbons between 4 ◦ C and 31 ◦ C. Environ. Toxicol. Chem. 18:
1905–1912.
Bamford, H.A.; Poster, D.L.; Huie, R.E.; Baker, J.E. 2002, Using extrathermodynamic relationships
to model the temperature dependence of Henry’s law constants of 209 PCB congeners. Environ.
Sci. Technol. 36: 4395–4402.
Banerjee, S. and Baughman, G.L. 1991, Bioconcentration factors and lipid solubility. Environ. Sci.
Technol. 25: 536–539.
Bartkow, M. 2004, ENTOX, Queensland, Australia. Personal communication.
Berner, R.A. 1980, Early Diagenisis. A Theoretical Approach. Princeton University Press: Princeton,
NJ.
82 Chapter 3
Bird, R.B.; Stewart, W.E.; Lightfoot, E.N. 1960, Transport Phenomena. John Wiley & Sons: New York,
NY.
Booij, K. and van Drooge, B.L. 2001, Polychlorinated biphenyls and hexachlorobenzene in atmosphere,
sea-surface microlayer, and water measured with semi-permeable membrane devices (SPMDs).
Chemosphere 44: 91–98.
Booij, K.; Hofmans, H.E.; Fischer, C.V.; van Weerlee, E.M. 2003a, Temperature-dependent uptake rates
lene membranes. Environ. Sci. Technol. 37: 361–366.
Booij, K.; Hoedemaker, J.R.; Bakker, J.F. 2003b, Dissolved PCBs, PAHs, and HCB in pore waters and
Boudreau, B.P. and Guinasso, N.L. 1982, The influence of a diffusive sublayer on accretion, dissolution,
and diagenisis at the sea floor. In The Dynamic Environment of the Ocean Floor. Fanning, K.A.,
Manheim, F.T., Eds.; Lexington Books: Toronto.
Burkhard, L.P. 2000, Estimating dissolved organic carbon partition coefficients for nonionic organic
chemicals. Environ. Sci. Technol. 34: 4663–4668.
Cornelissen, G.; Rigterink, H.; Ferdinandy, M.M.A.; Van Noort, P.C.M. 1998, Rapidly desorbing
fractions of PAHs in contaminated sediments as a predictor of the extent of bioremediation.
Cornelissen, G.; Rigterink, H.; Van Noort, P.C.M.; Govers, H.A.J. 2000, Slowly and very slowly
desorbing organic compounds in sediments exhibit Langmuir-type sorption. Environ. Toxicol.
Chem. 19: 1532–1539.
Cornelissen, G.; Rigterink, H.; ten Hulscher, D.E.M.; Vrind, B.A.; Van Noort, P.C.M. 2001, A simple
Tenax extraction method to determine the availability of sediment-sorbed organic compounds.
Danckwerts, P.V. 1970, Gas-Liquid Reactions. McGraw-Hill: New York, NY.
de Maagd, P.G.J.; ten Hulscher, D.T.E.M.; van den Heuvel, H.; Opperhuizen, A.; Sijm, D.T.H.M.
1998, Physicochemical properties of polycyclic aromatic hydrocarbons: Aqueous solubilities,
n-octanol/water partition coefficients, and Henry’s law constants. Environ. Toxicol. Chem. 17:
251–257.
Divine, C.E. and McCray, J.E. 2004, Estimation of membrane diffusion coefficients and equilibration
times for low-density polyethylene passive diffusion samplers. Environ. Sci. Technol. 38: 1849–
1857.
Dobarganes, M.C. and Marquez-Ruiz, G. 1998, Analytical evaluation of fats and oils by size-exclusion
chromatography. Analusis 26: M61–65.
Flynn, G.L. and Yalkowsky, S.H. 1972, Correlation and prediction of mass transport across membranes
I: Influence of alkyl chain length on flux-determining properties of barrier and diffusant. J. Pharm.
Sci. 61: 838–852.
Gobas, F.A.P.C.; Opperhuizen, A.; Hutzinger, O. 1986, Bioconcentration of hydrophobic chemicals in
fish: Relationship with membrane permeation. Environ. Toxicol. Chem. 5: 637–646.
Gustavson, K.E. and Harkin, J.M. 2000; Comparison of sampling techniques and evaluation of semiper-
meable membrane devices (SPMDs) for monitoring polynuclear aromatic hydrocarbons (PAHs)
in groundwater. Environ. Sci. Technol. 34: 4445–4451.
Harner, T. and Mackay, D. 1995, Measurement of octanol-air partition coefficients for chlorobenzenes,
PCBs, and DDT. Environ. Sci. Technol. 29: 1599–1606.
Devices. MSc Thesis, Netherlands Institute for Sea Research: Den Burg, The Netherlands.
27: 2489–2496.
Huckins, J.N.; Petty, J.D.; Orazio, C.E.; Zajicek, J.L.; Gibson, V.L.; Clark, R.C.; Echols, K.R. 1994,
Semipermeable Membrane Device (SPMD) Sampling Rates for Trace Organic Contaminants in
Air and Water. Abstract of the 15th Annual meeting of Society of Environmental Toxicology and
Chemistry; Denver, CO.; Abstract MB01.
Huckins, J.N.; Petty, J.D.; Lebo, J.A.; Orazio, C.E.; Prest, H.F.; Tillitt, D.E.; Ellis, G.S.; Johnson, B.T.;
Manuweera, G.K. 1996, Semipermeable Membrane Devices (SPMDs) for the Concentration and
Assessment of Bioavailable Organic Contaminants in Aquatic Environments. In Techniques in
Aquatic Toxicology; Ostrander, G.K., Ed.; CRC Press: Boca Raton, FL; pp. 625–655.
Huckins, J.N.; Petty, J.D.; Prest, H.F.; Lebo, J.A.; Orazio, C.E.; Gale, R.W.; Clark, R.C. 1997, Important
Considerations in Semipermeable Membrane Devices (SPMDs) Design, Performance, and Data
Comparability. Presented at the 18th Annual Meeting of Society of Environmental Toxicology
and Chemistry; San Francisco, CA; November 16, 1997; Abstract PTP055.
Huckins, J.N.; Petty, J.D.; Gale, R.W.; Booij, K.; Prest, H.F.; Clark, R.C. 1998, Observations on
Declining SPMD Sampling Rates for High Kow Compounds. Presented at the 19th Annual Meeting
of Society of Environmental Toxicology and Chemistry; November 15, 2004; Charlotte, NC;
Abstract PWA173.
Technol. 33: 3918–3923.
Huckins, J.N.; Petty, J.D.; Robertson, G.L.; Clark, R.C.; Cranor, W.L.; Alvarez, D.A.; Gale, R.W. 2001,
Calibration of Lipid-containing Semipermeable Membrane Devices (SPMDs) for Vapor-phase
Contaminants. Presented at the 22nd Annual Meeting of Society of Environmental Toxicology
and Chemistry; Baltimore, MD; November 11, 2001; Abstract PT258.
R.C.; Mogensen, B.B. 2002a, Development of the permeability/performance reference compound
W.L.; Johnson, B.T. 2002b, A Guide for the Use of Semipermeable Membrane Devices (SPMDs)
Landrum, P.F.; Nihart, S.R.; Eadie, B.J.; Gardner, W.S. 1984, Reverse-phase separation method for
determining pollutant binding to Aldrich humic acid and dissolved organic carbon of natural
waters. Environ. Sci. Technol. 18: 187–192.
Lebo, J.A.; Almeida, F.V.; Cranor, W.L.; Petty, J.D.; Huckins, J.N.; Rastall, A.; Alvarez, D.A.;
Mogensen, B.B.; Johnson, B.T. 2004, Purification of triolein for use in semipermeable mem-
brane devices (SPMDs). Chemosphere 54: 1217–1224.
Lefkovitz, L.; Crecelius, E.; McElroy, N. 1996, The use of Polyethylene Alone to Predict Dissolved-
phase Organics in the Columbia River. Presented at the 17th annual meeting Society of Environ-
mental Toxicology Chemistry; November 17–21, 1996; Washington, D.C.
84 Chapter 3
Levich, V.G. 1962, Physicochemical Hydrodynamics. Prentice Hall, Inc.: Englewood Cliffs, NJ.
Lloyd, D.R. and Meluch, T.B. 1985, Selection and evaluation of membrane materials for liquid sep-
arations. In Materials Science of Synthetic Membranes; Lloyd, D.R., Ed.; American Chemical
Society: Washington, DC.
Lohman, R.; Corrigan, B.P.; Howsam, M.; Jones, K.C.; Ockenden, W.A. 2001, Further developments
in the use of semipermeable membrane devices (SPMDs) as passive air samplers for persistent
organic pollutants: Field application in a spatial survey of PCDD/Fs and PAHs. Environ. Sci.
Technol. 35: 2576–2582.
Luellen, D.A. and Shea, D. 2002, Calibration and field verification of semipermeable membrane
devices for measuring polycyclic aromatic hydrocarbons in water. Environ. Sci. Technol. 36:
1791–1797.
Mackay, D. 1994, Unraveling the Choreography of Contaminant Kinetics: Approaches to Quantifying
the Uptake of Chemicals by Organisms. In Bioavailability: Physical, Chemical and Biological
Interactions; Hamelink, J.L., Landrum, P.F., Bergman, H.L., Benson W., Eds.; CRC Press: Boca
Raton, FL.; pp. 171–177.
McCarthy, J. 2004, Oak Ridge National Laboratory, Oak Ridge, TN. Personal communication.
Moisan, J.Y. 1981, Diffusion des additives du polyethylene-IV. Eur. Polym. J. 17: 857–864.
Möller, K. and Gevert, T. 1994, An FTIR solid-state analysis of the diffusion of hindered phenols in
low-density polyethylene (LDPE): the effect of molecular size on the diffusion coefficient. J. Appl.
Polym. Sci. 51: 895–903.
Ockenden, W.A.; Corrigan, B.P.; Howsam, M.; Jones, K.C. 2001, Further developments in the use of
semipermeable membrane devices as passive air samplers: Application to PCBs. Environ. Sci.
Technol. 35: 4536–4543.
Paasivirta, J.; Sinkkonen, S.; Mikkelson, P.; Rantio, T.; Wania, F. 1999, Estimation of vapor pressures,
solubilities and Henry’s law constants of selected persistent organic pollutants as functions of
temperature. Chemosphere 39: 811–832.
Pauly, S. 1989, Permeability and Diffusion Data. In Polymer Handbook, 3rd ed., Brandrup, J. Immergut,
E.H., Eds.; Wiley & Sons: New York, NY; Section 6, pp. 435–449.
Petty, J.D.; Huckins, J.N.; Zajicek, J.L. 1993, Application of semipermeable membrane devices
(SPMDs) as passive air samplers. Chemosphere 27: 1609–1624.
Prest, H.F.; Jacobson, L.A.; Huckins, J.N. 1995, Passive sampling of water and coastal air via semiper-
meable membrane devices. Chemosphere 30: 1351–1361.
Rantalainen, A.-L.; Ikonomou, M.G.; Rogers, I.H. 1998, Lipid-containing semipermeable-membrane
Rantalainen, A.-L.; Cretney, W.; Ikonomou, M.G. 2000, Uptake rates of semipermeable membrane
devices (SPMDs) for PCDDs, PCDFs, and PCBs in water and sediment. Chemosphere 40: 147–
158.
Reid, R.C.; Prausnitz, J.M.; Poling, B.E. 1987, The Properties of Gases and Liquids. 4th Edition;.
McGraw-Hill: New York, NY.
Reynolds, G.W.; Hoff, J.T.; Gillham, G.W. 1990, Sampling bias caused by materials used to monitor
halocarbons in groundwater. Environ. Sci. Technol. 24: 135–140.
Robertson, G, 2004, U.S. EPA, National Exposure Research Laboratory: Las Vegas, NV; Personal
communication.
Sabaliūnas, D. and Södergren, A. 1997, Use of semipermeable membrane devices to monitor pollutants
in water and assess their effects: A laboratory test and field verification. Environ. Pollut. 96: 195–
205.
Saleem, M.; Asfour, A.A.; De Kee, D. 1989, Diffusion of organic penetrants through low density
polyethylene (LDPE) films: Effect of size and shape of the penetrant molecules. J. Appl. Polym.
Sci. 37: 617–625.
Sander, R. 2003, Compilation of Henry’s Law Constants for inorganic and organic species of potential
importance in environmental chemistry (version 3). Max Planck Institute for Chemistry, Mainz,
Germany. http://www.mpch-mainz. mpg.de/∼sander/res/henry.html (accessed November, 2004)
Shiu, W.Y. and Ma, K.C. 2000, Temperature dependence of physical-chemical properties of selected
chemicals of environmental interest. II. Chlorobenzenes, polychlorinated biphenyls, polychlori-
nated dibenzo-p-dioxins, and dibenzofurans. J. Phys. Chem. Ref. Data 29: 387–462.
persistent organic pollutants. Environ. Sci. Technol. 36:4142–4151.
Staudinger, J. and Roberts, P.V. 2001, A critical compilation of Henry’s law constant temperature
dependence relations for organic compounds in dilute aqueous solutions. Chemosphere 44: 561–
576.
ten Hulscher, Th.E.M.; van der Velde, L.E.; Bruggeman, W.A. 1992, Temperature dependence of
Henry’s Law constants for selected chlorobenzenes, polychlorinated biphenyls, and polycyclic
aromatic hydrocarbons. Environ. Toxicol. Chem. 11: 1595–1604.
ten Hulscher, Th.E.M.; Vrind, B.A.; van den Heuvel, H.; van der Velde, L.E.; van Noort, P.C.M.;
Beurskens, J.E.M.; Govers, H.A.J. 1999, Triphasic desorption of highly resistant chlorobenzenes,
polychlorinated biphenyls, and polycyclic aromatic hydrocarbons in field contaminated sediment.
Tucker, W.A. and Nelken, L.H. 1982, Diffusion coefficients in air and water. In Handbook of Chemical
Property Estimation Methods. Lyman, W.J., Reehl, W.F., Rosenblatt, D.H., Eds.; McGraw-Hill:
New York, NY.
US EPA 2003, Estimation Program Interface (EPI) Suite. Version 3.11; U.S. Environmental Protection
Agency; http://www.epa.gov/opptintr/exposure/docs/episuite.htm (accessed on November, 2004).
Verweij, F.; Booij, K.; Satumalay, K.; van der Molen, N.; van der Oost, R. 2004, Assessment of
bioavailable PAH, PCB, and OCP concentrations in water using semipermeable membrane devices
(SPMDs), sediments and caged carp. Chemosphere 11: 1675–1689.
Williamson, K.S.; Petty, J.D.; Huckins, J.N.; Lebo, J.A.; Kaiser, E.M. 2002, Sequestration of pri-
ority pollutant PAHs from sediment pore water employing semipermeable membrane devices.
Worch, E. 1993, Eine neue gleichung zur berechnung von diffusionskoeffizienten gelöster Stoffe. Vom
Wasser 81: 289–297.
Chapter 4
Study Considerations
4.1. OVERVIEW
Before deploying SPMDs, it is important that users are aware of a number of

study considerations and quality control (QC) issues. Some of the QC issues apply
to environmental sampling in general, whereas some are specifically related to
SPMDs and to a few other passive samplers. For example, SPMDs generally have
relatively large surface areas (e.g., ≈460 cm2 cm−3 of triolein) to enhance uptake
rates of solutes or vapor phase chemicals. Thus, special care must be used to prevent
contamination of the SPMDs or the loss of target compounds (e.g., through per-
vaporation or photodegradation) during SPMD assembly, transport, deployment,
recovery and storage. Potential pitfalls that specifically relate to SPMD processing,
cleanup and analysis are discussed in Chapter 5.
4.2. SOURCES OF SPMDs
The SPMD technology is the subject of two United States Government

Patents, (Huckins et al.,1992, 1995a) and a Canadian Patent (Huckins et al.,
1996). The United States Department of Commerce granted Environmental Sam-
pling Technologies (EST), 1717 Commercial Drive, St. Joseph, MO, USA ex-
clusive license to manufacture and sell SPMDs in the USA. The license also
covers the organic solvent dialysis procedure. At this time, SPMDs are commer-
cially available from EST or Exposmeter AB, Trehörningen 34, 922 66 Taveljö,
87
88 Chapter 4
Sweden. However, it is important to examine key aspects of the SPMD preparation

process and associated QC issues.
4.3. EVALUATION OF COMPONENTS AND PREPARATION

OF SPMDs
Careful attention to pre-cleaning all SPMD components is a critical part of

ensuring that potential interferences are at acceptably low levels. To that end,
the layflat, low-density polyethylene (LDPE) tubing used for SPMDs contains
no antioxidants, blockers, slip additives, plasticizers, etc. Prior to use, it is batch
extracted with high purity hexane (see Chapter 5). Typical treatment consists of
three, 24 hour extractions at about 26 ◦ C with a minimum of 2.2 mL of hexane per
cm of tubing (2.5 cm wide). This step removes most of the lower-molecular-weight
polyethylene oligomers (the so-called polyethylene waxes which are present in
all LDPE) and other potentially interfering (analytically) compounds present in or
sorbed by the LDPE. Following this cleanup or preparation process, representative
samples of the LDPE membrane are further extracted, as described for SPMDs (see
Chapter 5). These extracts are purified using size exclusion chromatography (SEC)
to remove residual LDPE oligomers, then evaluated by gas chromatography (GC).
Detection of potential interferences is by means of an electron capture detector
(ECD), a flame ionization detector (FID) or a mass spectrometer (MS). Examples of
these analytical techniques are presented in more detail in Chapter 5. If the analyses
of concentrated LDPE extract (≤1 mL) show only negligible levels of analytical
interferences, the cleaned tubing is stored at room temperature (refrigerated if
long term storage/preservation is planned) in sealed metal cans under an inert
atmosphere (usually argon) until use.
Triolein is also examined for potential analytical interferences and contami-
nant residues before use in SPMDs. Generally, the triolein used for fabrication of
SPMDs is purified and its purity subsequently verified by GC or GC-MS methods.
This purification process has been described in detail previously (Lebo et al., 2004)
and is summarized in Section 5.4.
Because SPMDs have high sampling rates (Rs s) for vapor phase contaminants,
all SPMDs are assembled in an environmentally controlled room equipped with
an activated carbon air filtration system for the removal of airborne contaminants.
SPMDs of almost any length can be prepared after allowance of space for the
molecular welds or heat seals (i.e., ≈ 2.5 cm for each end). However, different
lengths of SPMDs should maintain the standard surface-area-to-volume (A V −1 )
ratio, which is given below. At room temperature, triolein is viscous and can cling
to surfaces. This characteristic can cause errors in the volume of triolein delivered
by common pipettors. To ensure accurate volumetric delivery, a pipettor equipped
with a total displacement plunger is employed to deliver triolein into the LDPE
tubing. Afterwards, the triolein is formed into a thin film throughout the length
Study Considerations 89
of the tubing, using a gloved hand (caution: gloves must be powder-free and pre-
rinsed to remove surficial contaminants). Care is used to force any air out when
forming the triolein film. Then, the open end of the tubing is heat-sealed. To ensure
durability of the closure, three to four seals are normally used at each end. Finally,
tether loops of LDPE tubing (no triolein) can be welded to both ends of the SPMD
membrane to facilitate deployment. The desirability of tether loops is dependent
on the method of deployment.
Standardization of SPMDs is essential to permit a more universal compa-
rability of passive sampler data. Toward this goal, a standard SPMD design has
been defined (Huckins et al., 2002). Because commercially available SPMDs are
modeled after the original USGS design, are of uniform construction, are used
globally, and represent the configuration used in most calibration studies, the spec-
ifications of these devices are operationally defined as “standard”. Depending on
study objectives, the use of non-standard SPMDs or other passive samplers may be
advantageous. However, we recommend the inclusion of a few standard SPMDs
in these studies as well to aid in data comparability.
4.4. SPECIFICATIONS OF THE STANDARD SPMD
The design of commercially available SPMDs consists of a specified length

(e.g., 91.4 cm between the inner-LDPE welds for 1 mL of triolein) of additive
free, 2.5 cm wide layflat LDPE tubing. The LDPE wall thickness ranges between
70–95 µm and the triolein used is ≥95% purity. Note that in the USA, all commer-
cially available SPMDs are fabricated with ≥99% purity triolein. The A V −1 ratio
is about 90 cm2 cm−3 (lipid plus membrane), or about 460 cm2 mL−1 of triolein.
The standard SPMD thereby consists of approximately 20% triolein. For the 1 mL
triolein configuration, the whole device typically weighs about 4.4 to 4.6 g.
Any length of SPMD with an A V −1 ratio of about 460 cm2 mL−1 of ≥95%
triolein, having an approximate 0.25 lipid-to-membrane-mass ratio (i.e., 20% lipid)
and a 70–95 µm wall thickness is considered a standard SPMD. However, the
aforementioned QC certification requirements for the membrane and triolein must
be met. Because most SPMD calibration data and field data are based on the
standard configuration, the use of non-standard designs for monitoring must be
weighed against the lack of data comparability.
4.5. PRE-EXPOSURE CONSIDERATIONS
Although the goals or endpoints of SPMD studies vary widely, a number

of questions should at least be considered prior to initiation of exposures. These
questions include the following: 1) are there threshold limit values (air) or water
quality criteria for chemicals of concern, and if so, has the lowest environmental
concentration of concern (Cc ) been established for target compounds; 2) will study
90 Chapter 4
sampling protocols provide adequate masses of target compounds for analysis or

toxicity assessment at levels of concern; 3) is there any information available
on turbulence and flow rates, temperature, biofouling potential, and turbidity at
exposure sites; 4) are SPMD calibration data available for target compounds;
5) do target compounds undergo photolysis (i.e., photodegradation), and if so,
will deployment devices and site conditions (e.g., light penetration of site water or
turbidity, natural shading and albedo or light reflectance of site surfaces) adequately
protect SPMD concentrates from photodegradation; 6) will SPMD deployment
devices be secure from vandalism or theft at study sites; and 7) will residues
accumulated in SPMDs represent linear (integrative), curvilinear or equilibrium
uptake kinetics?
Clearly, question 1 is not applicable when the project is a reconnaissance for
unknown toxics. In regard to question 2, investigators can sometimes (assuming
key parameters such as Rs values are known or can be approximated and that uptake
of the chemicals of interest is linear throughout the exposure) use the following
simple relationship for guidance.
Rs tn Cc Pr E t > MQLVi (4.1)
Where Rs is the sampling rate given as the volume of sampled matrix extracted
of analyte per day, t is days of exposure, n is the number of SPMDs per sample,
Cc is defined above, Pr is the overall procedural or method recovery (given as
a fraction of one; thus 1 is taken as 100% recovery) for the analyte, E t is the
fraction of the total of sample extract injected into the instrument used for analyte
quantitation, MQL is the method quantitation limit and Vi is the volume of the
standard injected. For example, if 2, 2 , 5, 5 -tetrachlorobiphenyl’s aqueous Rs is
6.4 L d−1 at 12 ◦ C and <1 cm s−1 flow (Meadows et al., 1998) and the Cc is
0.2 ng L−1 , then a single 1 mL-triolein SPMD (standard configuration as defined
in Section 4.4) exposed to 0.2 ng L−1 of aqueous 2, 2 , 5, 5 -tetrachlorobiphenyl
would sample about 38 ng over the course of 30 days. Assuming an instrumental
MQL of 1 pg µL−1 , a Vi of 1 µL, a Pr of 0.8 and an E t of 0.001, then 31 pg is
>1pg and the relationship given above (Eq. 4.1) holds, assuming uptake remained
linear. When the Rs of a compound of concern is not available, for this exercise it
is acceptable to use the Rs of another chemical with a similar K ow . Although this
approach of forecasting sampling and analytical outcomes is useful for planning,
it is clearly only a rough estimation because of the many variables involved. In
particular, commonly available laboratory-derived Rs values may not reflect actual
in situ Rs values, which can only be accurately derived when using performance
reference compounds (PRCs; see Chapter 3 for details).
Generally, there is little data available on site conditions with the possible
exceptions of seasonal average temperatures and river flow rates at USGS gaging
stations (question 3). Thus, reconnaissance of study sites is often needed to assess
exposure conditions and potential problems. A reconnaissance of study sites is also
an important part of assessing the potential for vandalism and theft (question 6)
and the precautions needed. The available sampling rate data related to question 4
can be found in Appendix A.
Unfortunately, question 5 has received little attention in the passive sampling
literature, even though the potential for photolysis of certain compounds accumu-
lated in SPMDs and other passive samplers may be high. For example, Kochany
and Maguire (1994) have reviewed the literature on the photolysis of polycyclic
aromatic hydrocarbons (PAHs) in water. They found that photolysis of most PAHs
occur within the 300–400 nm wavelength range (i.e., both UV-A and UV-B wave-
lengths) and that photoreaction rates are rapid in clear river water and in clear areas
of the upper layer of the ocean (i.e., top 35 m). In clear laboratory water, photolysis
half-lives of 11 PAHs ranged from 0.13 hour for 9-methylanthracene to 71 hours
for naphthalene. In general, it appears that the alkyl-substituted PAHs are more
photolabile than unsubstituted PAHs and that smaller PAHs photolyze less rapidly
than larger PAHs (Kochany and Maguire, 1994). However, photolysis is markedly
attenuated in turbid water, and Lee et al. (1978) found that in outdoor ecosystem
enclosures, only the higher molecular weight PAHs were removed primarily by
photolysis. Because LDPE does little to quench UV-A and UV-B wavelengths of
solar radiation, photolysis of residues in SPMDs has been used as an activation
method for the Mutatox assay (Johnson, 2001).
In view of the potential for photolysis of chemicals in clear water and air,
Orazio et al. (2002) exposed SPMDs spiked with the 16 priority pollutant PAHs
to sunlight in a 1 m deep pond (relatively clear water) and outdoor air. In both
cases, SPMDs were deployed in stainless steel canisters (Figure 4.1), designed
by Harry Prest (Long Marine Laboratory, Santa Cruz, CA, USA), and naked, i.e.,
tethered without any protective covering. The canisters were placed so that the
longitudinal axis was perpendicular to the sun at noon (i.e., they were mounted
horizontally) and in some cases aluminum foil was used to further shield the
canisters or SPMDs inside canisters. Exposure periods were as long as one week for
these experiments. During these exposures, significant losses (due to volatilization)
of naphthalene, acenaphthylene and acenaphthene from SPMDs were expected.
Also, Orazio et al. (2002) investigated the effects of brief (≤2 h), direct sunlight
exposure on polybrominated diphenyl ethers (PBDEs) in SPMDs.
The major finding of the Orazio et al. (2002) study included the following.
PAHs in SPMDs deployed in Prest-type canisters (both with and without addi-
tional aluminum foil shielding) for one week in pond water (1 m deep), showed
no photolysis. The PAHs in contiguously deployed naked SPMDs were found to
have suffered extensive photolysis. Atmospheric exposures appear to be of even
greater concern. For example, certain PAHs in SPMDs exposed to direct sunlight
were photolyzed after only two minutes. After one week exposure (air) to direct
sunlight, PAHs in SPMDs underwent photolysis even with additional foil shielding
over the canisters. In a similar manner, a two minute exposure to direct sunlight
(air) resulted in photolysis of a number of higher molecular weight PBDEs. These
data suggest that the likelihood for photolysis of compounds in SPMDs is much
greater in atmospheric exposures than in aqueous exposures. This suggestion is
92 Chapter 4
FIGURE 4.1 A stainless steel deployment device designed by Harrry Prest, Santa Cruz, CA, which
has the capacity for four 1 mL triolein SPMDs. The whole apparatus, loaded with SPMDs, fits
in a 3.85 L gas tight steel can (Figure 4.2) for transport to and from the field.
in line with the proposed mechanism of PAH photolysis (Kochany and Maguire,
1994), where oxygen may be rate limiting. Oxygen levels are much lower in wa-
ter than in air (also see discussion on oxygen permeability of LDPE in Section
3.9.2.). Recently, Bartkow et al. (2004) found that photolysis of deuterated-PAHs
(i.e., PRCs) occurred after a 32 day exposure of SPMDs to air inside galvanized
iron chambers with louvers on all sides and open bottoms. In view of these find-
ings, we recommend the development of atmospheric deployment devices that
exclude all light, yet allow for air exchange. Perhaps, the mounting of Prest-type
canisters (see Section 4.7. for description) inside chambers similar to the Bartkow
et al. (2004) design but with louvered metal bottoms, would result in deployment
devices protective of photosensitive compounds. For exposures in clear, fast water,
canisters with double walls and offset holes may be required, whereas canisters as
described in Section 4.7. can be used in more quiescent-shallow waters if effec-
tive shading structures or chambers can be mounted over the deployment devices.
Finally, the use of a photosensitive, high K ow deuterated-PAH spiked into SPMDs
will serve as an indicator that photolysis of some classes of target compounds has
occurred.
The following equations can be used for predicting times that SPMD sampling
will represent linear, curvilinear and equilibrium kinetics (question 7), assuming
key parameters such as Rs values are known or can be approximated.
t1/2 = − ln 0.5K sa/w Vs /Rs ≈ 0.693K oa/w Vs /Rs (4.2)

t95 = − ln 0.05K sa/w Vs /Rs ≈ 2.99K oa/w Vs /Rs (4.3)
Where t1/2 and t95 are the times to reach 50% and 95% of the equilibrium concen-
trations, respectively, K sa/w is the equilibrium SPMD-air (a) or SPMD-water (w)
partition coefficient, Vs is the volume of the SPMD, and K oa/w and Rs were de-
fined in Chapter 3. Note that Eq. 4.3 is essentially Eq. 1.2. Sampling can be
considered as integrative during the linear uptake phase or about one t1/2 , and
exposure concentrations derived from SPMD levels (see Chapter 3) represent time
weighted average concentrations. To confirm the assumption of linear uptake nec-
essary for the use of Eq. 4.1 presented earlier, we use Eq. 4.2 and the example of
2, 2 , 5, 5 -tetrachlorobiphenyl. Again, we assume an Rs of 6.4 L d−1 for a 1 mL
triolein SPMD (Vs = 5 cm3 ), and from Huckins et al. (1993), we derived a K sw of
1.19 × 105 . Using these data, the computed t1/2 for 2,2 ,5,5 -tetrachlorobiphenyl is
64 d. Thus, the linear uptake assumption for Eq. 4.1 is likely valid as we assumed
only a 30 day exposure. This exercise is unnecessary when using the exponential
Eq. 3.21.
When exposure time exceeds slightly more than four-t1/2 s or ≈t95 for a
chemical, the SPMD is essentially at equilibrium with the ambient environment.
Therefore, the time period in between the t1/2 and t95 values represents the curvi-
linear region of uptake. Because it is not possible to accurately predict in situ
SPMD exchange rates (e.g., Rs ) a priori, this exercise is for planning only and is
no substitute for the use of PRCs as discussed in Chapter 3.
4.6. STORAGE, TRANSPORT AND RETRIEVAL
Sample preservation during storage, transport and retrieval must be consid-

ered to maintain QC. Prior to field deployment, SPMDs can be stored under argon
at ≤−15 ◦ C, in solvent rinsed, gas-tight sealed metal cans (Figure 4.2.). The cans
can be purchased or are provided by the commercial supplier of SPMDs. If PRCs
(see Chapters 1 and 3) are used in any of the SPMDs, the PRC-containing SPMDs
must be kept separate from the others. The canned samplers should be shipped to
the field on ice in efficient coolers that are designated only for SPMD transport.
Although it may not be completely essential to transport SPMDs without PRCs to
the field at low temperatures (the SPMDs are in an inert atmosphere until the seal
on the can is broken), it is always preferable to maintain the samplers frozen or at
near-freezing temperatures. In particular, SPMDs with PRCs should be maintained
at freezing or near-freezing conditions during transport to and from sampling sites
to minimize losses of these QC compounds. A variety of coolants can be used for
94 Chapter 4
FIGURE 4.2 For storage and shipping, SPMDs (shown on a deployment apparatus rack) are placed
in a clean metal can, flushed with argon, and sealed with a gas tight lid. This figure reproduced
courtesy of the American Petroleum Institute (Huckins et al., 2002).
shipping SPMDs, including ice, blue ice, and dry ice. After SPMD deployment,
the lids are resealed on the shipping cans and the cans are stored refrigerated until
retrieval of the SPMDs.
Following retrieval from the exposure medium, SPMDs are immediately
sealed inside the same labeled metal cans and transported (frozen or near frozen)
back to the analytical laboratory in a cooler. If it is necessary to delay the shipping
of exposed SPMDs more than a few hours, then they must be stored frozen at
≤−15 ◦ C in the sealed metal cans. (Caution: failure to maintain exposed SPMDs
under freezing conditions can result in significant losses of analytes with rela-
tively high fugacities [e.g., naphthalene]). However, no measurable losses of 2,4,5-
trichlorophenol (high fugacity from SPMDs at room temperature) were observed
from SPMDs stored at −15 ◦ C for 6 months in sealed cans (Huckins, 1995b).
4.7. DEPLOYMENT DEVICES
SPMDs have been successfully deployed in a variety of deployment devices

(Ellis et al., 1995; Lebo et al., 1995; Petty et al., 1995). The commercially available
FIGURE 4.3 A commercially available stainless steel deployment canister (see Section 4.7), which
has a capacity of five 1-mL triolein SPMDs. Each SPMD is placed on a separate deployment
rack and the five deployment racks are held in place by a threaded center pin.
deployment canister (Environmental Sampling Technologies [EST], St Joseph,

MO) shown in Figure 4.3, is the most widely used system. This stainless steel
canister holds a maximum of five standard SPMDs mounted on individual racks,
but the whole apparatus is too large to fit into a 3.85 L gas-tight steel can (Figure 4.2
shows a 3.85 L can with an SPMD mounted on a rack). Although the Prest stainless
steel canister (Figure 4.1) only holds four standard SPMDs, the SPMD-loaded
canister easily fits into the 3.85 L cans for transport to the field. Thus, using the
Prest system has an advantage of minimizing handling and time required to deploy
and recover SPMDs at sampling sites.
Regardless of the choice of SPMD deployment structure, certain generic
guidelines should be used in its design and construction. These include the follow-
ing: 1) metal containment structures must be free of cutting oils or other potential
interferences, i.e., they must be decontaminated before each use; 2) use of most
plastic components should be minimized (Teflon and perhaps some grades of PVC
are exceptions), due to the possible presence of leachable organic residues, and
in some cases, competitive sorption of analytes by the plastic; 3) the design of
96 Chapter 4
the structure should minimize abrasion of the LDPE membrane, even in turbulent
environments, should reduce site-to-site differences in the effective thickness of
the SPMD aqueous or air boundary layers (note that container designs that baffle
flow can be used to accomplish these goals), and should maintain adequate ex-
change rates at the membrane surfaces, i.e., sampling should not cause significant
or differential depletion of chemical concentrations at the SPMD-exposure media
interface; 4) once the SPMDs are mounted in the deployment device, the lipid-
containing portion of the layflat tubes should not make contact with container walls,
and any SPMD tether loops should not self-adhere thereby reducing the effective
surface area (generally, some tension on SPMD tether loops and/or the use of a
Möbius configuration [Lebo et al., 1992] prevents this problem); 5) as discussed
in Section 4.5., additional shading structures beyond the deployment canisters or
double wall canisters may be required to protect photosensitive target compounds
(e.g., PAHs and brominated diphenyl ethers) from sunlight; 6) the structure should
be adequately tethered to prevent loss during flood events; 7) because vandalism is
always a potential problem in the field, the structure should be amenable to hiding;
and 8) designs that minimize “silting in” should be used if deployments are at the
sediment-water interface.
As suggested earlier, deployment systems that baffle flow-turbulence and
protect SPMDs from sunlight are needed for many exposure scenarios. However,
any reduction in flow-turbulence at the membrane-exposure medium interface will
result in a reduction in SPMD sampling rates for aqueous solutes with log K ow s >
4.5 or for vapors with log K oa s > 8.5, as postulated in Chapter 3. Louch et al. (2003)
measured river flow velocity just outside an EST SPMD deployment apparatus
and inside the apparatus loaded with SPMDs and found that inside flow was about
a 50% less than outside flow. Figure 3.6, suggests that measurements of bulk-
media flow are poor predictors of SPMD sampling rates. Thus, we cannot assume
that sampling rates are halved but we can assume that they are reduced. Potential
reductions in SPMD sampling rates due to container baffling effects and biofouling
should be considered when estimating the number of SPMDs required per sample
to achieve satisfactory quantitation limits for a project. Studies have shown that
C.V.s of hydrophobic contaminant concentrations in replicate SPMDs deployed
inside the canisters shown in Figures 4.1 and 4.3 are generally <25% (Huckins,
1998; Louch et al., 2003). Consequently, it seems unlikely that some SPMDs in a
canister would experience reduced uptake relative to others.
4.7.1. Precautions and Procedures During Deployments

Because SPMDs sequester a wide variety of organic solutes or vapors of
hydrophobic chemicals, care must be used to prevent inadvertent contamination of
the devices. Of particular concern for SPMDs destined to be used for environmental
sampling is the fact that SPMDs clear large volumes of vapor phase chemicals from
air. For example, under low flow conditions (<5 cm s−1 ) at about 22 ◦ C the Rs
or SPMD sampling rate (1 mL standard device) of vapor phase phenanthrene is

4.3 m3 d−1 , which is about the same rate as that observed for dissolved phase
phenanthrene (i.e., 3.6 L d−1 at <1 cm s−1 flow and 18 ◦ C), after the differences
in the density of air and water are taken into account (note that we are not inferring
that density adjustments alone can account for volumetric differences in Rs values
for air and water ). An atmospheric sampling rate of 4.3 m3 d−1 is equivalent
to about 3 L min−1 (many other vapors are sampled at an even higher rates) of
air cleared of vapor. Clearly, SPMD air exposure must be minimized to prevent
sample contamination.
Although SPMDs do require special considerations, proper handling of
SPMDs generally consists of logical precautions, which can be learned without
special training and are related to good laboratory practices. SPMDs are used in a
wide variety of environmental systems, ranging from wetlands and lakes to more
energetic systems such as rivers, estuaries and ocean environments, as well as
a wide range of atmospheric exposure conditions. However, the following prac-
tices and considerations apply to all deployments. 1) Before deployment and prior
to retrieval, inspect study sites for nearby sources of vapor-phase contaminants,
including fumes from engines, oils, tars, gasoline, diesel fuel, paints, solvents,
cigarette smoke (i.e., no smoking during deployments and retrievals), asphalt pave-
ment, etc. Record any findings of potential sources of contamination for each site.
2) Atmospheric exposure time during aqueous deployment and retrieval should
be minimized, because when the SPMD is exposed to site air, sampling begins
immediately. Atmospheric levels of some target compounds may be higher than
their concentrations in aquatic systems to be sampled. Also remember that during
SPMD recovery, photolysis of some compounds may be rapid. 3) Ensure that cans
with SPMD field blanks (the field blank SPMDs provide a record of any chemical
accumulated in SPMDs during transport, deployment and retrieval; see Chapter 5
for detailed definitions of QC samples and their suggested use) are open to the air
while sample SPMDs are being deployed and retrieved. 4) If waterborne chemi-
cals are visible as surface layers of oils, tars, gasoline, etc., or a biofilm is visible
on the surface of the water, where target compounds are potentially elevated (i.e.,
relative to the water column), precautions may be needed to reduce contamination
during placement of the deployment devices into the aquatic system. 5) Hand lo-
tions, perfumes, colognes, powdered gloves (use powder-free gloves), etc, should
not be used when handling samplers or deployment devices as they may contain
chemicals accumulated by the SPMDs (Petty et al., 2000). 6) The procedure for
retrieval of the SPMDs is essentially the reverse of the deployment sequence and
the same precautions apply. 7) Following retrieval, immediately place the SPMDs
back into the original metal cans, as provided by the supplier, and seal the lids on
the cans (Caution: if the lid is not completely sealed on the can, contamination of
SPMDs from airborne chemicals during transport back to the analytical laboratory
is highly probable). 8) Finally, place the cans containing the SPMDs into a cooler
and maintain frozen pending and during shipment to the processing laboratory.
98 Chapter 4
The procedures given above are designed to prevent any contamination of

SPMDs, to minimize losses of PRCs and accumulated analytes, and to reduce
the possibility of analyte photolysis. These precautionary steps are particularly
important when target compounds are at ultra-trace levels, as it may be difficult
to delineate handling, storage and shipping-related contamination from analytes
concentrated during the exposure.
Because environmental variables affect the uptake of target analytes, regard-
less of analyte types or sampling approach (i.e., kinetic or equilibrium), a record
should be kept on site conditions during exposures. Relevant data include temper-
ature (a minimum of the beginning and end of the deployment), the visual extent
of fouling (i.e., light, medium, heavy, none), and an estimation of turbulence-
flow rates (i.e., cm s−1 ) for water deployments. This type of data is helpful even
when PRCs are used, because it provides supporting information on the possible
causes of significant differences between calibration Rs s and in situ Rs s and the
differences of in situ Rs s among sites. As with any research project, notes should
be taken describing the site location and characteristics not discussed above and
events occurring during deployment and retrieval that may affect data quality.
This information may prove invaluable if QC-related questions arise during sam-
ple processing and analytical procedures, and will often be helpful in conducting
site assessments.
Successful applications of the SPMD technology under a wide variety of field
conditions have been demonstrated by a number of researchers (see Appendix B or
CERC, 2004). The common threads among successful applications of SPMDs are
a basic understanding of potential sources of sample contamination and losses, the
functional aspects of the SPMD technique, and the adherence to sound sampling
approaches and good laboratory practices.
4.8. REFERENCES
38: 5983–5900.
CERC 2004, SPMD Home Page. USGS Columbia Environmental Research Center: Columbia, MO;
http://wwwaux.cerc.cr. usgs.gov/SPMD/index.htm.; (accessed December, 2004).
of lipid-containing semipermeable membrane devices (SPMDs) for monitoring organochlo-
rine contaminants in the Upper Mississippi River. Environ. Toxicol. Chem. 14: 1875–
1884.
Huckins, J.N.; Lebo, J.A.; Tubergen, M.W.; Manuweera, G.K.; Gibson, V.L.; Petty, J.D. 1992, Binary
Concentration and Recovery Process. U. S. Patent 5,098,573, March 24, 1992.
27: 2489–2496.
Huckins, J.N.; Petty, J.D.; Zajicek, J.L.; Gibson, V.L. 1995a, Device for the Removal and Concentration
of Organic Compounds from the Atmosphere. U. S. Patent 5,395,426, March 7, 1995.
Huckins, J.N. 1995b, Stability and residence times of chemicals in frozen SPMDs. USGS Columbia
Environmental Research Center: Columbia, MO; Unpublished work.
Huckins, J.N.; Lebo, J.A.; Tubergen, M.W.; Manuweera, G.K.; Gibson, V.L.; Petty, J.D. 1996, Binary
Concentration and Recovery Process. Canadian Patent 2,037,320, December 17, 1996.
Huckins, J.N. 1998, Effect of deployment devices on performance reference compounds (PRCs) loss.
USGS Columbia Environmental Research Center: Columbia, MO; Unpublished work.
Johnson, B.T. 2001, USGS Columbia Environmental Research Center: Columbia, MO. Personal
communication.
Kochany, J.; Maguire, R.J. 1994, Abiotic transformations of polynuclear aromatic hydrocarbons and
polynuclear aromatic nitrogen hetrocycles in aquatic environments. Sci. Total Environ. 144: 17–31.
Lebo, J.A.; Zajicek, J.L.; Huckins, J.N.; Petty, J.D.; Peterman, P.H. 1992, Use of semipermeable mem-
brane devices for in situ monitoring of polycyclic aromatic hydrocarbons in aquatic environments.
Lebo, J.A.; Gale, R.W.; Petty, J.D.; Tillitt, D.E.; Huckins, J.N.; Meadows, J.C.; Orazio, C.E.; Echols,
K.R.; Schroeder, D.J.; Inmon, L.E. 1995, Use of the semipermeable membrane device (SPMD) as
an in situ sampler of waterborne bioavailable PCDD and PCDF residues at sub-part-per-quadrillion
concentrations. Environ. Sci. Technol. 29: 2886–2892.
Lebo, J.A.; Almeida, F.V.; Cranor, W.L.; Petty, J.D.; Huckins, J.N.; Rastall, A.; Alvarez, D.A.;
Mogensen, B.B.; Johnson, B.T. 2004, Purification of triolein for use in semipermeable mem-
brane devices (SPMDs). Chemosphere 54: 1217–1224.
Lee, R.F.; Gardner, W.D.; Anderson, J.W.; Blaylock, J.W.; Bardwell-Clarke, J. 1978, Fate of polycyclic
aromatic hydrocarbons in controlled ecosystem enclosures. Environ. Sci. Technol. 12: 832–838.
Louch, J.; Allen, G.; Erickson, C.; Wilson, G.; Schmedding, D. 2003, Interpreting results from field
deployments of semipermeable membrane devices. Environ. Sci. Technol. 37: 1202–1207.
Orazio, C.E.; Haynes, S.A.; Lebo, J.A.; Meadows, J.C.; Huckins, J.N.; Petty, J.D. 2002, Potential
for Photodegradation of Contaminants During SPMD Sampling. Presented at the 23rd Annual
National Meeting of the Society of Environmental Toxicology and Chemistry; November 16–20,
2002; Salt Lake City; UT; Abstract P192.
Petty, J.D.; Huckins, J.N.; Orazio, C.E, Lebo, J.A.; Poulton, B.C.; Gale, R.W.; Charbonneau, C.S.;
Kaiser, E.M. 1995, Determination of bioavailable organochlorine pesticide residues in the Lower
Missouri River. Environ. Sci. Technol. 29: 2561–2566.
Petty, J.D.; Gale, R.W.; Huckins, J.N.; Cranor, W.L.; Alvarez, D.A.; Clark, R.C. 2000, Development and
Application of Techniques for Sampling Bioavailable Airborne Contaminants-Tentatively Iden-
tified Compounds by Gas Chromatography/Mass Spectrometry. USGS Columbia Environmental
Research Center: Columbia, MO; Unpublished report to U.S. EPA National Exposure Assessment
Laboratory: Las Vegas, NV.
Chapter 5
Analytical Chemistry Related

to SPMDs
5.1. ANALYTICAL SPEED, SELECTIVITY

AND QUANTITATION LIMITS
Based on the results from more than 50 interlaboratory comparisons, Horwitz

et al. (1980) showed that, as target compound concentrations in complex matrices
decrease, the relative standard deviation or coefficient of variation (C.V.) of analyt-
ical results increases exponentially. There are a number of potential causes for this
problem, but at very low environmental concentrations, method selectivity (i.e., the
ability to distinguish the analyte from interferences) is a major factor. Because pas-
sive samplers such as the SPMD accumulate a broad range of nonpolar chemicals
(including complex mixtures of anthropogenic and biogenic compounds charac-
teristic of the site and sample matrix), sampling must be considered non-selective.
Fortunately, passive samplers with nonpolar sequestration phases concentrate po-
tential interferences with low octanol-water partition coefficients (K ow s) less than
the nonpolar target compounds. However, this enrichment effect is offset when
trace (<1 µg L−1 ) or ultra trace (<1 ng L−1 ) detection or quantitation limits are
required due to the much higher concentrations of potential interferences (relative
to analyte concentrations).
Two approaches are generally used to develop methods with lower detection
and quantitation limits for target compounds. One approach involves the use of
sample cleanup methods such as size exclusion chromatography (SEC) for the
101
102 Chapter 5
selective removal of interferences and various fractionation techniques to obtain

class separations prior to analysis. The other involves the use of instrumentation
capable of high analyte selectivity such as a gas chromatograph (GC) or a liquid
chromatograph (LC) interfaced with a mass spectrometer (MS; especially high
resolution MS and MS-MS systems). In the first case, the use of multiple cleanup
steps must be balanced by the potential for analyte losses during each step and
the associated propagation of errors or measurement uncertainty. However, use of
good laboratory practices and performance-based methods for multi-step cleanup
procedures have increased analyte recoveries and analytical precision, and reduced
detection and quantitation limits.
Advancements have also been made in the use of analytical instrumenta-
tion for both sample cleanup and quantitation at trace levels. However, a number
of problems still remain. For example, matrix effects or interfering compounds
can cause ionization-suppression or ionization enhancement (Jones-Lepp, 2004;
Meyer, 2004) using MS and poor chromatographic resolution with co-eluting com-
pounds and shifting retention times. These factors combine to reduce the sensitivity
of the analysis and the certainty of compound identification. Some progress is be-
ing made with innovations such as the ChromatoProbe device for dirty sample
injection (Maštovská and Lehotay, 2003), but for extracts with complex mix-
tures of compounds, it is doubtful that the need for class fractionation can be
avoided.
To reduce costs of sample analyses, efforts are often directed toward increas-
ing analytical speed (e.g., Maštovská and Lehotay, 2003). Generally, several trade-
offs result from increased speed of analysis, which can be illustrated as follows:
SPEED
SELECTIVITY
QUANTITATION LIMITS
Solid-phase microextraction (SPME) fibers provide a good example of a passive

method that can be used for rapid analysis of water samples. After turbulent sample
extraction (<1 hr), the fibers can be directly introduced (i.e., the entire sample is
analyzed) into a fast GC system (Gorecki and Pawliszyn, 1997). Depending on
the analytical instrument used, SPME method quantitation limits (MQL) may
approach pg levels. For a particular compound and instrument, this limitation
stems from the sorbent-phase volume (<1 µL for most SPMEs), the partition
coefficient of the analyte, and the relative concentration and partition coefficient of
interferences present in a sample. To obtain lower quantitation limits with passive
samplers, the sorbent phase volume must be increased and some sample cleanup is
required. From Table 1.1, a 1 mL triolein SPMD has about 5 × 103 times greater
sorbent phase volume than an SPME, and during the same time interval, the volume
of water or air extracted by the SPMD is hundreds of times greater than the
SPME. The large volume of sample matrix extracted allows pg L−1 detection
and quantitation limits when coupled with selective cleanup and fractionation
Analytical Chemistry Related to SPMDs 103
procedures. Using composite samples of SPMDs (n = 9 SPMDs per sample) and

rigorous cleanup and fractionation, McCarthy and Gale (2001) have quantified the
homologous- tetrachlorodibenzo- p-dioxins and tetrachlorodibenzofurans in river
water at <pg L−1 .
Procedures used for the analysis of SPMD samples are similar to those typ-
ically utilized for the determination of organic contaminant concentrations in en-
vironmental matrices such as aquatic organisms. However, extracts from aquatic
organisms can vary widely in the quantities and compositions of lipids and other
co-extracted biogenic materials, while SPMD extracts (i.e., dialysates) are more
uniform, well characterized, and generally contain significantly less lipid. Thus,
sample processing procedures and methods for the analysis of SPMD samples are
generally more amenable to standardization than those used for aquatic organism
tissues, and in some cases, require fewer labor-intensive steps.
Contaminants accumulated by SPMDs include a broad array of waterborne
and airborne organic chemicals and comprise most of the bioconcentratable or-
ganic chemicals of concern to environmental scientists. In fact, nearly any non-
polar organic chemical with a molecular cross sectional diameter no greater than
about 10 Å will be accumulated by the SPMD, given appropriate exposure con-
ditions. Consequently, the SPMD approach provides a convenient means for as-
sessing complex mixtures of organic chemicals present in air, water, sediments
and soil. Because of the non-specificity of nonpolar contaminants accumulated
by SPMDs, the physicochemical properties of individual compounds (i.e., those
relevant to chromatographic separations and instrumental analysis) in extracts
containing complex mixtures may overlap. Furthermore, analyte levels for some
chemicals may be close to quantitation limits, even though a single 1 mL triolein
SPMD may extract >100 L of water or 100 m3 of air during a 30 d exposure. Sim-
ply put, these sample characteristics necessitate the use of analytical methods that
are performance-based and specific for the analytes of interest. Sample analyses
based on dilute-and-shoot or extraction and instrumental desorption generally do
not provide satisfactory results.
The use of sophisticated instrumental systems such as high-resolution GC-MS
does not guarantee satisfactory quantitation of the hundreds of chemicals some-
times present in SPMDs without some fractionation of sample residues. Thus, the
complexity of target residues, as well as interferences from the matrix sampled
can be determinants in the cleanup and separation procedures needed for satisfac-
tory analyses. The following discussion presents the salient features of the typical
processing and analytical procedures applied to SPMD samples.
5.2. QUALITY CONTROL
The application of appropriate Quality Control (QC) procedures or criteria

is a mandatory consideration in the deployment and analysis of SPMDs (e.g.,
Petty et al., 2000a). Similar to any performance-based methodology or approach,
104 Chapter 5
QC samples must address specific issues of analyte recovery, background in SPMD

components, and any contamination incurred during transport, deployment, re-
trieval, storage, processing, enrichment, and fractionation operations. The exact
level of QC required should be determined during the development of a project’s
experimental design. In general, QC samples represent 20–50% of a “sample set”.
We operationally define a laboratory sample set as a group of samples (includes
both exposed SPMDs and QC samples from the same study) that are processed
and analyzed together. The number of samples in a set generally ranges from 6
to 30. The upper size limit of a sample set is often constrained by non-automated
analytical procedures (e.g., chromatographic cleanup and fractionation), where the
analyst must monitor the performance of one or more steps.
For projects needing stringent QC, control charts are recommended to monitor
analyte recoveries throughout an investigation (see Taylor, 1987, for a detailed
discussion). Briefly, during each quarter of a project, the last 20 observations of
recoveries from QC spikes are used to generate a control chart. Control limits are
established for the analytical process as described by Taylor (1987). When control
limits are exceeded, sample analyses are suspended until the problem step(s) can
be identified and corrected. These actions must follow an appropriate protocol for
corrective action. The results of this type of investigation or procedure modification
become a part of the permanent record of the sample set and the project.
Herein, we describe the basic QC samples and parameters related to the
performance of SPMD studies, and elucidate their role in conducting studies.
Also, a general overview of SPMD analytical procedures and data applicability
are given.
5.2.1. SPMD-Fabrication Blanks

This type of QC blank consists of a batch or subset of individual SPMDs of the
same size and material as those prepared for a specific project. After preparation,
SPMD-fabrication blanks are maintained frozen in vapor-tight metal cans under
argon at −10 to −20 ◦ C in the laboratory until the analysis of the project SPMDs.
Processing and analysis of these blanks is concurrent with and identical to that of
environmentally exposed SPMDs. The primary purpose of this type of QC sample
is to account for any background contribution due to interferences from SPMD
components, and for contamination incurred during laboratory storage, processing,
and analytical procedures.
5.2.2. SPMD-Process Blanks

This type of QC blank consists of a subset of SPMDs, made just prior to
initiation of the analysis of an SPMD sample set. Operationally, the only differ-
ence between SPMD-process blanks and SPMD-fabrication blanks is the time of
preparation and that the SPMD-process blanks are not subjected to storage, but
are immediately processed and analyzed along with the environmentally exposed
SPMDs. Use of this type of blank is generally limited to laboratories that assem-
ble SPMDs. If the numbers of SPMD-fabrication blanks are inadequate, SPMD-
process blanks can be used to determine analyte recovery and the precision of
the overall analytical method. Also, this type of QC sample can be used for other
purposes, such as determining potential effects of storage or changes in batches or
lots of SPMD materials.
5.2.3. Reagent Blanks

These blanks consist of portions of all solvents (volumes identical to those
used for SPMD samples) used during the processing, enrichment, and instrumental
analysis of an SPMD sample, that are carried along with SPMD samples through
the entire analytical procedure. This type of QC sample (at least one for each
sample set) provides information on background due to laboratory reagents and
procedures. The use of reagent blanks is strongly recommended, because they
greatly facilitate diagnosis of any interference problems encountered during SPMD
analysis.
5.2.4. Field/Trip-Blank SPMDs

These blanks are a subset of the SPMDs (at least one per sampling site)
prepared for a specific field study and are identical to those that will be deployed.
Field-blank SPMDs are used to account for contamination during transport (both
to and from study sites) and during deployment and retrieval of exposed SPMDs.
Field-blank SPMDs can also be spiked with performance reference compounds
(PRCs) to provide the “day 0” concentration of the PRC (see Chapter 3) used
in the assessment of the effects of environmental conditions on analyte sampling
rates (i.e., derivation of an exposure adjustment factor) which facilitates in situ
calibration (Booij et al., 1998). These blanks are treated the same as deployed
devices, with the exception that they are not exposed to the matrix of interest
at the study sites and are stored frozen during the exposure period. In the case
of air sampling, these QC samples are the trip blank SPMDs, i.e., these blanks
accompany the deployment SPMDs during transport to and from the deployment
site, but unlike aqueous exposures, the container is not opened during deployment
and retrieval.
As suggested earlier, field-blank SPMDs are taken to the field in sealed metal
cans and one or more cans are opened to the atmosphere at each site (note that
field-blank SPMDs are typically left inside the open cans) during both deployment
and retrieval of SPMDs from aquatic systems. The time periods that field-blank
SPMDs are exposed to site air should be the same as that required to deploy
and retrieve SPMDs. Afterwards, the cans with the field-blank SPMDs are re-
sealed and shipped to the analytical laboratory along with the deployed SPMDs.
Non-spiked (i.e., no PRCs) field-blank SPMDs and PRC-containing field-blank
SPMDs are processed and analyzed exactly as are deployed SPMDs. However,
106 Chapter 5
when perdeuterated polycyclic aromatic hydrocarbons (PAHs) are used as PRCs,

then GC-MS, or GC-FID must be used for separation and quantitation of native
PAHs and their perdeuterated counterparts.
5.2.5. PRC Samples

When environmental conditions at an exposure site differ from those used for
laboratory calibrations or when calibration data for an analyte are not available,
at least one SPMD per site is spiked with PRCs. The type of compounds used for
PRCs and their spiking levels were discussed earlier. PRC samples and standard
SPMD samples (i.e., field-deployed SPMDs) differ only by the presence of the
PRCs. Handling, processing and analysis are also identical. As implied above, the
purpose of the PRC sample is to provide data for estimation of in situ sampling
rates of target compounds.
5.2.6. SPMD Spikes

Spiked SPMDs (at least one per sample set) are used to determine the re-
coveries of target compounds and to establish “control limits” for the analytical
process. The C.V. for each analyte is used to set control limits for that compound.
SPMD-fabrication or -process blanks are used for this type of QC sample. Sam-
ple processing and analysis of SPMD spikes is exactly the same as for deployed
samples. The triolein of individual SPMD blanks is directly fortified with target
compound mixtures. The amounts of target compounds used for SPMD spikes
(note that the carrier solvent volume should not exceed 10% of the lipid volume)
vary but are based on achieving an instrumental response that is near the midpoint
of the appropriate calibration curves. For example, 2 µg of each priority pollutant
PAH is generally spiked into 1 mL of triolein in a standard SPMD. Assuming a
75% recovery, the concentration of each priority pollutant PAH will be 1.5 µg
mL−1 (final sample volume of 1 mL), and upon analysis with a GC-Mass Specific
Detector (MSD), the instrumental response will fall near the midpoint of the PAH
calibration curve. In the case of organochlorine pesticides (OCPs), 40 ng of each
compound is generally spiked, and based on the same assumptions given above,
the concentration of each OCP will be 30 ng mL−1 , at a final volume of 1 mL. As
in the example above, the GC-electron capture detector (ECD) response will fall
at about the midpoint of the OCP calibration curve.
5.2.7. Procedural Spikes

These type of spikes are used when a rapid and independent assessment of
individual steps in sample processing and enrichment is desired. The spikes used
are radiolabeled (14 C- or 3 H- labeled) compounds, which typically consist of a high
molecular weight PAH or a chlorinated compound. For assessment of the dialytic
step of overall analyte recovery from SPMDs a procedural blank is spiked directly
with the radiolabeled compound(s). For assessment of additional processing steps,

the radiolabeled compound is injected into an appropriate solvent which is treated
the same as sample extracts at the same point in the processing sequence. The
fortification levels used for these spikes are about the same as described earlier for
the SPMD spikes. These QC samples are used as a troubleshooting aid to rapidly
identify abnormalities in specific steps that might contribute to low recoveries of
analytes from SPMD spikes. Also, procedural spikes are used to develop control
charts and are an integral part of any corrective action assessment.
5.2.8. Expected QC Results

Based on the examination of analytical data from polychlorinated biphenyls
(PCBs), OCPs and PAHs spiked into SPMDs, which have subsequently been sub-
jected to the entire SPMD analytical procedure described herein, recoveries are
generally >75% with good precision (i.e., C.V.s ≤ ± 20%). Surprisingly, the C.V.s
for the analysis of contaminants present in replicate SPMDs deployed contigu-
ously at the same sites and treated identically during analysis are often equivalent
to C.V.s of SPMD spikes. This observation suggests that the variability of analyte
sampling rates of replicate SPMDs in the field is small and that the analytical
methods used for field-deployed SPMDs are robust.
DeVita and Crunkilton (1998) have examined QC associated with the use
of SPMDs. The results of their study demonstrated that quality control measures
applied to SPMD analysis met or surpassed conventional guidelines (EPA Method
610 for PAHs in water was used for this comparison) for precision and accuracy.
This elevated level of data quality was achieved even though measurements of both
overall precision and accuracy of SPMD data encompassed more steps (each with
the potential for variability) than the conventional method. In summary, DeVita and
Crunkilton (1998) found that QC measures could be used to validate data from the
analysis of SPMDs used in the field. In view of the state of SPMD QC, it appears
that the SPMD approach for monitoring hydrophobic organic contaminants is
equivalent to some EPA-approved methods.
5.3. SAMPLE PREPARATION
The processing, enrichment, and fractionation of SPMDs for instrumental

analysis of residues have been described in a number of publications (Ellis et al.,
1995; Lebo et al., 1995; Petty et al., 1995, 1998a, 1998b, 2000a; Huckins et al.,
1996; Bergqvist et al., 1998a). Preparation of SPMD samples for residue analysis
generally involves the following steps: 1) removal of exterior surficial periphyton
and debris; 2) organic solvent dialysis (OSD); 3) size exclusion chromatography;
and 4) class-specific fractionation using chromatographic techniques. All solvents
used in these procedures are of high purity (e.g., chromatography grade, spec-
troscopy grade, etc.). Although Figure 5.1 illustrates an approach commonly used
108 Chapter 5
FIGURE 5.1 Key aspects of the SPMD sampling and residue analysis process. Often class fraction-
ation is required following SEC when extracts contain complex mixtures of chemicals. Reprinted
with permission from the American Petroleum Institute (Huckins et al., 2002).
for analytical chemistry of SPMDs, the specific sequence of cleanup and fraction-
ation steps needed for a particular project depends on the goals of the project.
These goals dictate the methods used for sample analysis (e.g., residue analysis
of complex mixtures of contaminants or bioassay) and the required analytical de-
tection and quantitation limits or assay sensitivity. Furthermore matrix effects or
interferences related to the medium sampled and the sampler components, and the
selectivity of the instrument or bioassay used must be taken into consideration.
Herein, the focus is on the analytical chemistry of SPMDs but some of the same
methods are used for bioassays as well.
5.3.1. Cleaning Exposed SPMDs

Before dialytic recovery of concentrated analytes, any periphyton or biofilm,
carbonate salts, etc., on the SPMD membrane are removed by the following se-
quence of steps. First, each individual SPMD is immersed in about 200 mL of
hexane in a glass beaker for about 20 to 30 seconds. Afterwards, the hexane is dis-
carded. Then, SPMDs are placed in a stainless steel pan and washed with copious
amounts of running water (tap water is generally used but a 1 L sample should
be analyzed for any potential interferences prior to use), while being scrubbed
vigorously with a clean toothbrush. At this point the SPMDs are examined for
small holes in the membrane. If a hole is found, and other replicate SPMDs are not
available, the hole is isolated by heat-sealing. After the integrity of each SPMD
has been ensured, they are submerged in a tank of 1M HCl for approximately 30 s
to remove any adhering mineral salts. Following the HCl treatment, the SPMDs
are again rinsed with running water to remove the acid. All water on the membrane
surfaces is removed by brief rinses of high purity acetone, followed by high purity
isopropyl alcohol. After cleaning, SPMDs are allowed to air dry for a minimal
time period (typically <6 min.) on a piece of solvent-rinsed aluminum foil.
5.3.2. Extraction and Cleanup of SPMDs

Because of the very low levels of interferences observed in SPMDs (Lebo
et al., 1995), individual devices can be combined to create a composite sample.
This allows for lower detection and quantitation limits and provides increased
contaminant mass for use in bioassays or other endpoints. Analytes are recovered
from intact SPMDs by OSD (Huckins et al., 1990, 1993; Petty et al., 2000a) in
glass jars fitted with solvent-rinsed aluminum foil under screw-type lids. Interest-
ingly, this dialytic technique is also used for the cleanup of extracts from other
environmental samples and is the subject of several journal articles (Meadows
et al., 1993; Bergqvist et al., 1998b; Strandberg et al., 1998). A minimum of
180 mL of high purity hexane per standard 91.4 cm SPMD per dialytic treatment
is used. SPMDs are dialyzed individually. The procedure consists of 18 hours (h)
of OSD, followed by 4 to 6 h of OSD with fresh solvent (total of 360 mL of
hexane). Dialytic separations are performed at a constant temperature of 18 ◦ C,
because this sub-ambient temperature has been shown to minimize the amount of
co-dialyzed lipid components and low density polyethylene (LDPE) waxes, while
maintaining good-to-excellent recoveries of analytes. (Note that this standardized
extraction method contrasts with the wide variety of solvents and solvent mixtures
used for the extraction of biomonitoring organism [BMO] tissues. Randall et al.
[1991] have shown that extracted lipids vary by about four-fold depending on the
extracting solvent used.) The two dialysates for each sample are combined and
quantitatively transferred to round bottom flasks. The volumes of the dialysates
are reduced to approximately 5 mL using rotary evaporation, quantitatively trans-
ferred to test tubes by filtration through a pre-rinsed glass fiber filter, and the
volumes subsequently reduced to approximately 1 mL. At this point, carryover
of lipids and LDPE waxes for a standard 1 mL triolein (≥95% purity) SPMD
should be <30 mg. When using SPMDs with triolein purified by the Lebo et al.
(2004) method, SPMD dialysates generally can be used for bioassay-biomarker
tests without additional cleanup (see Chapter 6).
The following chromatographic techniques are representative of those used
by a number of investigators for the further enrichment and fractionation of ana-
lytes. The concentrated dialysate is subjected to SEC to remove co-dialyzed lipid
110 Chapter 5
materials and LDPE waxes. A typical SEC system consists of a high performance
liquid chromatograph, equipped with an autosampler, a fraction collector, a UV
detector operated at 254 nm and a 300 mm × 21.2 mm i.d. Phenogel column
(10 µm particle size, 10 nm pore size). The SEC column is from Phenomenex,
Torrance, CA, USA. Equivalent components and columns can be used for the SEC
treatment. The mobile phase consists of 2% methanol in 98% dichloromethane.
The SEC procedure results in the elimination (discarding) of nearly all lipid ma-
terials, LDPE oligomers, and elemental sulfur. The chromatography system must
be calibrated on a daily basis. For example, this can be accomplished by injecting
a solution containing di-2-ethylhexylphthalate (DEHP), biphenyl, naphthalene,
coronene, and elemental sulfur. These compounds elute in the order listed. Fol-
lowing calibration of the SEC system, the fraction collector is adjusted such that
two fractions, “dump” and “collect” are produced for each injection. The “dump”
fraction begins upon sample injection and stops upon initiation of the “collect”
fraction. The “collect” fraction is initiated between the apex of the DEHP peak
and the biphenyl peak. The “collect” fraction is terminated at 70% of the time
between the apex of the coronene chromatographic peak and the apex of the sulfur
chromatographic peak. This SEC procedure removes elemental sulfur (often found
in aqueous deployed SPMDs) from the SPMD sample extracts.
Not all laboratories use SEC for the cleanup of SPMD dialysates. For example,
Booij et al. (2003) used a 0.6 cm i.d. column containing 2 g of silica gel 60
(deactivated with 6% water [wt/wt]) obtained from Merck, Whitehouse Station, NJ,
USA; to purify dialysates from 1 mL triolein SPMDs. The concentrated dialysates
were applied to the silica gel columns, and PAHs and PCBs were quantitatively
eluted with 40 mL of high purity pentane. Less than 0.01 mg of non-target residues
coeluted with analytes in the pentane.
Because different SEC collect-fractions and subsequent enrichment tech-
niques are used for PAHs than for PCBs and OCPs, the sample extracts are split
into two portions before SEC. For the PAH portion of the samples, the “collect”
fraction is initiated at the point 70% of the time between the apex of the DEHP
chromatographic peak and the apex of the biphenyl chromatographic peak. For
the OCP and PCB portion of the samples, the “collect” fraction is initiated at the
point 50% of the time between the apex of the DEHP chromatographic peak and
the apex of the biphenyl chromatographic peak. The “dump” fraction contains the
co-extracted lipid components, the LDPE oligomers, etc. and is discarded to waste.
The fractions collected are amended with about 2 mL of isooctane, reduced to a
volume of about 1 mL on a rotoevaporation system, and quantitatively transferred
with hexane into test tubes. Each collect fraction is reduced in volume to approxi-
mately 1 mL using rotary evaporation and nitrogen (high purity) blow-down. After
transfer to vials or test tubes, the resulting concentrate is typically adjusted to a
final volume of about 2 mL.
Chemical enrichment and class fractionation procedures vary from laboratory
to laboratory. Thus, no universal method for further enrichment and fractionation
exists. However, the following provides some specific examples of commonly used
enrichment and fractionation methods. Following SEC treatment, the SPMD sam-
ple extracts are enriched using open column (glass) adsorption chromatography.
The portion designated for analysis of PCBs and OCPs (1 mL), is applied to an
activated Florisil (heated at 475 ◦ C for 8 hrs and subsequently stored at 130 ◦ C)
column (5 g) and target compounds are eluted with 60 mL of 75:25 (V V −1 ) methyl
tert-butyl ether: hexane. Following volume reduction (1 mL), the eluate is applied
to an activated (130 ◦ C) silica gel column (SG-60; 5 g). Two fractions are collected;
fraction SG-1 (46 mL of hexane) and SG-2 (55 mL of 40:60 [V V −1 ] methyl tert-
butyl ether: hexane). The PCB residues and nonpolar OCPs elute in SG-1, and the
remaining OCP residues elute in SG-2 (several nonpolar OCPs are split between
SG-1 and SG-2). Those portions of the post-SEC sample extracts that are destined
for analysis of PAHs are treated as follows. The solutions (1 mL) are applied to
tri-adsorbent columns consisting of (top to bottom) 3 g phosphoric acid/silica gel;
3 g potassium silicate (KS); and 3 g of activated silica gel (Petty et al., 2004). The
PAHs are eluted from the tri-adsorbent column with 50 mL of 4% methyl tert-butyl
ether in hexane.
5.4. POTENTIAL INTERFERENCES
While the extracts of SPMDs are generally less difficult to purify than are
extracts of tissue or sediment, certain interferences can be problematic for some
types of analyses. The most important of these potential interferences are co-
dialyzed polyethylene oligomers (i.e., the so-called polyethylene waxes), oleic
acid, and methyl oleate. The latter two interferences are residual from the synthesis
of the triolein. Also, oxidation products of triolein may be present in dialysates
of SPMDs that have been exposed (especially in the presence of light) to air for
periods exceeding 30 d. For a standard 1-mL triolein SPMD, the mass of all these
interferences in dialysates is generally <30 mg or about 6 mg g−1 of SPMD
(Huckins et al., 1996). Another potential interference is elemental sulfur, which is
often present in sediment pore water and is concentrated by SPMDs. However, both
polyethylene waxes and elemental sulfur are readily removed using the previously
described SEC procedure.
Unfortunately, small amounts of oleic acid and methyl oleate are generally
present in the post-SEC sample extracts. Both of these lipids can be a source of
analytical interference when the concentrated SEC eluate is evaluated by GC-
MS. However, the interference from oleic acid and methyl oleate is generally
greater for SPMD field blanks, fabrication blanks, and process blanks than it
is for environmentally exposed SPMDs. The lower level of interfering lipid in
environmentally exposed SPMDs is due to the diffusion of much of both methyl
oleate and oleic acid to the exterior membrane surface (during exposures), where
the residues dissipate, degrade or are removed during membrane cleaning.
Oleic acid can be completely removed by using a tri-adsorbent cleanup pro-
cedure or by using a small column of KS (5 g) and eluting with dichloromethane or
112 Chapter 5
any weaker solvent that will successfully elute the targeted compounds. Also, the
previously described Florisil cleanup procedure removes any residual oleic acid in
the post-SEC extracts. However, methyl oleate is often more problematic. While
most of the methyl oleate is eliminated during the aforementioned SEC treatment,
a small portion remains in the sample extract. Because the methyl oleate contains
a polar functional group, it is found in the SG-2 fraction rather than SG-1. Methyl
oleate causes little or no problem when the analysis is performed using GC-ECD
or GC-PID. However, it interferes with GC-FID determination of PAHs or full
scan GC-MS analyses. If necessary, methyl oleate concentrations can be further
reduced (≈99.6% reduction) by another pass through SEC. Also, residual methyl
oleate can be completely removed using destructive techniques, such as cleanup
with sulfuric acid impregnated silica gel. Unfortunately, this approach is only ap-
plicable when targeted compounds (e.g., PCBs, chlorinated dioxins and furans,
and selected OCPs) do not degrade in the presence of strong acid.
Gustavson et al. (2000) developed a convenient and novel solid phase ex-
traction (SPE) method for the removal of methyl oleate from SPMD dialysates
containing PAHs. A small SPE column (1 g or 0.5 g) containing a dual-zone silica
(normal phase)-based restricted-access sorbent (Diazem, Midland, MI, USA) is
used for the separation. The capacity of this sorbent to remove methyl oleate is
about 1.8% (lipid/sorbent; wt wt−1 ). The PAHs are eluted with 19 mL of hexane
and methylene chloride (97:3; V V −1 ) and recoveries of all PAHs are typically
≥72%.
More recently, commercially available SPMDs contain triolein of ≥99% pu-
rity. Furthermore, Lebo et al. (2004) developed an improved approach to greatly
reduce potential interferences caused by methyl oleate, oleic acid, and other in-
terferences, by removing these chemicals from triolein prior to its use in SPMDs.
The method is tailored to the purification of kg quantities of triolein, but it can
be scaled down to accommodate as little as 1 g of triolein. The following is an
example of the purification procedure. One kg of triolein is distributed equally
among 40, 250 mL polypropylene centrifuge tubes. The 25 g portions of triolein
are each partitioned with 200 mL of high purity methanol, and then the tubes
are centrifuged and placed in a freezer (<−20 ◦ C) overnight. The next day, the
methanol supernatants are decanted and discarded. This partitioning step is re-
peated six times. Subsequently, the purified triolein is consolidated in a 2 L round
bottomed flask. Residual methanol is removed using rotary evaporation and the
triolein is distributed among about 20, 50 mL amber bottles (or glass ampoules),
blanketed with argon and stored frozen (<−20 ◦ C). Throughout the procedure,
exposure to light (UV-A and B) and air must be minimized.
The purity of the purified triolein is verified as follows. According to standard
procedures, replicate standard SPMDs are made using the purified triolein. The
SPMDs are dialyzed and the dialysates are subjected to SEC fractionation. (Prior to
SEC, dialysates should have <500 µg methyl oleate, and essentially no oleic acid.)
The fractions collected from the SEC are evaluated by GC with ECD, FID, and
MS. In order to pass the pre-use certification, the dialysate from a standard SPMD,
after SEC, should contain less than 5 µg of methyl oleate, GC-ECD chromatograms
should have no coincident peaks (OCPs typically analyzed for in environmental

samples) at levels greater than 1 ng SPMD−1 , and no coincident ions using MS
detection at a level greater than 20 ng SPMD−1 of the priority pollutant polycyclic
aromatic hydrocarbons and other targeted analytes.
When using purified triolein, most samples are amenable to bioassay after di-
alytic enrichment. For example, Microtox bioassay of dialysates of SPMDs shows
that the SPMDs made with the purified triolein have lower acute toxicities than
dialysates from SPMDs made from unpurified triolein (Johnson, 2001). Finally, ex-
amination of the dialysates using the yeast estrogen screen (YES) assay (Routledge
and Sumpter, 1996) demonstrated that the purification procedure removes all back-
ground estrogenic activity (Lebo et al., 2004). Use of triolein purified by this
process expands the potential applicability of SPMD sample extracts to include
numerous bioassay procedures (see Chapter 6) and GC-MS as a standard analysis
technique.
5.5. INSTRUMENTAL ANALYSIS
After solvent exchange, the SPMD dialysate can be analyzed directly by

using high performance liquid chromatography (HPLC). If GC or GC-MS is the
instrument of choice, then SEC cleanup and other cleanup and fractionation steps
are generally required for best results (Figure 5.1). Although, almost any analytical
technique used for determining the presence and concentrations of chemicals in
environmental matrices can be applied to the analysis of chemicals in SPMD
extracts, the types and levels of chemicals expected to be present often dictate
the choice of instrumental requirements. For example, the need for instrumental
specificity is underscored by the work of Petty et al. (2000b), where analysis of
a single SG-2 fraction (i.e., the OCP fraction) of SPMDs exposed to indoor air
revealed approximately 400 detectable components (see Chapter 8 for more details
of this study). In this case, it was essential to employ highly selective, mass specific
instrumentation such as GC-MS or HPLC-MS to confirm the presence of the
analytes and to identify unknowns. These considerations also apply to the analysis
of other environmental matrices. However, for the majority of SPMD projects,
investigators have flexibility to choose analytical instrumentation and instrumental
methods assuming good laboratory and chromatographic practices are followed
(e.g., Meadows et al., 1998; Petty et al., 1998a; McCarthy and Gale, 2001).
5.6. DATA FORMAT AND COMPARABILITY
Care should be taken to supply sufficient background information when re-

sults from SPMD deployments are reported. The purpose of this background
information is to elucidate how the reported results are related to the raw data,
to allow users to compare the reported results to those obtained in other studies,
and to provide data of known quality for future reference.
114 Chapter 5
5.6.1. Important Background Information

Historically, the results obtained from SPMD deployments have been reported
in a number of different units: ng per gram SPMD, ng per mL of SPMD, ng per
SPMD, ng per gram triolein, or ng per L water. The general practice is to ana-
lyze whole SPMDs, but some investigators discard the LDPE and only analyze
the triolein. In addition, some SPMD designs have significantly deviated from the
standard configuration. In order to avoid confusion about the design of SPMDs
used in a study, the following characteristics should be specified in reports: LDPE
membrane thickness, SPMD length and width, triolein purity, and triolein mass
fraction. These characteristics fully define the SPMD design used. Other charac-
teristics, such as SPMD surface area, mass, and volume may be listed as well. The
label “standard design” is not sufficient to fully characterize the SPMD design,
because the standard design allows for a range of LDPE membrane thicknesses
(70–95 µm) and triolein purities (≥95%).
A number of exposure-site characteristics should also be listed in order to
document possible effects of flow, temperature, and biofouling on the uptake rates.
Flow conditions at the membrane surface are particularly important because of
the potentially large impact on SPMD sampling rates. Next to external or bulk
flow rates, the design or geometry of the deployment apparatus and mounting
devices may affect the flow conditions at the SPMD surface the most (Louch et al.,
2003), and therefore, should be briefly described. Although exchange kinetics
information may adequately be summarized by listing the release rate constants
(ke s) of PRCs, the documentation of exposure conditions is invaluable for a better
understanding and interpretation of rate constants, both within and among studies.
The necessity of including a brief general description of the exposure sites, their
geographical coordinates, the beginning and end of the exposure period and any
events that may result in field blank contamination may seem self-evident, but
should not be forgotten. Exposure site pH is a relevant parameter when acidic
or basic contaminants are targeted. Other useful data include: suspended matter,
particulate and dissolved organic carbon (POC and DOC, respectively) contents
of exposure water and the dust content of air. In regard to atmospheric exposures,
the predominant wind directions should be noted as it may help to interpret the
absorbed amounts in terms of geographical sources.
Raw data from SPMD exposures should include target compound levels in
the various blanks, and the amounts found in exposed SPMDs in ng per sample.
When the triolein phase is separately analyzed, the raw data should include the
amounts in the LDPE phase as well. When PRCs are used, the measured t = 0
levels and the calculated spike levels should be included as an additional check on
PRC recovery.
Information on whether the average blank levels were subtracted should also
be supplied with the processed data. In addition, the method detection limits
(MDLs) and MQLs should be listed, as well as the criteria used to determine MDLs
and MQLs. The methods or equations used to calculate aqueous or atmospheric
TABLE 5.1 Important Background Information to be Supplied with SPMD Data
Stage Required Information
SPMD design LDPE membrane thickness

length and width (excluding mounting loops)
triolein mass fraction
triolein purity
Exposure water or air flow rate
exposure cage geometry
air samplers: wind direction
location (coordinates, and preferably a digital picture)
start and end date and time
physical appearance after exposure (description, and preferably a digital picture)
temperature
pH (when organic acids or bases are the target analytes)
suspended matter, and POC and DOC levels (water) or particle (air) levels
other observations relevant to QC
Raw data PRC levels in non-exposed SPMDs (experimental and calculated)
blank levels (fabrication, process, reagent, field)
fraction processed (whole SPMD recommended!)
amount of analyte per sampler
Processed data blank corrections, MDL, MQL
concentration in SPMD (when applicable)
Cw or Ca calculation method (when applicable)
concentrations should be fully documented. The main reason for recording this
information is that calculation methods are subject to continual improvement, as
more calibration data become available. Other important information includes the
sources of sampling rates and partition coefficients, the methods used to calculate
PRC-derived sampling rates, and the schemes applied for estimating the sampling
rates of non-PRC analytes. Because there often is some ambiguity in the choice
of literature values of log K ow (aqueous exposures) and log octanol-air partition
coefficient (K oa ) values (atmospheric exposures), a full listing of these parameters
is essential.
A summary of essential background information to be supplied with SPMD
data is given in Table 5.1. Compliance with this list secures valuable data for the
future, with a relatively small investment of time.
5.6.2. The Triolein Phase and Lipid Normalization

Processing of the triolein alone (i.e., without analyzing the LDPE as well)
is not encouraged for a number of reasons. First of all, the LDPE constitutes a
significant part of the total SPMD sorption capacity, in contrast to biota, where the
sorption capacity of the non-lipid phase is often considered to be negligible. Data
on membrane-lipid partition coefficients (K mL ) is very limited, but the available
116 Chapter 5
information suggests that these partition coefficients are in the range 0.1 to 0.6 g g−1
(Huckins et al., 1990, 1993; Hofmans, 1998). Given the fact that the LDPE mass in
SPMDs is four times as large as the triolein mass, these values imply that the LDPE
membrane contains about 30 to 70% of the analyte amounts that are absorbed by
the SPMDs. Discarding the LDPE phase would therefore result in a significant
loss of analyte. A second reason why the LDPE phase should not be discarded
is that by far the greatest part of the calibration data are based on whole-SPMD
data rather than on triolein-only data. Because of the limited numbers of measured
K mL values, it is in practice very difficult, if not impossible, to accurately convert
calibration data for whole-SPMDs into sampling rate data for the triolein phase
alone.
The practice of lipid normalization finds its origin in biomonitoring research
and equilibrium partition (EP) theory, where it was observed that biota with larger
lipid contents generally had higher levels of bioconcentratable chemicals. The
main prerequisite for lipid normalization of contaminant concentrations in BMOs
is that thermodynamic equilibrium is attained for the compounds of interest. When
concentrations in BMOs are to be compared among different species, lipid nor-
malization is only helpful when this condition is met for all BMO samples. Other
assumptions related to lipid normalization are that differences in the composition
of organism lipids (Schneider, 1982) and in the physiological state among the
BMOs are negligible (Huckins et al., 2004). In regard to biomagnification (see
Section 7.8 for definition)-mediated deviations from EP theory, corrections are
made by using food chain multipliers. Because SPMDs can be viewed as a lipid
pool contained in a non-lipid matrix, it is tempting to apply most of these con-
cepts of lipid normalization to SPMDs as well. Unfortunately, this practice leads
to results that are often difficult to interpret and to conclusions that may be plainly
wrong. First, the non-lipid phase of SPMDs contains a significant fraction of the
analytes. This contribution of the membrane to the total sorption capacity of the
device could be accounted for in terms of a “lipid-equivalent mass” (m Leq )
m Leq = m L + K mL m m (5.1)
where m L and m m are the mass of the triolein and the LDPE membrane, re-
spectively. A complicating factor is that the lipid-equivalent mass differs among
compounds, and often cannot be calculated because of lack of K mL data. Using
the range of K mL values given earlier (K mL = 0.1 to 0.6 g g−1 ), and the mass frac-
tion of a standard-design SPMD (20% triolein and 80% membrane), m Leq values
range from 28 to 68% of the SPMD, depending on the compound. The compound-
dependent lipid-equivalent mass of an SPMD is difficult conceptually, and thus
may lead to errors in the normalization process.
As suggested earlier, all samples must attain equilibrium before lipid nor-
malization is appropriate. However, SPMDs are not designed to reach equilibrium
during typical exposure periods of one month or less. In fact, attainment of equi-
librium may take years for some compounds and exposure conditions. Prior to
equilibrium, the amount of a contaminant accumulated by an SPMD or BMO is
not limited by the size of the lipid pool, but rather by the contaminant sampling rate,
which is proportional to the surface area of the sampler-water interface. In this case,
lipid-normalization clearly leads to erroneous results. For example, Meadows et al.
(1998) reported that the uptake rate constants (ku s) of SPMDs (≈30–70% effective
lipid content) and brown trout (Salmo trutta) (≈2–3% lipid content) are similar
when expressed on a wet weight (tissue) and whole SPMD basis. If this study
was extended and trout had just attained equilibrium concentrations, the levels in
SPMDs would represent only about 4–10% of the equilibrium concentrations. At
this point in time, lipid normalization of trout and SPMD concentrations would
result in an apparent ≥10-fold higher concentration in trout than in SPMDs, even
though the actual amounts of residues sequestered per gram of whole sampling
matrix would be equivalent. Unfortunately, normalization artifacts or errors occur
quite often in reports comparing SPMDs and BMOs.
5.7. REFERENCES
Bergqvist, P.-A.; Strandberg, B.; Ekelund, R.; Rappe, C.; Granmo, Å. 1998a, Temporal monitoring
of organochlorine compounds in seawater by semipermeable membranes following a flooding
episode in Western Europe. Environ. Sci. Technol. 32: 3887–3892.
Bergqvist, P.-A.; Strandberg, B.; Rappe, C. 1998b, Lipid removal using semipermeable membranes in
Booij, K.; Hofmans, H.E.; Fischer, C.V.; van Weerlee, E.M. 2003, Temperature-dependent uptake rates
lene membranes.Environ. Sci. Technol. 37: 361–366.
DeVita, W.M. and Crunkilton, R.L. 1998, Quality Control Associated with Use of Semipermeable
Membrane Devices. In Environmental Toxicology & Risk Assessment, Seventh Volume; Little,
E.E., Delonay, A.J., Greenberg, B.M., Eds.; ASTM STP 1333; American Society for Testing and
Materials: West Conshocken, PA; pp. 237–245.
Górecki, T. and Pawliszyn, J. 1997, Field-portable solid-phase microextraction/fast GC system for
trace analysis. Field Anal. Chem. Tech. 1: 227–284.
Gustavson, K.E.; DeVita, W.; Revis, A.; Harkin, J.M. 2000, A novel use of a dual-zone restricted
access sorbent: Normal phase separations of methyl oleate and polynuclear aromatic hydrocarbons
stemming from semipermeable membrane devices. J. Chromatogr. A 883: 143–149.
Devices. MSc Thesis, Netherlands Institute for Sea Research: Den Burg, The Netherlands.
Horwitz, W.; Kamps, L.R.; Boyer, K.W. 1980, Quality assurance in the analysis of foods for trace
constituents. J. AOAC 63: 1344–1354.
Huckins, J.N.; Tubergen, M.W.; Manuweera, G.K. 1990, Semipermeable membrane devices contain-
27: 2489–2496.
118 Chapter 5
Johnson, B.T. 2001, USGS Columbia Environmental Research Center, Columbia, MO. Personal com-
munication.
Jones-Lepp, T. 2004, U.S. EPA, National Exposure Research Laboratory: Las Vegas, NV; Personal
communication.
Lebo, J.A.; Almeida, F.V.; Cranor, W.L.; Petty, J.D.; Huckins, J.N.; Rastall, A.; Alvarez, D.A.; Mo-
gensen, B.B.; Johnson, B.T. 2004, Purification of triolein for use in semipermeable membrane
devices (SPMDs). Chemosphere 54: 1217–1224.
Maštovská, K. and Lehotay, S.J. 2003, Practical approaches to fast gas chromatography-mass spec-
trometry. J. Chromatogr. A 1000: 153–180.
McCarthy, K.A. and Gale, R.W. 2001, Evaluation of persistent hydrophobic organic compounds
in the Columbia River Basin using semipermeable membrane devices. Hydrol. Process 15:
1271–1283.
Meyer, M. 2004, USGS Organic Geochemistry Research Laboratory: Lawrence KS; Personal commu-
nication.
Petty, J.D.; Poulton, B.C.; Charbonneau, C.S.; Huckins, J.N.; Jones, S.B.; Cameron, J.T.; Prest, H.F.
1998a, Determination of bioavailable contaminants in the Lower Missouri River following the
flood of 1993. Environ. Sci. Technol. 32: 837–842.
Petty, J.D.; Huckins, J.N.; Cranor, W.L.; Clark, R.C. 1998b, Determination of Dieldrin in Ground Water
from the George Air Force Base; California. USGS Columbia Environmental Research Center:
Columbia, MO; Unpublished report to U.S. EPA, San Francisco; CA.
Petty, J.D.; Orazio, C.E.; Huckins, J.N.; Gale, R.W.; Lebo, J.A.; Meadows, J.C.; Echols, K.R.; Cra-
nor, W.I. 2000a, Considerations involved with the use of semipermeable membrane devices for
monitoring environmental contaminants. J. Chromatogr. A 879: 83–95.
Petty, J.D.; Gale, R.W.; Huckins, J.N.; Cranor, W.L.; Alvarez, D.A.; Clark, R.C. 2000b, Devel-
opment and Application of Techniques for Sampling Bioavailable Airborne Contaminants-
Tentatively Identified Compounds by Gas Chromatography/Mass Spectrometry. USGS Columbia
Environmental Research Center : Columbia, MO; Unpublished report to U.S. EPA National
Exposure Assessment Laboratory: Las Vegas; NV.
contaminants. Chemosphere 54: 695–705.
Randal, R.C.; Lee, H. II.; Ozretich, R.J.; Lake, J.L.; Pruell, R.J. 1991, Evaluation of selected lipid
methods for normalizing pollutant bioaccumulation. Environ. Toxicol. Chem. 10: 1431–1436.
Routledge, E.J. and Sumpter, J.P. 1996, Estrogenic activity of surfactants and some of their degradation
products assessed using a recombinant yeast screen. Environ. Toxicol. Chem. 15: 214–248.
Schneider, R. 1982, Polychlorinated biphenyls (PCBs) in cod tissues from the Western Baltic: Sig-
nificance of equilibrium partitioning and lipid composition in the bioaccumulation of lipophilic
pollutants in gill-breathing animals. Meeresforschung 29: 69–79.
526–533.
Taylor; J.K. 1987, Quality Assurance of Chemical Measurements; Lewis Publishers: Boca Ratan, FL.
Chapter 6
Bioassay of SPMD
Extracts or Diluents
6.1. OVERVIEW AND RATIONALE OF THE SPMD-TOXICITY

SCREENING APPROACH
In addition to instrumental methods of qualitative and quantitative analyses,

chemicals sequestered by SPMDs are amenable to examination by a variety of
bioassays (Huckins et al., 1996; Zajicek et al., 1996; Parrott and Tillitt, 1997;
Johnson, 1998; Johnson et al., 2004; Petty et al., 1998, 2000, 2004; Parrott et al.,
1999; Sabaliūnas et al., 1999, 2000; Rastall et al., 2004). These assays include
biomarker or bioindicator tests, immunoassays, and classic toxicity tests. In this
chapter the words “bioindicator” and “biomarker” are used interchangeably but the
tests refer to in vitro and in vivo indicators of toxicity and exposure. Bioassays used
to assess extracts or diluents of SPMDs include but are not limited to the following:
Microtox, Mutatox, mixed function oxygenase (MFO) induction-ethoxyresorufin-
O-deethylase (EROD) activity, sister chromatid exchange, vitellogenin (VGT)
induction via interperitoneal injection of test species, enzyme-linked immunosor-
bent assay (ELISA), Daphtoxkit F, Ames mutagenicity test, and yeast estrogen
screen (YES) assay. Some of these assays incorporate metabolic activation sys-
tems (e.g., Mutatox and EROD). This approach makes it possible to account for
metabolic transformation processes that may enhance or reduce the effects of chem-
ical residues in organism tissues. The marriage of SPMDs and compatible bioas-
says offers many avenues of investigation, all potentially providing information
121
122 Chapter 6
concerning the relative toxicological significance of exposure to chemicals present

in the environmental matrices sampled.
A major challenge for ecotoxicologists is to obtain relevant samples suitable
for testing from environmental systems. Most hydrophobic organic contaminants
are present in environmental matrices only at trace- to ultra-trace- levels. However,
the sometimes-slow processes of bioconcentration (uptake via respiration-dermal
absorption) and bioaccumulation (uptake from both respiration-dermal absorption
and diet) can lead to elevated concentrations of contaminants in exposed organisms,
which can result in a variety of adverse effects. These chemical uptake processes are
especially relevant to the magnitude of adverse environmental effects of persistent
hydrophobic organic contaminants. Those contaminants with log octanol-water
partition coefficients (K ow s) between 4 and 7 are of particular concern. In many
cases, bioassays (especially the more rapid cellular-based assays) do not fully ac-
count for the effect of bioaccumulation on tissue concentrations of aquatic organ-
isms. Also, a number of the aforementioned assays have relatively low sensitivities
for many common pollutants. For example, the Microtox test often requires high
ng to low µg amounts of priority pollutants to elicit a measurable response. Direct
testing of environmental waters with this bioassay may lead to false-negative errors
in assessing the potential risk of waterborne residues to aquatic life. To avoid this
type of error and expand the use of biomarker tests for ranking toxicity potential,
a pre-concentration method is needed that mimics the bioconcentration process.
SPMDs offer several advantages over other potential pre-concentration methods,
including: 1) a biomimetic design; 2) only bioavailable, dissolved-phase (in the
case of waterborne chemicals) residues are sampled; 3) sampling is generally in-
tegrative, which permits the SPMD sample to reflect contribution from episodic
contamination events; and 4) significant statistical advantages, due to high re-
producibility, relative to biomonitors (Prest et al., 1997; Huckins et al., 1998).
Consequently, the SPMD-bioassay assessment approach enhances an investiga-
tor’s ability to screen for the relative toxicological significance of environmental
hydrophobic organic chemical (HOC) residues.
If an investigator demonstrates that an SPMD extract is toxic or genotoxic
when using a specific biomarker test, questions may arise as to the relevance of the
finding in regard to risk assessment. Clearly, the SPMD-bioassay combination is
useful as a screening tool for ranking the potential toxicities of bioconcentratable
residues at multiple sites and for determining sources of pollutants. However, the
justification for a specific level of pre-concentration prior to use of bioassay tests
to account for the toxicological effects of residue bioconcentration in tissues is
less clear. This is, in part, due to the wide variation in bioconcentration factors
(BCFs) for the same chemical among different species. Also, depending on the
specifics of biomarker experimental conditions (e.g., exposure or incubation time)
and the type of organisms or cell lines used, chemicals in the incubation medium
are also concentrated by the test organisms to some unknown degree. Although
the exposure duration is only five minutes for the basic Microtox test, the high
surface area to volume ratio of bacteria likely facilitates uptake rates.
Bioassay of SPMD Extracts or Diluents 123
FIGURE 6.1 Illustration of the SPMD in vitro bioassay and immunoassay protocol, which in-
cludes sampling, different levels of processing and enrichment, and bioassay or immunoassay.
Reprinted with permission from the American Petroleum Institute (Huckins et al., 2002).
Regardless of the complexity of predicting an appropriate pre-concentration

factor, some level of pre-concentration is often needed to permit application of
bioindicator tests and certain other bioassays to the assessment of trace or ultra-
trace contaminants. Obviously, the maximum level of pre-concentration, or the
time-dependent SPMD concentration factor (CF; see definition in Section 7.5)
should not exceed the measured or estimated BCF of a chemical for the species of
concern. However, this may not be true when organisms biotransform compounds
of concern. Clearly, metabolic activity differences among species can affect residue
body burdens (Connell, 1990). Often, investigators choose SPMD exposure dura-
tions on the basis of convenience. Generally, exposure periods seldom exceed 30
days (EROD for ultra-trace levels of dioxins, furans and coplanar PCBs may be an
exception). Exposures of 30 days or less generally result in SPMD CFs less than
the reported BCFs of many stable hydrophobic compounds in test organisms.
SPMD dialysates (extracts), rinses of the exterior membrane surface (only
SPMDs exposed to air), and aliquots thereof often contain a number of classes of
chemicals. Figure 6.1 shows various levels of processing and enrichment used for
SPMD derived bioassay samples. We strongly recommend the use of SPMDs with
triolein purified by the method of Lebo et al. (2004) for bioassays to reduce the
probability of false-positive results or for controls that fail to meet quality control
124 Chapter 6
(QC) criteria. After transferring the enriched SPMD extract to an appropriate

carrier solvent such as dimethyl sulfoxide (DMSO), the toxicity of the sample
can be assessed with a wide variety of bioassays. If the test endpoint is an effects
concentration for 50% of test organisms (EC50 ), the units are a specific mass or
volume (mg or mL) of whole SPMD per volume (mL or L) of carrier solvent.
Obviously, investigators must keep track of sample splits, dilutions, etc. When
analytical chemistry is also performed and SPMD uptake rate data are available
for the detected residues, the measured EC50 value can be related to the volume
of water (Vw−tox ) sampled at a site that contained sufficient mass of bioavailable
contaminants to elicit the observed toxicological response. If the above conditions
are met, the following model can be used to derive Vw−tox
Vw−tox = ku t EC50 Vc (6.1)
where the ku t represents the volume of water extracted per g of SPMD, and Vc
is the volume of carrier solvent (includes correction for any serial dilutions) used
in each determination of the EC50 value. If ku s are available for the chemicals
identified in the SPMDs and for the aquatic organisms of interest (e.g., see Mackay
et al., 1992a; 1992b; 1997), then Vw−tox Cw (represents the mass of bioavailable
contaminants eliciting the bioassay response) can be compared to organism ku t
Mo Cw (mass of bioavailable contaminants exposed to organism of concern during
a specified time interval [t]); where Mo is organism mass in g. Obviously, this
approach is limited by a general lack of applicable rate constant data, the potential
for unidentified contaminants to elicit toxic responses, and the paucity of data
on chemical concentrations in biomarker cell lines or organisms (i.e., the CF
associated with the transfer of chemicals from the test medium to the test cells or
organisms).
Kočı́ et al. (2003) developed the Vtox approach for assessing the relative
toxicity of SPMD extracts. They defined Vtox(50) as the media volume that is needed
to reduce the toxicity of the extract of a 1-day exposed standard SPMD to a 50%
effect level. The following relationship is used to determine Vtox
Vtox(50) = 1/(sn EC50 d) (6.2)
where Vtox has units of mL d−1 , sn is the number of standard SPMDs per mL of
carrier solvent used for the assay, the EC50 is the 50% effect concentration (mL
carrier solvent per mL test solution), and d is days of exposure.
The following sections describe several commonly used bioassay tests that
have been successfully used in conjunction with SPMDs. However, no attempt
was made to cover all of the in vitro assays used for SPMD extracts.
6.2. MICROTOX AND MUTATOX
Several investigators (e.g., Huckins et al., 1996; Cleveland et al., 1997;

Johnson, 1998; Johnson et al., 2004; Sabaliūnas et al., 1999, 2000; Petty et al.,
TABLE 6.1 Toxicological Evaluation of Polycyclic Aromatic Hydrocarbons (PAHs)

with Microtox Basic Test and Mutatoxa . Reprinted with permission from the
American Petroleum Institute (Huckins et al., 2002)
Microtox
Mutatox
Compounds EC50 b CIc genotoxicityd
acenaphthylene 0.34 0.25–0.47 positive

phenanthrene 0.48 0.33–0.68 positive
fluorene 0.50 0.35–0.70 positive
anthracene 0.64 0.53–0.78 positive
benz[a]anthracene 0.73 0.65–0.81 positive
acenaphthene 0.75 0.69–0.81 positive
2-aminoanthracene 0.75 0.49–1.2 positive
fluoranthene 0.83 0.63–1.08 positive
naphthalene 0.90 0.85–0.99 positive
chrysene 0.92 0.85–0.99 positive
2-aminonaphthalene 1.3 1.1–1.5 positive
2-acetamidofluorene 2.3 1.3–4.1 positive
2-aminofluorene 4.1 2.5–6.4 positive
benzo[a]pyrene 10.7 6.4–18.2 positive
3-methylcholanthracene 19.9 18.3–21.5 positive
7,12-dimethylbenzanthracene 33.1 14.6–74.7 positive
pyrene >500 — positive
DMSO (control) NDe — negative
a
Data from Johnson (1998).
b
5 minute EC50 = µg mL−1 .
c
CI = 95% confidence interval.
d
1% rat S9 activation.
e
ND = not detected.
2000) have determined the toxicity and/or genotoxicity (i.e., DNA-damaging po-
tential) of purified SPMD extracts, of SPMD lipid, or of liver homogenates obtained
from organisms exposed to SPMD extracts using the Microtox and Mutatox assays
(Microbics, 1992). Unfortunately, the Mutatox cell line is currently unavailable
from the supplier, but we include information on the test in the hope of future avail-
ability. The Microtox in vitro test is based on the chemically induced reduction
in the level of light generated by bioluminescent bacterium Vibrio fischeri, while
the Mutatox in vitro test is based on a chemically induced increase in light from
a dark-mutant strain of V. fischeri. The degree of the decrease in light (Microtox)
or increase in light (Mutatox), when compared to controls, indicates the relative
acute toxicity (i.e., the basic Microtox test) and genotoxicity (Mutatox), respec-
tively of the sample extract. The toxicological endpoint for the Microtox test is
an EC50 value and 95% confidence interval (i.e., the test is quantitative), whereas
the endpoint for the Mutatox test is qualitative, providing a yes or no assessment
of the presence of DNA-damaging substances. Johnson (1998) has determined
the acute toxicities and genotoxicities of many chemicals, including PAHs (see
Table 6.1). Table 6.2 gives an example of using Microtox and Mutatox to determine
126 Chapter 6
TABLE 6.2 Use of Microtox and Mutatox to Determine the Toxicity of SPMD
Concentrates. Reprinted with permission from the American Petroleum Institute
(Huckins et al., 2002)
Microtox Mutatox
Sample type toxicity EC50 a genotoxicity
SPMDs
Winter Quarters Bayb 3.1 (2.9 – 3.3) f negative
McMurdo Soundb 88 (28 – 275) negative
Flat Branchc NAg positive
Quality control
procedural blankd NDh negative
laboratory blank SPMDe ND negative
microtox phenol reference toxicant (µg mL−1 H2O) 19 (17 – 21) NA
mutatox benzo[a]pyrene reference toxicant(1 µg /Vial) NA positive
a
Assays were conducted on lipid diluent or dialysates and EC50 values represent mg of SPMD lipid mL−1 carrier
solvent.
b
SPMDs exposed to Antarctica sediments in microcosms (Huckins et al., 1996).
c
SPMDs exposed to a small urban stream (Lebo et al., 1992).
d
Solvents and reagents used in tests.
e
Freshly prepared blank SPMD, carried through Microtox and Mutatox test.
f
Microtox values are 5-minute EC50s with 95% confidence intervals (in parentheses).
g
None analyzed.
h
None detected.
the potential toxicities of SPMD extracts from two separate studies (Huckins et al.,
1996). Note that EC50 values given in Table 6.2 are given in units of mg SPMD
lipid per mL of carrier solvent. As discussed in Section 5.6.2., SPMD extracts
generally represent whole SPMDs, thus, units of mg or mL of whole SPMDs are
advised. The positive Mutatox response for the Flat Branch sample (Table 6.2) is
not surprising because of the relatively high levels of PAHs detected there by Lebo
et al. (1992).
Microtox responds well to a wide array of hydrophobic chemicals (see the
work of Johnson [1998] in which a variety of pesticides, industrial chemicals
and petroleum products were tested). Apparently, Microtox is highly responsive
to compounds with a narcosis mode of toxicity (Johnson, 1998; Sabaliūnas et al.,
1998). As shown by Johnson (1998) chemicals that elicit narcosis include multiple
chemical classes.
Clearly, the mode(s) of action eliciting a genotoxic response are more
chemical-structure specific (Johnson, 1998). As suggested by Johnson (1998),
in vitro metabolic activation is required to assess the genotoxicity of SPMD
residues with Mutatox. Typically, a rat liver S9 fraction is used for the exoge-
nous metabolic activation step (Microbics, 1992). At first glance, Mutatox appears
to be well suited for the assessment of SPMD extracts. However, Sabaliūnas et al.
(2000) have pointed out several potential difficulties and shortcomings of the test
in its present form. These include reduced light intensity due to cytotoxicity or
cell death, delays in the genotoxic response of some samples beyond standard
measurement times, and lower sensitivity of measurements based on reverse mu-

tations. Test sensitivity is not an issue when SPMDs are used to pre-concentrate
samples. Also, in the Mutatox protocol (Johnson, 1998), a sample is designated
as genotoxic only when two positive responses are recorded, each at different
concentrations within a single dilution series. Generally, cytotoxic effects should
be evident from the shape of the dose-response curve, and increased turbidity
in exposed samples relative to controls is also used as a cytotoxicity indicator
(Johnson, 1998; Johnson et al., 2004). However, the concerns raised by Sabaliūnas
et al. (1999), require further investigation before Mutatox can be used with total
confidence.
6.3. MIXED FUNCTION

OXIDASE-7-ETHOXYRESORUFIN-O-DEETHYLASE (MFO-EROD)
SPMDs are often applied to concentrate trace levels of environmental con-

taminants that induce mixed function oxidase (MFO) activity (Huckins et al., 1996;
Parrott and Tillitt, 1997; Parrott et al., 1999; Petty et al., 2000). MFOs are a category
of enzymes that aid in the biotransformation and clearance of many hydrophobic
compounds. One of the most widely recognized MFO enzymes is cytochrome
P4501A1 (often measured as ethoxyresorufin-O-deethylase [EROD] activity). In-
creased EROD activity indicates exposure to certain types of organic compounds.
These include planar polycyclic aromatic compounds, either halogenated (e.g.,
polychlorinated dioxins and furans and planar PCBs) or PAHs. Because determi-
nation of EROD activity is relatively simple (Parrott et al., 1999), this endpoint is
often used as a screening tool for the assessment of sites potentially contaminated
with these classes of chemicals. Specific cell lines, e.g., H4IIE rat hepatoma cells
and the PLHC-1 fish (Poeciliopsis lucida) hepatoma cells are most commonly
used for the measurement of EROD activity (Tillitt et al., 1991; Lebo et al., 1995;
Huckins et al., 1996; Parrott and Tillitt, 1997, Whyte et al., 2004). However, ex-
posed organisms can also be employed in determining EROD activity (e.g., Petty
et al., 1998, 2000; Whyte et al., 2000). In this approach, enzymatic activity is stan-
dardized to cellular protein content using the method of Lorenzen and Kennedy
(1993). This method corrects for the attenuation of EROD due to cytotoxicity.
EROD activity is measured in the H4IIE cells as follows. The cells are seeded
at 7,000 cells per well in 250 µL of Dulbecco’s modified Eagles culture media
(Tillitt et al., 1991). After an initial incubation period of 24 hours, the cells are
dosed with 5 µL volumes of enriched SPMD extracts (cleanup of extracts gener-
ally includes dialysis and size exclusion chromatography [SEC]) and incubated
for an additional 72 hours. Sample dose is typically expressed as g-equivalents tri-
olein or whole SPMD per mg cellular protein. Multiple exposures are performed at
each of six (typically) sample concentrations, using a dilution series. Afterwards,
the microtiter plates are washed three times with distilled water to lyse the cells.
EROD activity (pmol mg−1 cellular protein per min) in each sample is measured
128 Chapter 6
FIGURE 6.2 Ethoxyresorufin-O-deethylase (EROD) induction in H4IIE rat hepatoma cells exposed
to an SPMD dialysate. Four Standard SPMDs were deployed for a 28 day period in Bayou Meto,
Arkansas. Doses of dialysate were normalized to gram-equivalents of triolein per mg of cellular
protein. This figure was generated by Don Tillitt, USGS-CERC, Columbia, MO, USA and is
reprinted with permission from the American Petroleum Institute (Huckins et al., 2002).
kinetically, and the linear portion of the sample dose-response curve is compared
to a 2,3,7,8-tetrachlorodibenzo- p-dioxin (TCDD) standard response (pg TCDD
mg−1 cellular protein) curve. The standard response curve is generally based on
eight concentrations of TCDD, and it is used to quantify the total toxic equiv-
alents (TEQs) of samples (Gale et al., 2000). TEQ values of samples represent
the concentrations of TCDD required to give equivalent EROD responses. Ankley
et al. (1991) and Whyte et al. (2004) have given details of the procedure for TEQ
calculation.
Figure 6.2, shows a specific example of the use of the H4IIE-EROD assay for
purified SPMD extracts. After purification of SPMD extracts, Gale et al. (2000)
showed that H4IIE-determined TEQs were well correlated with instrumentally-
determined TEQs.
6.4. NEUROTOXICITY ENDPOINTS
The use of SPMDs to sequester hydrophobic contaminants for incorporation

into bioindicator test-based screening is increasing in both frequency of application
and in the array of modes of action. For example, as a focused part of a broader
survey of ecological conditions, species diversity, and habitat quality, SPMDs

were deployed in two major water sources of the Santa Cruz River, near Nogales,
Arizona, USA. The Santa Cruz River rises in the United States, flows southward
into Mexico, turns north and flows back into the United States near Nogales,
Arizona. This river is a major source of water in this very arid region and is of
critical importance to both the United States and Mexico. Because of the multiple
uses of the water of the Santa Cruz River, it is important to determine the presence of
waterborne contaminants and the potential for adverse effects on aquatic organisms
and ultimately humans. The SPMD samplers were deployed in the effluent of the
International Wastewater Treatment Plant (IWWTP) and the Nogales Wash, which
are two major water sources of the Santa Cruz River at Nogales, Arizona, USA
(Petty et al., 2000).
Following dialysis and treatment by SEC, the sample extracts were solvent
exchanged into sterile DMSO. Subsequently, four rainbow trout (Oncorhynchus
mykiss [RBT]) were placed in each of seven tanks (each tank is considered as a
treatment and a replicate is an individual fish within a tank) in 18 ◦ C well water
(280 mg L−1 hardness as CaCO3 ) using flow-through conditions. RBT were fed
once daily throughout the study. Following a 48 hour acclimation, RBT were
injected interperitoneally with 100 µL of a 1:1 mixture of an SPMD extract or ap-
propriate controls in DMSO or corn oil. Controls included non-deployed SPMD
extracts, SEC blanks, and DMSO blanks. The same injection procedure was re-
peated 6 days later. RBT were sacrificed 11 days after initial exposure to the
extracts, and the plasma, liver, gills, and brain were immediately removed from
each fish and maintained at −80 ◦ C until assayed.
Brain tissue was homogenized and portions were used to determine acetyl-
cholinesterase activity (AChE), and for muscarinic cholinergic receptor (MChR)
and β-andrenoceptor (βAR) binding in vitro assays. Determination of AChE was
performed according to the method of Gard and Hooper (1993) as modified by
Beauvais (1997) in which the activity of the enzyme is measured by hydrolysis
of acetylcholine and the reaction of the thiocholine product with a colorimetric
reagent. The MChR binding assay was performed according to the method of Jones
and King (1995) in which crude membrane preparations are incubated with a radio-
ligand specific for MChR. βAR binding was assayed following a standard method
(Steevens et al., 1996). After incubation, samples were filtered through GF/B
glass fiber filters, with radioactivity being measured using a liquid scintillation
counter at an efficiency of 46%. None of the blanks or control samples elicited any
measurable response in these assays, indicative of a controlled and interference-
free set of samples. The results of these assays are presented in Tables 6.3
and 6.4.
An examination of the data reveals that the average values of plasma
cholinesterase were depressed in fish exposed to the SPMD sample extracts from
the IWWTP site compared to all controls, although only significantly lower com-
pared to the SEC blanks ( p < 0.05). A similar trend was observed with the Nogales
Wash sample extracts. A number of chemicals determined to be present in the
130 Chapter 6
TABLE 6.3 Brain MChR Binding and Brain and Plasma Cholinesterase Activities in
Rainbow Trout (Oncorhynchus mykiss) Exposed to Sample Extracts and Controls.
Reprinted from Petty et al. (2000), copyright (2000); reproduced with permission
from Elsevier
Binding MChR Cholinesterasec
Exposure medium KDa BMAX b Brain Plasma
Controls
DMSOd , n = 3 62 ± 10 193 ± 23 14.73 ± 0.83 0.051 ± 0.009
reagent blanks, n = 2 55 ± 6 205 ± 10 14.38 ± 0.08 0.46 ± 0.014
SPMD blanks, n = 3 61 ± 5 203 ± 17 13.95 ± 0.74 0.061 ± 0.004
SECe blanks, n = 3 72 ± 9 205 ± 9 15.41 ± 0.38 0.055 ± 0.002
Samples
IWWTP f SPMDs, n= 4 83 ± 29 227 ± 17 14.40 ± 0.46 0.034 ± 0.002
Nogales Wash SPMDs, n = 4 65 ± 7 227 ± 17 15.46 ± 1.04 0.042 ± 0.003
a
Picomol
b
Femtomol mg−1 protein.
c
µmol min−1 g−1 tissue.
d
Dimethyl sulfoxide.
e
Size exclusion chromatography.
f
International Waste Water Treatment Plant; Significantly different from blanks, p = 0.05.
TABLE 6.4 Brain and Gill βAR Binding in Rainbow Trout (Oncorhynchus mykiss)
Exposed to Sample Extracts and Controls. Reprinted from Petty et al. (2000),
copyright (2000); reproduced with permission from Elsevier
Brain Gill
Exposure medium KDa BMAX b KD BMAX
Controls
DMSOc , n = 3 235 ± 118 18 ± 9 445 ± 66 116 ± 5
reagent blanks, n = 3 1,047 ± 139 55 ± 4 943 ± 20 171 ± 22
SPMD blanks, n= 3 1,047 ± 139 82 ± 39 589 ± 284 116 ± 19
SECd blanks, n = 3 745 ± 348 42 ± 16 471 ± 16 140 ± 20
Samples
IWWTPe , n = 4 507 ± 294 30 ± 15 449 ± 24 122 ± 12
Nogales Wash f , n = 4 413 ± 270 26 ± 17 532 ± 59 176 ± 18
a
picomol.
b
Femtomol mg−1 protein.
c
Dimethyl sulfoxide.
d
Size exclusion chromatography.
e
International Waste Water Treatment Plant.
f
Significantly different from blanks, p = 0.05.
SPMD sample extracts, e.g., certain organochlorine pesticides (OCPs), are known
to inhibit cholinesterase activity. Therefore, these results were not unexpected.
However, it was surprising that a similar response was not observed with brain
cholinesterase activity. It is possible that brain cells can more readily metabolize
the chemicals, that the chemicals did not pass the brain blood barrier or that the
effects occurred earlier in the exposure period, effectively allowing the activity
to recover. Considering the numerous neurotoxic chemicals potentially entering
aquatic ecosystems or present as airborne vapor phase chemicals, the neurotoxic
mode of action related to exposure to contaminants is of increasing interest. Ev-
idence presented in this work demonstrate that SPMDs concentrate members of
this class of toxicants.
6.5. ENDOCRINE EFFECTS
Petty et al. (1998, 2000) used a vitellogenin (VGT) assay to assess the en-
docrine disrupting potential of contaminants in purified SPMD extracts. VGT is an
egg yolk phosphoprotein precursor that is synthesized in the liver of female teleosts
in response to estrogen from the ovary (Bailey, 1957). A wide variety of environ-
mental contaminants have been shown to have estrogenic activity (Colborn et al.,
1993). Equal portions of purified extracts from SPMDs, exposed in the Missouri
River after the flood of 1993 and from the IWWTP at the Nogales Wash deployment
were individually injected into immature rainbow trout (Oncorhynchus mykiss) as
described in Section 6.4. The SPMD extracts contained elevated levels of complex
mixtures of contaminants, including PAHs and pesticides. The fish injected with
these sample extracts exhibited VGT induction, while no induction was observed
in fish injected with any of the blank sample extracts.
Another approach for determining potential endocrine disrupting effects is
the YES assay (Routledge and Sumpter, 1996). In this assay, recombinant yeast
cells stably transfected with the gene for human estrogen receptor (hER) and
containing expression plasmids carrying strong promoter sequences and the lac-Z
(β-galactosidase) reporter gene are used to assess the estrogenic potential of chem-
icals or mixtures of chemicals. Using appropriate growth media, the yeast recom-
binant cells express the hER in a form capable of binding to estrogen response
elements (ERE) situated within a promoter sequence on the plasmid. Following the
binding of a suitable agonist, the agonist-hER complex interacts with various tran-
scription factors, binds to the ERE and initiates a cascade of events which results
in the expression of the lac-Z reporter gene and the secretion of β-galactosidase
into the assay medium. By incorporating a chromogenic substrate, chlorophenol-
red-β-D-galactopyranoside (CPRG), the estrogenic potential can be determined
photometrically at 540 nm following the conversion of the CPRG from yellow to
red by the β-galactosidase released into the growth medium in response to the
presence of hER in the sample.
132 Chapter 6
FIGURE 6.3 YES assay of HPLC fractions of extracts from SPMDs exposed to the Elizabeth
River, VA, USA. Reproduced courtesy of Andrew Rastall, University of Heidelberg, Heidelberg,
Germany.
Petty et al. (2000) exposed SPMDs to river water (Elizabeth River, VA, USA)
and then fractionated extracts of exposed SPMDs using high performance liquid
chromatography (HPLC; C18 column). Compounds in 55 fractions were separated
on the basis of their K ow s and the estrogenic activity of each fraction was assessed
using the YES assay. Figure 6.3 shows the results of this test. Fractions 21 (log
K ow range = 4.98 to 5.12), 26 (log K ow range = 5.83 to 5.97), 34 (log K ow
range = 6.96 to 7.10), and 51 (log K ow range = 9.22 to 9.36) had significant
estrogenic activity. This work illustrates the complexity of identifying the causal
agents resulting in environmental effects. It also suggests that classic toxicity
identification and evaluation procedures (i.e., chromatographic fractionation of
toxic extracts with subsequent assay of individual fractions) can be successfully
applied to this problem.
6.6. EXPOSURE OF LABORATORY ANIMALS

TO SPMD EXTRACTS
In addition to biomarker tests, exposure studies using selected laboratory test

animals have recently been conducted to obtain data related to the consequences
of the presence of mixtures of chemicals in habitats of concern. For example, am-
phibian deformities have been widely reported in North America (Souder, 2000),
and have resulted in numerous attempts to delineate the causes for the high defor-
mity rates observed (Ouellet et al., 1997). Often, it is very difficult to ascertain the
effects of contaminants on natural amphibian populations, due principally to the
variety of environmental variables that may act in concert with the contaminants
to cause detrimental effects. Because SPMDs are integrative samplers that accu-
mulate the readily bioavailable lipophilic waterborne-chemicals present in aquatic
systems, the chemical mixtures so obtained are time-weighted averages (TWA) of
exposure. Thus, the exposure of organisms to the complex mixture of hydrophobic
chemicals present at a site can be performed using the purified SPMD sample
extract and a physiologically neutral carrier medium.
To ascertain whether waterborne bioavailable contaminants were present in
ponds in North-Central Minnesota, Bridges et al. (2004) deployed 16 standard
SPMDs at each of two sites for 30 days. At one site, a high rate of amphibian
deformities had been documented (ranging from 60 to 75% in the mink frog [Rana
septentrionalis] and 4 to 20% in the northern leopard frog [Rana pipiens], Canfield
et al., 2000). The reference site had a stable amphibian population (Helgen, 1999)
exhibiting no unusual rates of deformities. Following dialysis and SEC cleanup as
described in Chapter 5, the extracts from the deployed SPMD samples within each
site were pooled into a single composite sample. Extracts from fabrication, field
and process blanks were similarly treated. The composite samples in high purity
hexane were solvent-exchanged into sterile DMSO so that each 1-mL extract
contained residues equivalent to about one day of exposure (i.e., represents the
amount of bioavailable residues sequestered by a standard SPMD in 24 hours).
Each of these extracts was added to 1 L of water. The test organisms were exposed
to UV radiation, (15.6 µW/cm2 UV-B, 914.9 µW/cm2 UVA; <1 µW/cm2 UV-B,
0.02 µW/cm2 UVA; respectively) in the presence and absence of SPMD samples
extract, to determine the potential effects of the interaction of contaminants and
UV radiation (see Bridges et al., 2004 for a detailed description of the exposure
system and statistical design of the experiment).
Following a 45 day exposure (at initiation of the exposure, tadpoles were free-
swimming but had not developed hind legs; test solutions were renewed every third
day, i.e., static renewal exposure) the R. pipiens tadpoles were shipped to the De-
partment of Developmental and Cell Biology at the University of California-Irvine
where they were held in culture until metamorphosis at which time deformities
were assessed (Gardiner and Hoppe, 1999). At metamorphosis the presence of
the following types of deformities were recorded: bony triangles, skin webbing,
misoriented limbs, and rigid limbs. Bony triangles were defined as the limb bone
being bent back on itself such that the proximal and distal ends of the bone were
adjacent to one another (Gardiner and Hoppe, 1999). Individuals with skin web-
bing displayed continuous bands of skin stretching across a joint, which restricted
motion of the limb (Meteyer, 2000). Individuals were classified as having multiple
deformities if more than one deformity type was noted on a single individual.
At metamorphosis there was equal mortality in both UV treatments, con-
sequently the UV treatments were pooled within SPMD treatments to increase
statistical power when analyzing the deformity data. When analyzed indepen-
dently of SPMD treatment, no significant deformities were found to occur due to
the effects of UV ( p > 0.05). SPMD extracts from the site exhibiting amphibian
134 Chapter 6
deformities produced significantly more bony triangles ( p = 0.025) and skin web-
bing ( p = 0.0183). The percentage of skin webbings in tadpoles exposed to ref-
erence site SPMD extracts and field blanks were not significantly different from
water controls. The effects of SPMD treatment on the number of frogs exhibit-
ing multiple deformities was marginally significant ( p = 0.052). There was no
difference among SPMD exposures in the number of misoriented limbs observed
( p = 0.65). Additional results are described in detail in Bridges et al. (2004).
Based upon these results, it is highly probable that waterborne lipophilic
chemicals contribute to the observed deformities in amphibian populations at the
impacted site. It is highly improbable that microorganisms or viruses originating
from the lake water could have caused the deformities because the transport cor-
ridors in the SPMD membrane are no more than 10 Å in cross-sectional diameter
(far too small to allow viruses to penetrate the sampler membrane).
Research studies employing exposures of organisms to SPMD extracts in a
physiologically neutral medium are increasingly being applied. We envision that
this approach will continue to find application in a wide variety of contaminant
assessment research studies.
6.7. OVERVIEW OF ADDITIONAL ASSAYS
As suggested earlier, a number of other assays have been successfully used

with SPMDs, including an immunoassay for PCBs and the Daphtoxkit F for in-
secticides. Zajicek et al. (1996) explored the use of a commercial ELISA kit from
Ohmicron Corporation (the Ohmicron PCB RaPID Assay) to analyze PCB residues
sequestered in SPMDs. He found a positive correlation (i.e., r 2 = 0.999, n = 3)
between the PCB concentrations in SPMDs measured by the ELISA and PCBs
measured by GC-ECD. ELISA kits are currently available for a number of types
of contaminants, determination of test results is rapid, and the kits are generally
inexpensive.
Sabaliūnas et al. (2000) showed that a Daphnia pulex immobilization test
(Daphtoxkit F) was far more sensitive than Microtox to a mixture of insecticides
sequestered in SPMDs. This is not surprising because the OCPs and pyrethroid
pesticides present in the enriched SPMD extracts are neurotoxins, and the effect
thresholds can be much lower than narcosis-type toxicants. Thus, if insecticides are
the contaminants of concern, the Daphtoxkit F approach may have some advantage
over Microtox.
6.8. POTENTIAL INTERFERENCES
The use of the above assays to evaluate the toxicity of SPMD extracts is
not without potential interferences. Sabaliūnas et al. (1999, 2000) examined the
potential role of oleic acid and elemental sulfur as contributors to the toxicity of
extracts from environmentally exposed SPMDs. The toxicity of fatty acids has
been attributed to their membrane disturbing properties, which include disruption
of the calcium pump by the formation of metal salts (Ewald and Sundin, 1993). As
mentioned earlier, oleic acid is an impurity in the triolein used in SPMDs, and may
also be produced by biotic or abiotic hydrolysis of methyl oleate and triolein (note
that this has not been definitively demonstrated to occur during SPMD exposures).
During environmental exposures, a significant portion of this triolein impurity
diffuses to the exterior surface of an SPMD, where dissipation and degradation
occur. Unfortunately, little or no attenuation occurs to the oleic acid levels in lab-
oratory SPMD-field blanks, -fabrication blanks and -process blanks (i.e., SPMDs
not environmentally exposed). Thus, the potential for a differential response exists
among field exposed SPMDs and associated QC SPMD samples. Elemental sulfur
is taken up by bacterial cells and may be reduced to toxic sulfides (Brouwer and
Murphy, 1995). Many types of sediment contain relatively large amounts of ele-
mental sulfur and elemental sulfur is readily accumulated by SPMDs. However, a
number of analytical methods can be used to remove these potential interferences
from SPMD extracts. These include SEC, as described in the Chapter 5 (both oleic
acid and sulfur), acid-treated copper wool (sulfur only, see Petty et al., 1995), and
KS (oleic acid). Note that other cleanup techniques are also available for these
interferences, especially for oleic acid.
Using a preemptive approach, Lebo et al. (2004) have shown that oleic acid
and methyl oleate can be removed from triolein prior to use of the triolein in SPMDs
(see Chapters 4 and 5). Dialysates from SPMDs prepared using triolein purified
by the Lebo et al. (2004) method exhibited lower acute toxicity (Microtox assay)
than SPMDs prepared with unpurified triolein. Also, the YES assay demonstrated
that the purification method had removed all background estrogenic activity from
SPMD extracts. For these reasons, the use of triolein purified by the method of
Lebo et al. (2004) is standard for all SPMD studies conducted at CERC, USGS.
Also, SPMDs with triolein purified by the Lebo et al. (2004) method are available
from the commercial vendor upon request.
Demonstrating the causal link between a specific chemical or mixture of
chemicals and the adverse effects observed in the ecosystem of concern is a
Sisyphean task. SPMDs offer an approach to sampling lipophilic chemicals in
aquatic and terrestrial ecosystems without having prior knowledge of their iden-
tities. Further, inclusion of other abiotic factors, e.g., UV radiation, provides an
opportunity to assess the impact of multiple stressors and increases the environ-
mental relevance of the research results. Following the determination of a link
between exposure to chemicals and observed effects, investigations to identify
the chemical or mixture of chemicals causing the effects can be designed and
conducted. In conclusion, combining the power of SPMDs to concentrate the
environmentally relevant bioavailable portion of complex mixtures of chemicals
present in aquatic and terrestrial environments with selected biomarker tests and
exposure studies appears to be a useful screening approach for determining the
potential consequences of exposure to bioconcentratable contaminants.
136 Chapter 6
6.9. REFERENCES
Ankley, G.T.; Tillitt, D.E.; Giesy, J.P.; Jones, P.D.; Verbrugge, D.A. 1991, Bioassay-derived 2,3,7,8-
tetrahlorodibenzo- p-dioxin equivalents in the flesh and eggs of Lake Michigan Chinook Salmon
and possible implications for reproduction. Can. J. Fish. Aquat. Sci. 48: 1685–1690.
Bailey, R.E. 1957, The effect of estradiol on serum calcium, phosphorus, and protein of goldfish. J.
Exp. Zool. 136: 455–469.
Beauvais, S.J. 1997, Factors Affecting Cholinesterase in Aquatic Animals. Ph.D. Thesis, Iowa State
University, Ames Iowa.
Bridges, C.M.; Little, E.E.; Gardiner, D.M.; Petty, J.D.; Huckins, J.N. 2004, Assessing the toxicity of
teratogenicity of pond water in North-Central Minnesota to amphibians. Environ. Sci. Pollut. R.
11: 233–239.
Brouwer, H. and Murphy, T. 1995, Volatile sulfides and their toxicity in freshwater sediments. Environ.
Toxicol. Chem., 14: 203–208.
Canfield, J.T., Kerstern, S.M., Vanselow, P. 2000, 1997–1999 Field Season Report. Unpublished report
to Minnesota Pollution Control Agency, Minneapolis, MN.
Cleveland, L.; Little, E.E.; Petty, J.D.; Johnson, B.T.; Lebo, J.A.; Orazio, C.E.; Dionne. J.; Crockett,
A. 1997, Toxicological and chemical screening of Antarctica sediments: Use of whole sediment
toxicity tests, Microtox, Mutatox, and semipermeable membrane devices (SPMDs). Mar. Pollut.
Bull. 34: 194–202.
Colborn, T.; vom Saal, F.S.; Soto, A.M. 1993, Developmental effects of endocrine-disrupting chemicals
in wildlife and humans. Environ. Health Persp. 101: 533–553.
Connell, D.W. 1990, Bioaccumulation of Xenobiotic Compounds. CRC Press: Boca Raton, FL. p 219.
Ewald, G. and Sundin, P. 1993, ATP Leakage from ELD cells after exposure to steric,
monochlorostearic, dichlorostearic, and oleic acids. Pharmacol. Toxicol. 73: 159–162.
Gale, R.W.; Long, E.R.; Schwartz, T.R.; Tillitt, D.E. 2000, Evaluation of planar halogenated and poly-
cyclic aromatic hydrocarbons in estuarine sediments using ethoxyresorufin-o-deethylase induction
of H4IIE cells. Environ. Toxicol. Chem. 19: 1348–1359.
Gard, N.W. and Hooper, M.J. 1993, Age dependent changes in plasma and brain cholinesterase activities
of Eastern Blue-birds and European Starlings. J. Wildlife Dis. 29: 1–7.
Gardiner, D.M. and Hoppe, D.M. 1999, Environmentally induced limb malformations in mink frogs
(Rana septentrionalis). J. Exp. Zool. 284: 207–216.
Helgen, J.C. 1999, Minnesota Pollution Control Agency, Minneapolis, M.N. Personal communication.
Huckins J.N.; Prest H.F.; Petty J.D.; Røe T.I.; Meadows J.C.; Echols K.R.; Lebo J.A.; Clark R.C.
1998, A Overview of the Results of Several Comparisons of Lipid-containing Semipermeable
Membrane Devices (SPMDs) and Biomonitoring Organisms for Assessing Organic Chemical
Exposure. Abstracts of the 19th Annual National Meeting; SETAC; Charlotte, NC; November
15–19, 1998; p 215.
Johnson, B.T. 1998, Microtox R
Toxicity Test System—New Developments and Applications. In Mi-
croscale Testing in Aquatic Toxicology; Wells, P.G., Lee, K., Blaise, C., Eds; CRC Press: Wash-
ington, D.C.; pp. 201–218.
Johnson, B.T.; Petty, J.D.; Huckins, J.N.; Lee, K. 2004, Hazard assessment of simulated oil spill on
intertidal areas of the St. Lawrence River with SPMD-TOX. Environ. Toxicol. 19: 329–335.
Jones, S.B. and King, L.B. 1995, Muscarinic cholinergic receptors in brain and atrial membranes
of adult Brook Trout Salvelinus fontinalis measured by radiolignad binding techniques. Comp.
Biochem. Physiol. 112: 43–50.
Koc̆ı́, V.; Mlejnek, M.; Kochánková, L. 2003, Toxicological evaluation of exposed SPMD membranes.
Cent. Eur. J. Chem. 1: 28–34.
Lorenzen, A. and Kennedy, S.W. 1993, A fluorescence-based protein assay for use with a microplate
reader. Anal. Biochem. 214: 346–348.
FL; Vol. V.
Meteyer, C.U. 2000, Field Guide to Malformations of Frogs and Toads with Radiographic Interpre-
tations. USGS National Wildlife Health Center, Madison, WI; Biological Science Report No.
USGS/BRD/BSR-2000-0005.
Microbics Corporation 1992, Condensed Protocol for Basic Test, Using Organic Solvent
Sample Solubilization. Microtox Manual, Vol. III; Condensed Protocols; Carlsbad, CA; 13,
226–232.
Ouellet, M.; Bonin, J.; Rodrigue, J.; DesGranges, J.L.; Lair, S. 1997, Hindlimb deformities (ectromelia,
ectrodactyly) in free-living anurans from agricultural habitats. J. Wildlife Dis. 33: 95–104.
Parrott, J.L. and Tillitt, D.E. 1997, The use of Semipermeable Membrane Devices (SPMDs) to Concen-
trate Inducers of Fish Hepatic Mixed Function Oxygenase (MFO). In Ecotoxicology: Responses,
Biomarkers, and Risk Assessment. An OECD Workshop; Zelikoff, J.T., Ed.; SOS Publications:
Fair Haven, NJ.; pp. 185–196.
Parrott, J.L.; Backus, S.M.; Borgmann, A.I.; Swyripa, M. 1999, The use of semipermeable mem-
brane devices to concentrate chemicals in oil refinery effluent on the Mackenzie River. Arctic 52:
125–138.
1998, Determination of bioavailable contaminants in the Lower Missouri River following the
Petty, J.D.; Jones, S.B.; Huckins, J.N.; Cranor, W.L.; Parris , J.T.; McTague, T.B.; Boyle, T.P. 2000,
An approach for assessment of water quality using semipermeable membrane devices (SPMDs)
and bioindicator tests. Chemosphere 41: 311–321.
138 Chapter 6
contaminants. Chemosphere 54: 695–705.
Prest, H.F.; Hodgins, M.M.; Petty, J.D.; Huckins, J.N.; Lebo, J.A.; Orazio, C.E.; Gibson, V. 1997,
Laboratory Comparisons of Oysters and Semipermeable Membrane Devices (SPMDs). USGS
Midwest Science Center, Columbia, MO; Report to the American Petroleum Institute (API):
Washington; DC; API Report No. DR 360; p. 105.
Rastall, A.C.; Neziri, A.; Vukovic, Z.; Jung, C.; Mijovic, S.; Hollert, H.; Nikcevic, S.; Erdinger, L. 2004,
The identification of readily bioavailable pollutants in Lake Shkodra/Skadar using semipermeable
membrane devices (SPMDs), bioassays and chemical analysis. Environ. Sci. Pollut. R. 11: 240–
253.
Routledge, E.J. and Sumpter, J.P. 1996, Estrogenic activity of surfactants and some of their degradation
products assessed using a recombinant yeast screen. Environ. Toxicol. Chem. 15: 214–248.
Sabaliūnas, D.; Lazutka, J.; Sabaliūniene, I.; Sodergren, A. 1998, Use of semipermeable membrane de-
vices for studying effects of organic pollutants: Comparison of pesticide uptake by semipermeable
membrane devices and mussels. Environ. Toxicol. Chem. 17: 1815–1824.
Sabaliūnas, D.; Ellington, J.; Sabaliūniene, I. 1999, Screening bioavailable hydrophobic toxicants in
surface waters with semipermeable membrane devices: Role of inherent oleic acid in toxicity
evaluations. Ecotox. Environ. Safe. 44: 160–167.
Sabaliūnas, D.; Lazutka, J.R.; Sabaliūniene, I. 2000, Acute toxicity and genotoxicity of aquatic hy-
drophobic pollutants sampled with semipermeable membrane devices. Environ. Pollut. 109: 251–
265. (was 2000b)
Souder, W. 2000, A Plague of Frogs: The Horrifying True Story. Hyperion Press: New York.
Steevens, J.A.; Bauman, P.C.; Jones, S.B. 1996, A comparison of β-adrenoceptors and muscarinic
cholinergic receptors in tissues of brown bullhead catfish Ameiurus nebulosus from the Black
River and Old Woman Creek Ohio. Environ. Toxciol. Chem. 15: 1551–1554.
Tillitt, D.E.; Giesy, J.P.; Ankley, G.T. 1991, Characterization of the H4IIE rat hepatoma cell bioassay as
a tool for assessing toxic potency of planar halogenated hydrocarbons in environmental samples.
Whyte, J.J.; Jung, R.E.; Schmitt, C.J.; Tillitt, D.E. 2000, Ethoxyresorufin-O-deethylase (EROD) activity
in fish as a biomarker of chemical exposure. Crit. Rev. Toxicol. 30: 347–570.
Whyte, J.J.; Schmitt, C.J.; Tillitt, D.E. 2004, The H4IIE cell bioassay as an indicator of dioxin-like
chemicals in wildlife and the environment. Crit. Rev. Toxicol. 34: 1–83.
Zajicek, J.L.; Tillitt, D.E.; Huckins, J.N.; Petty, J.D.; Potts, M.E.; Nardone, D.A. 1996, Application of
Enzyme-Linked Immunosorbent Assay (ELISA) for Measurement of Polychlorinated Biphenyls
(PCBs) from Hydrophobic Solutions: Extracts of Fish and Dialysates of Semipermeable Mem-
brane Devices (SPMDs). In Environmental Immunochemical Methods; ACS Symposium Series
646; American Chemical Society: Washington, D.C.; Chapter 26, pp 307–325.
Chapter 7
COMPARISONS TO
BIOMONITORING
ORGANISMS
7.1. BACKGROUND
Numerous side-by-side comparisons of the accumulation of hydrophobic or-

ganic chemicals (HOCs) by lipid-containing semipermeable membrane devices
(SPMDs) and a variety of biomonitoring organisms (BMOs) have been conducted
(Prest et al., 1992, 1995a, 1995b, 1997; Ellis et al., 1995; Herve et al., 1995;
Huckins et al., 1996, 1998a, 2004; Kolok et al., 1996; Peven et al., 1996; Gale
et al., 1997; Hofelt and Shea, 1997; Moring and Rose, 1997; Meadows et al.,
1998; Sabaliūnas et al., 1998; Wang et al., 1998; Axelman et al., 1999; Utvik et al.,
1999; Utvik and Johnsen, 1999; Echols et al., 2000; Granmo et al., 2000; Leppanen
and Kukkonen, 2000; Baussant et al., 2001; Booij et al., 2002; Følsvik et al., 2002;
Gatermann et al., 2002; Lu et al. 2002; Wang et al., 2002; Richardson et al., 2003;
Verweij et al., 2004). The primary goal of these comparisons was to determine the
commonality in the types and concentrations of HOCs accumulated by the two
monitoring matrices. Both similarities and differences have been observed in the
accumulation patterns of SPMDs and BMOs. The inconsistencies of these findings
are not surprising in view of the differences between inanimate SPMDs and living
organisms. Species of BMOs used for field assessments vary depending on the
environmental characteristics of deployment sites and project goals. Significant
139
140 Chapter 7
physiological, anatomical, and behavioral differences exist across BMO species

that may affect the accumulation of HOCs. In contrast, the accumulation of HOCs
by SPMDs is solely mediated by passive diffusional and partitioning processes,
which are the basis of equilibrium partition (EP) theory. Thus, good correlations
between analyte concentrations in side-by-side exposures of BMOs and SPMDs
(Hofelt and Shea, 1997; Meadows et al., 1998; Sabaliūnas et al., 1998; Utvik and
Johnsen, 1999; Wang et al., 2002) suggest that passive partitioning and diffusional
processes ultimately control the residue accumulation patterns in the BMOs, while
poor correlations (e.g., Gale et al., 1997; Moring and Rose, 1997) suggest that
more complex biological processes largely control residue accumulation patterns
in BMOs (Huckins et al., 2004).
For a standard SPMD (see Section 4.4.) of set size, the variables poten-
tially affecting HOC accumulation are limited to physicochemical properties of
the analyte, exposure site conditions, and exposure scenario factors such as the
constancy of chemical concentrations during the exposure period (Huckins et al.,
1993, 2002a, 2002b; Booij et al., 1998). A number of SPMD calibration studies
(see Appendix A) have been conducted to determine the effects of analyte physico-
chemical properties and site conditions such as temperature and flow dynamics
on SPMD accumulation rates and equilibrium partition coefficients. Beyond this,
PRCs (i.e., performance reference compounds) can be added to SPMD lipid prior
to exposures to account for the effects of site-specific conditions on SPMD sam-
pling (Huckins et al., 2002a; also see Chapters 1 and 3 for more information).
This ability to generate chemical-specific calibration data and then adjust these
values to site-specific conditions means that analyte concentrations obtained using
SPMDs are directly comparable across sample sites. Thus, SPMDs are well suited
for reliable determinations of HOC sources, concentration gradients and concen-
tration estimates, but their utility for the estimation of bioaccumulation factors
(BAFs; see Chapter 1 for definition) in species of concern is not as well defined.
The physicochemical properties of chemicals, site-specific exposure condi-
tions, and the exposure scenario, also affect the bioconcentration or bioaccumu-
lation of HOCs (Mackay et al., 1992a, 1992b; Goudreau et al., 1993; Baumard
et al., 1998a, 1998b). As suggested earlier, certain behavioral, physiological, and
anatomical characteristics of BMO species affect bioaccumulation (Huebner and
Pynnönen, 1992; Gilek et al., 1996; Björk and Gilek, 1997; Moring and Rose,
1997; Baumard et al., 1998a; Axelman et al., 1999; Baussant et al., 2001). For
example, Björk and Gilek (1997) found that laboratory BAFs of selected poly-
chlorinated biphenyl (PCB) congeners in blue mussels (Mytilus edulis) varied up
to two orders of magnitude due to differences in food ration. Surprisingly, the
largest BAF was observed at a relatively low food ration of about 0.08 mg L−1
particulate organic carbon (POC) or 1900 cells mL−1 , and higher rations resulted
in an exponential decline in BAFs. In regard to finfish, Jimenez et al. (1987) found
that benzo[a]pyrene BAFs in bluegill sunfish (Lepomis macrochirus) were at least
five times higher in unfed fish than in fed fish. Gilek et al. (1996) showed that
equilibrium BAFs of PCB congeners fell significantly with increases in the size
COMPARISONS TO BIOMONITORING ORGANISMS 141
and mass of blue mussels. In the presence of toxic chemicals, juvenile and adult
bivalve mollusks can close their valves for extended periods (Goudreau et al.,
1993; ASTM, 1996), which greatly impacts bioaccumulation. Also, xenobiotic
metabolism can greatly affect steady-state concentrations of some HOCs in BMO
tissues. Some variables known to affect bioaccumulation are difficult to quantify a
priori, such as intra- and interspecies variability, condition factors, and the role of
developmental stages (Franke et al., 1994). Finally, the potential impacts of many
of the aforementioned variables and others on equilibrium bioconcentration factors
(BCFs; see definition in Chapter 1) and BAFs are discussed in three ASTM Stan-
dard Guides (ASTM, 1984, 1996, 2001), and a review of data from Mackay et al.
(1992a, 1992b, 1997) shows orders of magnitude variations in BMO BCFs, and
uptake (ku s) and depuration (ke s) rate constants for the same chemical (Huckins
et al., 2004).
To reduce the effects of differences in uptake kinetics and lipid contents on
residue concentrations in BMOs, exposures are typically designed to attain equilib-
rium concentrations, and the concentrations are lipid normalized. When using the
EP approach to assess exposure concentrations, sources, concentration gradients,
etc., investigators assume that BAFs are independent of exposure concentration
while concentrations in tissues or lipids are proportional to exposure concentra-
tions, and that lipid normalization reduces the variability of whole body tissue
concentrations. However, these assumptions are not always borne out in studies
reported in the peer-reviewed literature. For example, Axelman et al. (1999) found
that BAFs of individual polycyclic aromatic hydrocarbons (PAHs) in blue mus-
sels deployed in marine water near a smelter were about one order of magnitude
greater than the equilibrium BAFs of simultaneously deployed blue mussels at a
nearby marine reference site that had similar levels of PAHs. Furthermore, the
reference-site BAFs were about one order of magnitude greater than the BAFs of
blue mussels reported in the literature. Also, there is debate in the literature about
the appropriateness of normalizing tissue concentrations to extractable lipid con-
tents (e.g., Schneider, 1982; Randall et al., 1991; Hebert and Keenleyside, 1995;
Stow, 1995; Gray, 2002). As discussed in Section 5.6.2., lipid-normalization of
SPMD concentrations is inappropriate in most cases.
Regardless of the marked differences between SPMDs and BMOs, there are
some fundamental similarities in the characteristics and processes affecting the
accumulation of HOCs in the two matrices. For example, diffusion of non-polar
compounds through nonporous organic polymers such as used in the SPMD has
been shown to be similar to diffusion across biomembranes (Lieb and Stein, 1969).
Furthermore, the processes of solute diffusion across the water boundary layer
(WBL) and the lipid-like or lipid-containing membranes of SPMDs and aquatic
organisms, and the partitioning between the lipids and the exposure water, are
important factors in the accumulation of HOC residues in both matrices. The
triolein used in SPMDs is a major lipid in fishes (Huckins et al., 1990) and is
representative of fats or the neutral lipid class (Chiou, 1985), which is the largest
storage site of persistent HOCs in many organisms.
142 Chapter 7
7.2. IMPLICATIONS OF SELECTED MODELS USED FOR SPMDs

AND BMOs
Herein, we examine several equations which are used to determine BCFs and
BAFs, SPMD-water partition coefficients (K sw s), and the more complex exchange
processes of some BMOs. Hopefully, the formulation and the assumptions behind
these equations further clarify some of commonalities and differences between
residue accumulation in BMOs and SPMDs.
Phillips (1980) and Phillips and Rainbow (1993) have stated that each species
of aquatic BMO exhibits unique uptake and elimination kinetics for a particular
HOC. The ramification of this statement is revealed in the following simple one-
compartment model, which is often used for the determination of steady-state
BCFs and K sw s.
BC F = ku /ke (7.1)
where BCFs are generally based on whole body residue concentrations, ku repre-
sents the linear portion of the uptake curve and ke represents the first-order depu-
ration curve. The following equation provides a more detailed formulation of ke :
ke = ko A/(Mwb BC F) (7.2)
where ko is the overall mass transfer coefficient, A is the surface area of the
membrane where solute or vapor exchange occurs, Mwb is the whole body tissue
mass and in this case BCF has units of mL g−1 . The group ko A can be viewed
as the apparent water sampling rate (Rs ), with units of L d−1 . When the sorption
capacity of non-lipid phases can be neglected, the BCF can be written in terms of
the lipid-water partition coefficient (K Lw , in mL mL−1 ) and the lipid volume (VL ).
BC F ≈ K Lw VL /Mwb (7.3)
Inspection of Eqs. 7.2 and 7.3 shows that ke is inversely proportional to
the lipid content of the biota (VL /Mwb ), whereas ku is independent of the lipid
content. Thus, the generally much lower neutral lipid contents of BMOs result
in much higher values of BMO ke s. Also, the BMO BCFs are generally much
smaller than the SPMD K sw s.
Equation 7.1 utilizes exchange coefficients to predict steady-state BCFs and
K sw s, and the model assumptions include a uniform lipid phase enclosed in a non-
interactive membrane. The model shows that the magnitude of a BMO’s BCF or an
SPMD’s K sw is affected by variations in ku and/or ke , unless both constants rise or
fall proportionally. In the case of SPMDs, Huckins et al. (1993, 2002a) have shown
that the uptake and release process is essentially isotropic for HOCs. When residue
exchange is isotropic, K sw s will remain relatively constant even when exposure
conditions affect SPMD ku and ke values. This is not always the case for BMOs,
yet isotropic exchange is a fundamental assumption of EP theory.
Metabolism of PAHs by the cytochrome P-450-dependent enzymes in fish
is a classic deviation from EP theory. When PAH biotransformation product data
are available (e.g., concentrations of PAH conjugates or other metabolites in bile),
two depuration rate constants can be used as follows

BC F = ku /(ke + kem ) (7.4)
where kem is the first-order rate constant for HOC metabolism and egestion of
metabolites via the bile, feces, and urine, and the rate constant ke represents outward
diffusion of HOCs across the gills or skin. Deviations of loss rates from first-order
kinetics invalidate the model for the determination of BCFs and result in anisotropic
exchange kinetics. Even when the overall depuration of PAHs by fish follows first
order kinetics, the resulting tissue concentrations are not always proportional to
ambient environmental concentrations. A study reported by Gale et al. (1997)
supports the contention that tissue residues are often not proportional to ambient
water concentrations (see Chapter 8).
In cases where the depuration of HOCs from BMOs involves enzyme-
mediated biotransformations (Eq. 7.4) or active transport mechanisms, and envi-
ronmental concentrations are high (e.g. near a point source), depuration rates have
been shown to follow Michaelis-Menten kinetics (Spacie and Hamelink, 1985).
Michaelis-Menten kinetics is elicited when an enzyme or active transport system
is saturated with a chemical. This type of kinetics is characterized by lower values
of ke s at sites with high HOC concentrations. If ku s are unchanged at high con-
centration sites, Michaelis-Menten kinetics will result in elevated BAFs. However,
if chemical concentrations become toxic, finfish likely avoid the area and sessile
organisms such as mussels may close their valves for extended periods (Huckins
et al., 2004).
A number of models have been developed that include physiologically con-
trolled bioaccumulation parameters (Bruggeman et al., 1981; Rand and Petrocelli,
1985; Gobas et al., 1993; Björk and Gilek, 1997). Following the work of Brugge-
mann et al. (1981), the relative roles of respiratory and dietary uptake, and residue
metabolism can be examined for BMOs using
dCB /dt = S FCf + (E x Rv /Mwb )Cw − ke CB − kem CB (7.5)
where CB is the HOC concentration in whole body BMO tissue, S is the fractional
assimilation or absorption efficiency of HOCs across the gut, F is the feeding or
ingestion rate per unit mass of organism (i.e., g−1 d−1 ), Cf is the HOC concentration
in the food, E x is the fractional extraction efficiency of HOCs from water ventilated
through the gills, Rv is the ventilation or filtration rate (bivalves or other filter
feeders) of water through the respiratory lamellae (gills) in L d−1 , Mwb is the mass
of whole body tissues, and Cw is given in units of g L−1 . A major assumption
underlying Eq. 7.5 is that uptake processes and removal processes are additive.
The group ke CB can be changed to (ke + G p )CB , where G is the relative growth rate
(g g−1 d−1 ), which can be positive or negative (note: changes in lipid content must
be considered as well). Both gill-extraction efficiencies (E x ) and gut-assimilation
efficiencies (S) vary considerably (e.g., E x ranges from about 7 to 60% (McKim
et al., 1985) and S ranges from about 1 to 90% (Wang and Fisher, 1999; Thomann,
et al., 1992), depending on HOC K ow and molecular structure (Huckins et al.,
1998b; Bucheli and Gustafsson, 2000), type of organism (Gobas et al., 1993; Wang
144 Chapter 7
and Fisher, 1999), and the quality of ingested material (Wang and Fisher, 1999;
Bucheli and Gustafsson, 2000). Also, behavioral factors such as the avoidance of
toxics and differential flow of blood to tissues are not modeled even though they
may affect tissue concentration. Equations such as 7.5 are not commonly used in
field studies because data to determine the associated rate constants are seldom
available. Furthermore, the expansion of models to include additional processes
must always be justified in light of the potential increase in error propagation.
Factors other than organism physiology contribute to the variability of BAFs
and related parameters (i.e., ku s and ke s) reported in the literature. Differences
in analytical chemistry conventions used for tissue weighting can be a source of
errors and may be problematic when comparing SPMD and BMO concentrations
(see Chapter 5). Briefly, tissue weighting approaches include wet weights, dry
weights and lipid weighting or normalization. To determine BAFs, ku s and ke s,
wet tissue concentrations are generally used for fish, while dry tissue weights
are generally used for bivalve concentrations. Unlike BMOs, SPMD weighting
is standardized to the weight of a whole SPMD. Lipid normalized data are gen-
erally derived from dividing tissue concentrations by total lipid concentrations.
Hebert and Keenleyside (1995) have pointed out that lipid normalization can re-
duce precision and the power of statistical tests to detect differences in the data.
The lipid normalization approach is based on the assumption that a significant re-
lationship exist between contaminant concentration in tissues and their total lipid
concentration. However, Schneider (1982) found poor correlations between HOC
concentrations in a marine fish and the total extractable lipid concentration of the
tissues. Only when using the neutral triglyceride fraction (i.e., storage fats such as
triolein) of the total lipid extracts did correlations improve to an acceptable level.
Unfortunately, many bivalves and invertebrates do not have significant levels of
triglycerides.
Using PCB levels in five species of freshwater finfish, collected over a course
of 20 years, Stow (1995) failed to find a significant relationship between residue
concentrations and percent lipid. The finding of Randall et al. (1991) may explain
part of the problem. They found that using different extraction solvents for tissues,
lipid concentrations can vary by 3.5 fold and that laboratories vary widely in the
type of solvents used for the extraction of HOC residues in tissues. Whole body
lipid levels across BMO species typically vary from about 1 to 15% (based on wet
tissue weights). Thus, the lipid mediated differences in BMO tissue concentrations
may be as high as 15 fold. Unlike BMOs, Standard SPMDs have a uniform lipid
content, which precludes any need for lipid normalization, and the extraction or
dialysis solvent is standardized.
7.3. COMPARISON OF SPMDs AND BMOs
The series of steps involved in the accumulation of HOCs in BMOs is more

complex than those associated with residue accumulation in SPMDs. For example,
FIGURE 7.1 Comparison of the patterns of organic contaminant uptake rates (as related to
log Kow s) by SPMDs and across fish gills (McKim et al., 1985). Reprinted with permission
from the American Petroleum Institute (Huckins et al., 2002).
the rate of residue transport from the gill epithelium via the blood to lipid storage
compartments is complicated by differences in blood flows to various tissue groups
(Barron, 1990). Thus, comparisons of the relative flux of chemicals across the bar-
riers associated with the water-blood interface (i.e., gills or skin) of organisms and
across the barriers to SPMD uptake should provide the best correlations between
BMOs and SPMDs, because of the lack of confounding factors such as differential
blood flow to tissues, in vivo biotransformation of residues, and marked variations
in the amounts and types of lipids. Using this approach, Figure 7.1 shows that
a plot of organic chemical uptake efficiencies across trout gills (McKim et al.,
1985), relative to their log K ow s, has the same parabolic shape as a plot of SPMD
146 Chapter 7
Uptake Rate (k u)
Water Solubility Limitations

Membrane Boundary Sorptive Attenuationn
Controll Layer controll
2 3 4 5 6 7 8 9
Log Kow
FIGURE 7.2 Plot of rate constants for the uptake of HOCs by SPMDs and fish, relative to com-
pound hydrophobicity. Also, potential rate-limiting steps/factors are illustrated as related to
compound hydrophobicity. Low to moderate flow and turbulence were assumed.
sampling rates of priority pollutant PAHs versus log K ow s. This illustration sug-
gests that similar rate limiting steps govern the uptake of organic chemicals across
the water-blood barriers of fish and across the barriers to solute flux into SPMDs.
The parabolic shape of the curves in Figure 7.1 can be partly explained by the
following discussion on mass-transfer phenomena (also, see Figure 7.2 and the re-
lated discussions in Chapter 3). For chemicals with log K ow s < 4, the rate limiting
step in chemical uptake appears to be permeation across the fish respiratory lamel-
lae (gills) and the SPMD membrane (McKim et al., 1985; Huckins et al., 1999
and 2002b). In this case, uptake rate constants and extraction efficiencies rise with
increasing membrane-water partition coefficients (K mw s) of chemicals. The region
of uptake rate constants and extraction efficiencies represented by compounds with
log K ow s ≥ 4 but ≤ 5 is characteristic of the transition from membrane rate control
to external WBL rate control. During this transition, resistance in the membrane
decreases (due to rising values of K mw ; see Eqs. 3.8 and 3.9) to the point where
it is similar in magnitude to the resistance of the WBL. Uptake rate constants
and extraction efficiencies are often highest for chemicals with log K ow s > 5 but
≤6, where membrane resistance becomes insignificant. Finally, when log K ow s
of HOCs increase beyond 6, uptake rate constants and extraction efficiencies are
expected to slowly fall due to the decreasing molecular diffusion coefficients of
compounds with increasingly large molar volumes (see solid line in Figure 7.2).
However, plots of uptake rate constants and extraction efficiencies of SPMDs and
BMOs often show greater declines in the sampling of high K ow compounds than
can be accounted for by molecular size related decreases in diffusion coefficients
(Figure 7.1 and dashed line in Figure 7.2). This phenomenon may be due to a com-
bination of factors described in Chapter 3 such as attenuation of very hydrophobic
solute concentrations by sorption to particulate and dissolved organic carbon (i.e.,
an artifactual overestimation of freely dissolved chemicals), solubility limitations,
a switch back to membrane control due to steric impedance (Barron, 1990), ventila-
tion volume limitations of very hydrophobic chemicals, and the reduced solubility
of high molecular weight HOCs in high molecular weight storage fats or neutral
triglycerides (Chiou, 1985; Schüürman and Klein, 1988). Much of this discussion
as well as Figure 7.2 is based on the assumption that the flow and turbulence is
low to moderate. When flow and turbulence are high, the switching point between
membrane control and WBL control (see vertical dotted line in Figure 7.2) would
be shifted to higher K ow s in the case of SPMDs. The effects of flow dynamics on
turbulence around BMO gill structures are unknown.
Table 7.1 provides a comparison of selected physical characteristics of a stan-
dard SPMD membrane and the gills of fish. The surface area per unit mass of SPMD
is at least 9-fold larger than the surface area of fish gills per unit mass of organism.
Earlier, Eq. 3.15 showed the direct relationship between surface area of the mem-
brane and uptake rates. The SPMD membrane is at least 8 times thicker than the gill
integument of fishes. Equation 3.50 showed the inverse relationship of membrane
or barrier thickness (δ) in the mass transfer of solutes. Table 7.1 also shows that
the estimated diameters of transient cavities in the low density polyethylene mem-
brane used for SPMDs ranges up to about 10 Å (Hwang and Kammermeyer, 1984),
and larger at elevated temperatures. Opperhuizen et al. (1985) postulated that the
size of transient cavities in lipoidal regions of biomembranes is ≤9.8 Å based
on the lack of octachloronaphthalene (log K ow = 8.4, molecular cross-sectional
diameter = 9.8 Å) accumulation in fish. More recently, Booij et al. (2002) found
that the polybromonated diphenyl ether (PBDE) 209 (decabromodiphenyl ether)
TABLE 7.1 Comparison of Selected Properties of Standard SPMDs and Fish Gills
Standard SPMD Fisha
Membrane:
composition low density polyethylene complex lipoprotein bilayer
molecular size cutoff ≈10 Å ≈9.5 Åb
surface area ≈100 cm2 g−1 1 – 9 cm2 g−1 tissue
thickness ≈86 µm 0.5 – 11 µm
blood-water barrier
Exchange kinetics:
rate control membrane if log K ow < 4.5 membrane if log K ow < 3
diffusion layer if log K ow ≥ 4.5 diffusion layer if log K ow ≥ 3
a
Data obtained from Hayton and Barron (1990) and Gobas et al. (1986).
b
Hypothesized molecular size cutoff by Opperhuizen et al. (1985).
148 Chapter 7
was not accumulated by blue mussels but was accumulated by SPMDs. The ef-
fective molecular cross-sectional diameter of all PBDEs with 2,5-Br substitution
on at least one ring is 9.6 Å. These results indicate that steric impedance is not a
likely explanation for the lack of PBDE 209 accumulation in mussels. Residues
of PBDE 209 were present in the gut of mussels but after a 24 hour depuration
period, these residues were essentially eliminated, indicating no assimilation and
tissue incorporation.
Not shown in Table 7.1 is the effective thickness of the WBLs (see Figure 3.1),
which varies with flow velocity and turbulence and often is the rate limiting step in
solute mass transfer. Under quiescent or stagnant conditions, the effective thickness
of a WBL can be as much as 1 mm and as much as several mm for an air boundary
layer (ABL). Under highly turbulent conditions, the effective thickness of the WBL
and the ABL can be thinned to only a few µm. Note that compound hydrophobicity
plays an important role in which barrier (e.g., membrane or water) has the greatest
resistance to mass transfer as shown by Eq. 3.9. Thus, Table 7.1 gives an estimate
of the log K ow values, where the rate-limiting step in mass transfer changes from
membrane to WBL control.
Except under turbulent exposure conditions, the effective thickness and re-
lated resistance to mass transfer of the SPMD WBL is expected to be greater than
that associated with the gills of an aquatic organism. This is due to the active pump-
ing action or ventilation required for respiration, which can be a little more than 1
L d−1 g−1 for fish under demanding conditions. When bivalves are filter feeding,
ventilation rates may be much higher and can be as high as 20 L d−1 g−1 (e.g.,
zebra mussel Dreissena polymorpha). A review of Eq. 7.5 shows the importance
of ventilation rate (Rv ), in combination with extraction efficiency, in the clearance
or uptake rates of HOCs. Figure 7.1 indicates that gill extraction efficiencies are as
high as about 60%. Unfortunately, there is little data on the effects of ventilation
volume or flow velocity through the gills on solute extraction efficiencies. Because
high flow and turbulence thins the WBL and reduces resistance to mass transfer,
extraction efficiencies would be expected to increase. However, high flows also
increase the probability of channeling (i.e., insufficient time for solute molecules
in the center of interlamellar channels to make random contact with membrane sur-
faces) through interlamellar spaces, reducing extraction efficiencies. Lower rates
of oxygen uptake have been shown for high flow rates (i.e., perfusion) through the
gills (Bayne et al., 1976) of rainbow trout (Oncorhynchus mykiss).
7.3.1. Similarity of Uptake Rate Constants

Meadows et al. (1998) conducted a 28 d exposure of brown trout (Salmo
trutta), standard SPMDs and hexane filled dialysis bags (Södergren, 1987) to spring
water (total organic carbon <1 mg L−1 ) contaminated with PCBs. Trout were not
fed during the exposure, and temperature and flow conditions remained constant
throughout the exposure. A good correlation (r 2 = 0.89) was found between the
uptake rate constants (ku,F s) for whole body trout and the uptake rate constants
10
9
8
7
k u,F (L g--1 d )
--1
6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 10
ku,S (L g--1·d--1)
FIGURE 7.3 Comparison of brown trout (Salmo trutta) and whole SPMD uptake rate constants
(ku,F s and ku,S s, respectively) for PCB congeners. Reprinted from Meadows et al. (1998), copy-
right (1998); reproduced with permission from American Chemical Society.
(ku,s s) for whole SPMDs (Figure 7.3). Congener uptake rate constants for SPMDs
averaged about 2-fold higher than those for trout over a 500-fold range in K ow s.
The uptake rates of congeners by both trout and SPMDs slowly declined with
increasing K ow s. This is consistent with WBL controlled uptake, where diffusion
coefficients in the WBL slowly decline with increasing molar volume or molecular
weight of solute or vapors. The least hydrophobic congener accumulated by fish
and SPMDs had a log K ow of 5.06, which is greater than the log K ow values
estimated for the switch from membrane control to WBL control of SPMD and
fish uptake rates (Table 7.1).
Even for less persistent HOCs, such as a diverse mixture of PAHs, reasonable
correlations have been found between the uptake rate constants for Pacific oysters
(Crassostrea gigas) and standard SPMDs in controlled side-by-side laboratory
exposures (Huckins et al., 2004). In this study, oysters and SPMDs were exposed
to three concentrations of PAHs (10 ng L−1 , 100 ng L−1 and 250 ng L−1 ; nominal
values) in a flow-through system for 20 d and samples were collected every five
days. Uptake rate constants were derived from residue concentrations of whole
tissue wet weights and whole SPMDs. Figure 7.4 shows that SPMD and oyster
ku s correlated well (r 2 = 0.81) for the high treatment level, where feeding activity
was not observed. Note that the open circles represent data points not included in
the regression analysis, because there was a systematic bias (likely an analytical
artifact) of values for PAHs with log K ow s within the range of 5.6 to 6.4. Overall,
SPMD ku s averaged 1.4 fold higher than the oyster ku s and the C.V.s associated
with mean SPMD and oyster ku s were 17% and 39%, respectively. Eastern oyster
150 Chapter 7
0.65
0.60
0.55
0.50
ku,S (L g--1 d--1)
0.45
0.40
0.35
0.30
0.25
0.1 0.2 0.3 0.4 0.5 0.6 0.7
--1 --1
ku,O (L g d )
FIGURE 7.4 Relationship between Pacific oyster (Crassostrea gigas) and SPMD uptake-rate con-
stants (ku,O and ku,S respectively), for test chemicals covering the range of test chemical log Kow s
(250-ng L−1 treatment). Test chemicals within the range of log Kow 5.6 to 6.4 are shown as open
symbols but are not used in the regression (see text for explanation). Reprinted from Huckins
et al. (2004), copyright (2004); reproduced with permission from Alliance Communication
Group.
(Crassostrea virginica) ku s for a number of PAHs have been reported by Bender

et al. (1988). The mean of the ratios obtained by dividing the SPMD ku s of Huckins
et al. (2004) into the eastern oyster ku s of Bender et al. (1988) for the same PAHs is
2.6 ± 1.4 (n = 10). This is a relatively small difference considering that exposure
conditions can affect both organism and SPMD ku s (e.g., Björk and Gilek, 1997;
Booij et al., 1998).
The correlation shown in Figure 7.4 would not be expected to hold for PAHs,
when using finfish as a test species. This observation is based on the much higher
levels of cytochrome P-450 dependent enzymes in finfish than in shellfish (Buhler
and Williams, 1989). The P-450 system mediates Phase I metabolism of PAHs,
which in this case consists of enzyme-catalyzed oxidation of the non-polar parent
PAH molecule.
7.4. RELATIVE AMOUNTS ACCUMULATED
In another controlled study, Sabaliūnas et al. (1998) exposed SPMDs and

lake mussels (Anodonta piscinalis) to four pesticides in a laboratory flow-through
FIGURE 7.5 Ratios of test chemicals in SPMDs and mussels given as percentages of the total
residues. Reprinted from Sabaliūnas et al. (1998), copyright (1998); reproduced with permission
from Alliance Communication Group.
system, equipped with a 105 L test chamber. The exposure period was 20 d and
samples were collected on d 3, 8, 14 and 20. The pesticides used were chlordane,
endosulfan, allethrin, and fenvalerate. The mussels were not fed during the exper-
iment. Based on wet tissue weights and whole SPMDs, SPMD ku s were 3.5 to 5.5
times greater than those for the same pesticides in mussels. However, the pattern
of accumulated residues was very similar as shown by Figure 7.5.
Data from the Meadows et al. (1998) study was further analyzed (Echols et al.,
1996) by using principle component analysis (PCA) to compare PCB congener
concentrations in technical Aroclor mixtures, contaminated spring water, caged
brown trout, SPMDs and hexane filled dialysis bags (Södergren, 1987; also, see
description in Chapter 1). As stated earlier, caged fish were not fed and exposure
conditions were quite stable during the 28 d exposure. Figure 7.6 shows the re-
sults of the PCA. The clustering of fish and SPMDs together indicates the close
similarity of the concentration profiles of major congeners. This PCA plot fur-
ther supports the good correlation between SPMD and brown trout rate constants
shown in Figure 7.3. However, in vitro bioassay of the SPMD and trout extracts
using the PLHC-1 (fish [Poeciliopsis lucida] hepatoma cells) ethoxyresorufin-O-
deethylase (EROD) test indicated that the toxicity of the two types of extracts
differ substantially (Meadows, 2005). When the EROD response was normalized
to the mass of the total extracted PCBs, EROD induction was thirty times greater
152 Chapter 7
FIGURE 7.6 Principle components analysis (PCA) of PCB congener concentrations in technical
Aroclor mixtures, contaminated water, caged brown trout, SPMDs, and hexane filled dialysis
bags. The plot shows that 77% of the variance of samples within the 95% confidence ellipse
is explained by PC1 and PC2 and that caged fish and SPMDs are clustered together (PCA plot
courtesy of Kathy Echols, USGS-CERC, Columbia, MO, USA).
for SPMD extracts than for fish extracts. This assay is highly sensitive to 2,3,7,8-
tetrachlorodibenzo-p-dioxin (TCDD) and dioxin-like compounds, which include
non-ortho Cl substituted PCB congeners with four or more chlorines. Therefore,
it is likely that difference in the toxicity of SPMD and trout extracts stems from
relatively higher levels of non-ortho Cl PCBs in the SPMD extracts. To explain
this finding, Meadows (2005) hypothesized that non-ortho Cl PCBs were selec-
tively metabolized by the fish. The work of Gale et al. (1997) and Peterman (2005)
indicate that levels of congener 77 (a non-ortho Cl tetrachlorobiphenyl) in channel
catfish (Ictalurus punctatus) are lower than expected based on EP theory, which
supports the selective metabolism hypothesis of Meadows (2005). Finally, because
non-ortho Cl PCBs represent <1% of the mass of commercial PCB mixtures, they
are not principle components of the PCA shown in Figure 7.6.
These studies show that when the accumulation of HOCs occurs solely by
respiration or dermal absorption, BMO and SPMD rate constants correlated very
well, but that concentrations in SPMDs are often higher than those in BMOs. How-
ever, good correlations between the “fingerprints” or patterns of HOC residues in
BMO tissues and SPMDs would not be expected to hold, when diet plays a major
role in the uptake of compounds with high K ow s (see Eq. 7.5). For example, Peven
et al. (1996) compared the accumulation of organic contaminants by transplanted
blue mussels and standard SPMDs at a known contaminated site in Dorchester

Bay, MA, USA. The exposure duration was 95 d and mussels likely fed during the
study. Mussel and SPMD concentrations were based on dry weights and whole
SPMDs, respectively, and the analytes measured were PAHs, PCBs and total DDT.
Mussels were not depurated; thus residue concentrations represented gut contents
as well as tissues. Overall, concentrations of total PAHs, PCBs and DDTs were 2 to
7 times higher in SPMDs than in mussels, indicating that differences would have
been even greater on a wet-weight basis. Also, Peven et al. (1996) found that when
PCB and PAH concentrations in the two sampling matrices were normalized to
the highest peak in each mixture, high K ow congeners or components represented
a larger fraction in mussels than the corresponding fraction in the SPMDs. Con-
versely, moderate to low K ow compounds were higher in SPMDs than in mussels.
However, the total amounts (e.g., ng) of these high K ow compounds accumulated
in an SPMD sample was about the same as that found in a similar sized BMO sam-
ple. Furthermore, Booij et al. (2002) and McCarthy and Gale (2001) have shown
that SPMDs accumulate sufficient residue mass for the quantitation of ultra-trace
levels of extremely hydrophobic contaminants such as decabromodiphenyl ether
and 2,3,7,8-tetrachlorodibenzo- p-dioxin.
7.5. INDEPENDENCE OF CONCENTRATION FACTORS

RELATIVE TO EXPOSURE CONCENTRATION
Few reports are available on the potential effect of chemical concentration

on the BAF in an aquatic organism (e.g., Mayer, 1976). Yet, a key assumption
of EP theory is the independency of BAFs relative to exposure concentration.
To our knowledge, there is only one report (Huckins et al., 2004) in the peer-
reviewed literature, where the effect of chemical exposure level on concentration
factors (CFs) or BAFs has been tested in side-by-side BMO and passive sampler
exposures. Huckins et al. (2004) defined CF as the ratio of the concentration in a
sample matrix (whole body [soft tissues in the case of bivalves] or whole SPMDs)
relative to the concentration in the ambient exposure medium at any moment in
time, whereas the K sw and BAF (includes biomagnification) represent the maximal
CF. Similar to K sw s and BAFs, CFs are expected to be independent of exposure
concentrations, when residue exchange follows first-order kinetics.
Figure 7.7 shows the CFs of 10, 100 and 250 ng L−1 (nominal treatment levels;
actual concentrations were less than nominal and were inversely proportional to
hydrophobicity) of PAHs in SPMDs and oysters (Crassostrea gigas) after a 20 d
exposure period (Huckins et al., 2004). Oyster CFs were generally independent
of test chemical concentrations for PAHs having low to moderate log K ow s (i.e.,
log K ow s of 3.4 to ≤5.3). Because oysters in the 10-ng L−1 treatment appeared
to feed while no evidence of feeding was observed in the 250-ng L−1 treatment,
feeding seemed to have little effect on the CFs of these chemicals. This observation
corresponds well with the pioneering work of Bruggeman et al. (1981). They
154 Chapter 7
FIGURE 7.7 20-d concentration factors (CF s) in whole SPMDs and Pacific oysters (Crassostrea
gigas) exposed to 10, 100, and 250 ng L−1 of PAHs.
found that food chain or dietary uptake of chemicals exceeded respiratory uptake
by fish only when chemicals were quite hydrophobic, i.e., log K ow s >5. Based
on Bruggeman et al. (1981) and Connell (1990), the apparent feeding activity
of oysters in the 10 ng L−1 treatment would not be expected to affect CFs of
compounds with low to moderate K ow s. For compounds with log K ow values
between 3.4 and 5.3, residue accumulation in oyster tissues followed first-order
kinetics and EP theory (Figure 7.7). Only the five PAHs with the lowest K ow s
attained steady state concentrations in oyster tissues at the end of the 20-d exposure,
while other PAHs tested were in the curvilinear phase of uptake.
As expected, the magnitude of PAH CFs in SPMDs is generally independent
of exposure concentrations (Figure 7.7). The possible exception was the 2-fold
higher CFs of PAHs (log K ow s ≥ 5.6 and ≤6.4) in the 10 ng L−1 treatment versus
the same PAHs in the 250 ng L−1 treatment. However, the authors concluded
that this difference was artifactual in nature (Huckins et al., 2004). SPMD CFs for
PAHs within the log K ow range of 3.4 to ≤5.3 were generally more than an order of
magnitude higher than the corresponding CFs for oysters (Figure 7.7). A plausible
explanation for the differences in SPMD and oyster CFs for low to moderate
K ow compounds is the much higher percentage of lipid in SPMDs (effectively
>20%; see Section 5.6.2.) than in oysters (≈1.6%; wet weight). Although the
lipid contents of BMOs and SPMDs do not directly affect the linear uptake rate
constants (Section 7.2.), a higher percentage of lipid does result in a smaller ke
and a longer duration of the integrative or linear uptake phase. Assuming that
metabolism of these test chemicals is insignificant, that the lipid in oysters and
SPMDs is of similar quality, and that equilibrium was attained in oyster tissues, it
seems reasonable to assume that the approximately 12-fold higher lipid content of
SPMDs compared to oysters explains much of the difference in CFs between the
two sampling matrices. Nevertheless, metabolism of these PAHs cannot be ruled
out as a causative factor in the observed differences between SPMD and oyster
CFs, because cytochrome P-450 levels in some mollusks are fairly high (Buhler
and Williams, 1989).
For PAHs with log K ow s ranging from 3.4 to 5.3, differences between the
CFs in SPMDs and in oysters decreased with increasing K ow (Figure 7.7). In
the case of PAHs with log K ow s ≥5.6, oyster CFs were much higher in the low
treatment (10 ng L−1 ), where feeding appeared to occur, than in the high treatment
(250 ng L−1 ), where feeding activity appeared to be minimal. The ability of juvenile
and adult bivalve mollusks to close their valves for extended periods to avoid
toxicants is well documented (Huebner and Pynnönen, 1992; Goudreau et al.; 1993;
ASTM, 1996, 2001; Moring and Rose, 1997). Furthermore, Toro et al. (2003) have
shown an inverse relationship between the ku s of PAHs in the marine giant mussel
Choromytilus chorus and PAH concentrations at field sites. Oyster CFs for high-
K ow HOCs were consistent with the valve closure or reduced feeding scenario,
because CFs were inversely dependent on exposure concentration (Figure 7.7).
The differences in oyster CFs or BAFs across exposure concentrations increased
with HOC K ow , and culminated in a 13-fold difference for benzo[g, h, i]perylene.
Oyster CFs are also 1.3- to 2.9-fold higher than the corresponding SPMD CFs for
PAHs with log K ow s ≥5.6 (10 ng L−1 SPMD and oyster treatments), apparently
as a result of their somewhat higher uptake rate constants.
7.6. SIMILARITY OF ELIMINATION OR EQUILIBRATION RATE

CONSTANTS
Although the relative magnitudes of SPMD and oyster ku s were similar under
the conditions of the Huckins et al. (2004) exposure, the ke s and the associated
half-lives (t1/2 s) of residues in the two sampling matrices were markedly different.
Table 7.2 summarizes selected bivalve and SPMD first-order ke s and t1/2 s for
PAHs. The t1/2 is related to ke by
t1/2 = ln 2/(ku /K sw ) (7.6)
In all cases, bivalve ke s are much greater than those of SPMDs, resulting in
much shorter t1/2 s of test compounds. These data are generally consistent with the
hypothesis that the fractional lipid content (effectively >20% for SPMDs, versus
about 1 to 3% for bivalves [wet weight]) controls the magnitude of ke s and t1/2 s.
However, closer inspection of Table 7.2 and ke and t1/2 values from several other
studies (e.g., Pruell et al., 1986; Sericano et al., 1996) suggest that bivalves depurate
PAHs with log K ow s >5.8 at greater rates than PAHs with log K ow s ranging from
5 to 5.8. These data do not follow a hydrophobicity model (Barron, 1990). Thus,
it appears that active physiological processes such as metabolism may play a role
in PAH clearance from bivalve tissues, especially for high K ow test chemicals.
TABLE 7.2 Comparison of Elimination Rate Constants (ke s; d−1 ) and Halflives (t1/2 s; d) Determined in SPMDs and Bivalves 156
Huckins et al. (2004)a Huckins et al. (1999) Bender et al. (1988) Tse et al. (2000)
Oysters SPMDs SPMDs Oysters Mussels SPMDs
Compound ke (t1/2 ) ke (t1/2 ) ke (t1/2 ) ke (t1/2 ) ke (t1/2 ) ke (t1/2 )
naphthalene —b 0.18 (3.8) 0.074 (9.6) — — —

biphenyl — 0.05 (14) — — —
1-methylnaphthalene — 0.07 (9.9) — — —
acenaphthylene — 0.06 (11) 0.06 (11) — — —
acenaphthene — 0.04 (17) 0.04 (17) — — —
1-ethylnaphthalene 0.31 (2.2) 0.02 (35) — — — —
1,3-dimethylnaphthalene 0.28 (2.5) 0.01 (69) — — — —
fluorene 0.31 (2.2) 0.02 (35) 0.02 (35) — — —
phenanthrene 0.17 (4.1) <MQLc 0.02 (35) 0.21 (3.3) — —
anthracene 0.13 (5.3) <MQL 0.01 (69) — 0.22 (3.1) 0.07 (9.9)
2,3,5-trimethylnaphthalene 0.14 (4.9) — — — — —
fluoranthene 0.05 (14) <MQL 0.02 (35) 0.12 (5.8) 0.31 (2.2) 0.01 (69)
pyrene 0.04 (17) <MQL 0.01 (69) 0.10 (6.9) 0.30 (2.3) 0.01 (69)
benz[a]anthracene 0.01 (69) NMLc <0.01 (173) 0.04 (17) — —
2-methylfluoranthene 0.01 (69) — — — — —
chrysene 0.03 (23) NML — 0.05 (14) — —
benzo[b]fluoranthene 0.03 (23) NML — 0.01 (69) — —
benzo[k]fluoranthene 0.04 (17) NML — — — —
benzo[a]pyrene 0.03 (23) NML — 0.03 (23) — —
perylene 0.03 (23) NML — 0.07 (9.9) — —
indeno[1,2,3-cd]pyrene 0.05 (14) NML — — — —
dibenz[a, h]antracene 0.07 (10) NML — — — —
benzo[g, h, i]perylene 0.06 (11) NML — 0.06 (11) — —
a
250 ng L−1 treatment, ke values are means (n = 4), and C.V.s (in parenthesis) for oysters ranged from 16 to 192%, whereas C.V.s for SPMDs ranged from 11 to 75%.
b
Not measured or achieved equilibrium.
c
Method quantitation limit (MDL) or no measurable loss (NML).
Chapter 7
FIGURE 7.8 Model simulation (Gale, 1998) of the response of oysters and SPMDs to a 3-d pulse
of fluorene. Reprinted from Huckins et al. (2004), copyright (2004); reproduced with permission
from Alliance Communication Group.
Clearly, SPMD ke s follow a hydrophobicity model, where ke s fall with increasing

K ow s (Table 7.2).
The lower ke values in SPMDs compared to BMOs have a major effect on
the retention of contaminants that are absorbed during episodic exposure events.
Figure 7.8 shows the results of a model simulation (Gale, 1998) of fluorene uptake
by oysters and SPMDs, following an episodic fluorene release into water, using
the exchange kinetics data reported by Huckins et al., (2004). According to the
model, readily detectable levels of fluorene were retained by the SPMDs 25 d after
the exposure event, whereas fluorene was completely cleared from oysters in 15
d following the exposure (Fig. 7.8). The first-order residence time (tm or 1/ke ;
the mean length of time that a molecule remains in a sampling matrix subject to
first-order exchange kinetics) of fluorene in oysters and SPMDs was 3 and 50 d,
respectively. This simulation highlights a major difference between bivalves and
SPMDs. Phillips and Rainbow (1993) have suggested that the implications of short
residence times of many chemicals in bivalves exposed to variable or episodic ex-
posure conditions has not been fully recognized by many investigators employing
BMOs in monitoring studies. To minimize this problem a “time bulking” approach
has been recommended (Phillips and Rainbow, 1993), which consists of matching
the end of the linear phase of BMO uptake kinetics to sampling times. Unfortu-
nately, BMO exchange kinetics are not known a priori, and sampling through time
is labor intensive and costly. However, differences between the clearance rates of
persistent HOCs from SPMDs and high lipid content BMOs, such as carp, are not
expected to be great.
158 Chapter 7
7.7. COMPARABILITY OF ESTIMATED AND MEASURED BCF s

OR BAF s
Burkhard et al. (2003) have stated that BAFs can only be measured using
field data, because bioaccumulation is the net result of chemical uptake from all
sources and processes. Also, a number of investigators have reported that very
long exposures (e.g., ≥100 d) are required to reach equilibrium with BMOs (Booij
et al., 2002; Verweij et al., 2004) for some high K ow compounds, which may be
a source of errors for some regression models. Although we agree with Burkhard
et al. (2003) that BAFs are unique to a particular environment and BMO species,
there is still a need for screening models and methods to estimate BAFs. Non-
physiological or EP based models assume that the log BCF (i.e., the BAF in the
case where diffusional exchange via the gills and skin is the only uptake and loss
mechanism) is linearly related to the K ow . The classic Veith et al. (1979) model is
an example of this approach.
log BCF = 0.76 log K ow − 0.23 (7.7)
This regression equation was derived using four species of fish and 84 or-
ganic compounds with a wide range of hydrophobicities and generally provides
reasonable predictions for compounds with log K ow s < 6. However, some investi-
gators (e.g., Connell and Hawker, 1988) suggested that when compounds covering
a very wide range of K ow s (includes very high to super hydrophobic compounds)
are considered, biological responses and BCFs deviate substantially from a linear
relationship. For example, Connell and Hawker (1988) successfully used the fol-
lowing polynomial equation to model BCFs of fish for compounds with log K ow s
ranging from about 3 to 9.5.
Log BCF = 0.0069(log K ow )4 − 0.185(log K ow )3
+ 1.55(log K ow )2 − 4.18 log K ow + 4.79 (7.8)
The maximum BCF of 4.6 is achieved at a log K ow of 6.7. Connell (1990) pointed
out that if only compounds with log K ow s within the range of 3 to 6 are considered
Eq. 7.8 reduces to
Log BCF = 0.94 log K ow − 1.0 (7.9)
which is similar to Eq. 7.7. Steady state K sw s correlate well to K ow s using the
following non-linear relationship (Eq. 3.28)
log K sw = a0 + 2.321 log K ow − 0.1618(log K ow )2 (7.10)
where the intercept a0 is equal to −2.61 for the chemicals examined in this section.
Table 7.3 gives SPMD K sw s (Eq. 7.10), lipid-normalized BCFs (Eqs. 7.7 and
7.8), and BCF/K sw ratios for model compounds with log K ow values ranging from
3.0 to 6.0. In this comparison, a lipid content of 5% was adopted for lipid normal-
ization of whole body fish concentrations (note that the whole body lipid content
TABLE 7.3 Comparison of Regression Model Estimates of Ksw s and

Lipid-Normalized BCFsa
log K sw log BCF log BCF BCF/K sw BCF/K sw

log K ow Eq. 7.10 Eq. 7.7 Eq. 7.8 Eq. 7.7 Eq. 7.8
3 2.9 3.4 3.1 3.2 1.6

4 4.1 4.1 4.1 1.0 1.0
5 5.0 4.9 5.1 0.8 1.3
6 5.5 5.6 5.8 1.3 2.0
a
A 5% lipid content was assumed for fish.
in fishes typically range from about 2 to 10%). The BCF/K sw ratios produced
when using Eqs. 7.7 and 7.8 (Table 7.3) show that SPMD K sw s are within about
three-fold of BCFs over 3-orders of magnitude and that the mean of BCF/K sw
ratios for Eqs. 7.7 and 7.8 is about 1.5. The BCF/K sw ratios (1.6 ± 1.1) produced
from Eq. 7.7 BCFs are more variable than the BCF/K sw ratios (1.5 ± 0.4) pro-
duced from Eq. 7.8 BCFs. Overall, the computed K sw s are generally lower than
lipid-normalized BCFs.
The work of Wang et al. (2002) appears to be a rare case where lipid normal-
ization of SPMD and feral fish concentrations may have been partly justified. This
observation is based on the facts that the flow during the exposure was about 100
cm s−1 (Huaihe River, China) and that the deployment device did little to buffer
the flow at the membrane surface. Flows of this magnitude are known to greatly
enhance sampling rates and thus reduce times to attain equilibrium concentrations
in SPMDs (see 90 cm s−1 sampling rate data in Table A.7 of Appendix A, and
discussions in Sections 3.6.2. and 3.6.5.). For example, examination of PCB con-
gener 52 (2,5,2 ,5 -tetrachlorobiphenyl) calibration data in Tables A.2 and A.7 of
Appendix A show that the log K sw is about 5.6 and that at a flow rate of 90 cm s−1
across the membrane surface, the SPMD ku = 26 L g−1 d−1 . With this data, the
t1/2 for attainment of equilibrium can be estimated from Eq. 7.6. Using this ap-
proach gives a t1/2 of about 10 d, which means that during the 28-day exposure
period about 3 t1/2 s elapsed and that SPMDs should have reached about 88% of
the equilibrium concentrations of congener 52. Wang et al. (2002) reported lipid
normalized PCB concentrations in SPMDs and fish for 21 congeners bracketed by
congeners 5 and 101. The discussion of this data is limited to 12 congeners brack-
eted by congeners 5 and 52 to increase the likelihood that equilibrium was attained.
In all cases within the designated congener range, lipid normalized SPMD concen-
trations were higher than or equal to fish concentrations. The mean of individual
values (n = 12) obtained by dividing SPMD concentrations by fish concentrations
(i.e., K sw /BAF) was 2.3 ± 1.1, which may be due to the fact that the sorption ca-
pacity of the membrane was not taken into account. Furthermore, this analysis did
not take into account that the effective lipid content of the whole SPMD varies
with analyte partition coefficients between the membrane and lipid (see Section
160 Chapter 7
5.6.2.). Also the study did not employ performance reference compounds, which
are required to validate attainment of equilibrium. We include this exercise to show
that lipid normalization is problematic even when both matrices attain equilibrium
and do not recommend the procedure without further proof of concept.
7.8. DIETARY UPTAKE AND BIOMAGNIFICATION
A major criticism of the use of SPMDs and other passive samplers for estimat-
ing screening level BAFs is that dietary uptake is not modeled and thus the potential
for biomagnification is not taken into account. Biomagnification is the sequential
increase in chemical concentration going up the trophic levels of a food chain. The
biomagnification factor BMF is determined by dividing the concentration in the
consumer by the concentration in the diet. However, Connell (1990), has observed
that the first step in most aquatic food chains is bioconcentration (i.e., the uptake
of dissolved-phase residues by autotrophic organisms such as phytoplankton). It
is also likely that the dominant route of chemical uptake is water even for food
chains based on hetrotrophs and fungi (Schmidt, 2004). Passive samplers, such
as SPMDs, provide the best available technology for determining dissolved phase
water concentrations of trace to ultra-trace bioaccumulative organics and thus en-
able in situ determination of HOC exposure to organisms at the lowest trophic
level in nearly all food chains. Furthermore, the contribution of dietary uptake is
generally very small for compounds with log K ow s < 5.5 (e.g., Connell, 1990;
Huckins et al., 2004).
Connell (1990) also proposed that, irrespective of whether food or water
is the primary source of accumulated chemical, BMF values are near unity in
aquatic food chains when differences in lipid content are taken into account. More
recently, there has been a general acceptance that even after taking differences
in lipid contents into account, BMFs >1 do occur in some aquatic food chains
(Macdonald, et al., 2002). Typically, BMFs in finfishes are small (e.g., 3.0-fold)
when compared to mammals or birds (e.g., 30-fold) fed similar diets. Finally, until
the advent of passive samplers such as the SPMDs, BMF multipliers have been
easier to estimate than the dissolved phase exposure concentrations. Knowledge
of dissolved phase chemical concentrations is a critical part of understanding
how aqueous exposure levels relate to the concentrations of residues measured in
organisms in various trophic levels of aquatic ecosystems.
7.9. DO SPMDs QUALIFY AS BIOMIMETIC SAMPLERS?
In the context of this chapter, biomimetic is defined as “the use of simple

synthetic media to mimic a complex biological process”. Earlier Fendler (1984)
defined membrane biomimetic chemistry as “processes in simple media that mimic
aspects of biomembranes”. Thus, classical biomimetic approaches target specific
sites and/or modes of action characteristic of a particular species, and are rarely
applied to a whole organism or a group of organisms. However, the processes of
diffusion and partitioning mediate biouptake at the base of aquatic food chains
and are fundamental to homeostasis of all organisms. Also, the bioavailability
of HOCs is largely controlled by diffusion and partitioning processes, which in
turn affect the accumulation of HOCs by both SPMDs and BMOs. In the case of
SPMDs, the rate-limiting step in chemical uptake and release kinetics is always
diffusion, and the maximum achievable concentration for a particular exposure
level is determined by the chemical’s partition coefficient (K sw ). Although diffu-
sion may not always be the rate-limiting step in chemical uptake by biota, it is
always a key mechanism in the overall uptake process. Likewise, partitioning gen-
erally plays an important role in HOC concentrations in BMOs, but the maximum
tissue level during a specified interval of time or an organism’s life cycle may
not correspond to a chemical’s partition coefficient because of variations in en-
vironmental exposure concentrations, kinetic limitations, biotransformation, and
biomagnification.
In this chapter we have shown that the relative fluxes of a wide range of
solutes across the gill membrane and the SPMD membrane appear to fit the same
pattern (Figure 7.1), suggesting that similar processes are involved. Also, uptake
rate constants (i.e., ku s) of SPMDs and BMOs generally appear to be of similar
magnitude, and in some cases are reasonably well correlated (Figures 7.3 and
7.4), but this finding may be limited to organism exposure scenarios where water
represents the major route of HOC uptake. A major difference between some
BMOs and standard SPMDs is in the much higher values of ke s found for BMOs
relative to SPMDs. This observation applies to those organisms with low levels of
total lipids (especially neutral triglycerides or storage fats) such as bivalves, and
those with enzymatic systems capable of metabolizing HOCs, but may not apply
to persistent HOCs accumulated in organisms with high fat contents (i.e., >10%).
Because the magnitude of the overall ke largely controls equilibration time, most
BMOs are expected to reach equilibrium more rapidly than SPMDs. Regardless
of kinetic differences, regression model predictions (Eqs. 7.7, 7.8 and 7.10) of
steady-state values of K sw s and BCFs in fish appear to agree within about 2-fold
after differences in lipid contents are taken into account (Table 7.3).
In summary, the large number of variables which potentially affect the ac-
cumulation of HOC in BMOs, suggest that it is unrealistic to expect any single
passive sampler to be biomimetic of all BMOs. Also, it is similarly unrealistic to
expect that one or two species of BMOs mimic bioaccumulation in all organisms of
concern. Literature values of BCFs or BAFs vary by several orders of magnitude for
the same chemical across test species and exposure scenarios (e.g. Mackay et al.,
1992a, 1992b, 1997), which underscores the problem of selecting representative
sampler designs or BMOs. Thus, SPMDs are biomimetic only when diffusional-
partitioning processes mediate HOC concentrations in organisms of concern (i.e.,
when residue accumulation in organism tissues follows EP theory). The similar-
ity of SPMDs and BMOs is reflected in the proportionality of their uptake rate
162 Chapter 7
constants (whole-weight basis) and partition coefficients (Table 7.3; ratio of lipid-
normalized BCF to whole-weight K sw ≈ 1.4) across a range of Kow s. Again, we
stress that assessment of the similarity of partition-coefficients can only be made
when both SPMDs and BMOs have attained equilibrium. Because SPMDs are
designed to remain in the linear uptake mode for compounds with log K ow s > 4.5
during typical 4-week exposure periods, the attainment of equilibrium by SPMDs
is an exception rather than a rule.
The primary role of SPMDs and other passive samplers is to provide conve-
nient, powerful analytical tools for determining dissolved and vapor phase HOC
concentrations in environmental systems. This chapter has shown that they are
also useful as biomimetic screening tools for estimating exposure of organisms to
bioconcentratable compounds and for deriving BCFs based on EP theory. Even
for those chemicals that are present at vanishingly small amounts in the dissolved
phase and are primarily accumulated via the dietary uptake, SPMDs generally
extract sufficient amounts of residues for analysis.
7.10. REFERENCES
American Society for Testing and Materials (ASTM) 1984, Standard Practice for Conducting Biocon-
centration Tests with Fishes and Saltwater Bivalve Mollusks. Annual Book of ASTM Standards
No. E 1022-84; Vol 11.05; Philadelphia; PA; pp. 286–303.
American Society for Testing and Materials (ASTM) 1996, Standard Guide for Conducting Acute
Toxicity Tests on Test Materials with Fishes, Macroinvertebrates, and Amphibians. Annual Book
of ASTM Standards, No. E 729-96; Vol 11.05; Philadelphia, PA; pp. 178–199.
American Society for Testing and Materials (ASTM) 2001, Standard Guide for Conducting In-situ Field
Bioassays with Marine, Estuarine, and Freshwater Bivalves. Annual Book of ASTM Standards;
No. E 2122-01; Vol 11.05; Philadelphia, PA; pp. 1562–1591.
Axelman, J.; Næs, K.; Näf, C.; Broman, D. 1999, Accumulation of polycyclic aromatic hydrocarbons in
semipermeable membrane devices and caged mussels (Mytilus edulis) in relation to water column
phase distribution. Environ. Toxicol. Chem. 18: 2454–2461.
Baumard, P.; Budzindki, H.; Garrigues, P. 1998a, PAHs in Arcachon Bay, France: Origin and biomon-
itoring with caged organisms. Mar. Pollut. Bull. 36: 577–586.
Baumard, P.; Budzinski, H.; Garrigues, P.; Sorbe, J.C.; Burgeot, T.; Bellocq, J. 1998b, Concentrations
of PAHs (polycyclic aromatic hydrocarbons) in various marine organisms in relation to those in
sediments and to trophic level. Mar. Pollut. Bull. 36: 951–960.
devices during chronic exposure to dispersed crude oil. Environ. Toxicol. Chem. 20: 1175–1184.
Barron, M.G. 1990, Bioconcentration: Will water-borne organic chemicals accumulate in aquatic or-
ganisms. Environ. Sci. Technol. 24: 1612–1618.
Bayne, B.L.; Thompson, R.J.; Widdows, J. 1976, Physiology: I. In Marine Mussels: Their Ecology
and Physiology; Bayne, B.L., Ed.; Cambridge University Press: London; pp. 121–206.
Bender, M.E.; Hargis, W.J. Jr; Huggett, J.R.; Roberts, M.H. Jr. 1988, Effects of polynuclear aromatic
hydrocarbons on fishes and shellfishes: An overview of research in Virginia. Mar. Environ. Res.
24: 237–241.
Björk, M. and Gilek, M. 1997, Bioaccumulation kinetics of PCB 31, 49, and 153 in the blue mussel,
Mytilus edulis L. as a function of algal food concentration. Aquat. Toxicol. 38: 101–123.
(Mytilus edulis). Chemosphere 46: 683–688.
Bruggeman, W.A.; Martron, L.B.J.M.; Kooiman, D.; Hutizinger, O. 1981, Accumulation and elimina-
tion kinetics of di-; tri- and tetrachlorobiphenyls in goldfish after dietary and aqueous exposure.
Bucheli, T. and Gustafsson, Ö. 2000, Quantification of the soot-water distribution coefficient of PAHs
provides mechanistic basis for enhanced sorption observations. Environ. Sci. Technol. 34: 5144–
5151.
Buhler, D.R. and Williams, D.E. 1989, Enzymes involved in metabolism of PAHs by fishes and other
aquatic animals: Oxidative enzymes (or Phase I enzymes). In Metabolism of Polycyclic Aromatic
Hydrocarbons in the Aquatic Environment; Varanasi, U., Ed.; CRC Press: Boca Raton, Fl; pp.
151–184.
Burkhard, L.P.; Cook, P.M.; Mount, D.R. 2003, The relationship of bioaccumulative chemicals in
water and sediment to residues in fish: A visualization approach. Environ. Toxicol. Chem. 22:
2822–2830.
Connell, D.W. and Hawker, D.W. 1988, Use of polynomial expressions to describe the bioconcentrations
of hydrophobic chemicals by fish. Ecotox. Environ. Safe. 16: 242–251.
Connell, D.W. 1990, Bioaccumulation of Xenobiotic Compounds. CRC Press: Boca Raton, FL; p. 219.
Echols, K.R.; Meadows, J.; Schwartz, T.; Borsuk, F.; Carline, R.; Tillitt, D. 1996, Comparison of
SPMDs, Caged Fish, and Hexane Filled Dialysis Bags for Sampling PCBs in Groundwater from
a Spring at Sherman’s Dale, PA. Presented at the 4th Annual SPMD International Workshop:
Kansas City, MO. June 11–13, 1996.
Echols, K.R.; Gale, R.W.; Schwartz, T.R.; Huckins, J.N.; Williams, L.L.; Meadows, J.C.; Morse, D.;
Petty, J.D.; Orazio, C.E.; Tillitt, D.E. 2000, Comparing polychlorinated biphenyl concentrations
and patterns in the Saginaw River using sediment, caged fish, and semipermeable membrane
devices. Environ. Sci. Technol. 34: 4095–4102.
Fendler, J.H. 1984, Membrane mimetic chemistry. C&E News; January 2; 1984; p. 25–39.
Følsvik, N.; Brevik, E.M.; Berge, J.A. 2002, Organotin compounds in a Norwegian fjord. A comparison
of concentration levels in semipermeable membrane devices (SPMDs), blue mussels (Mytilus
edulis) and water samples. J. Environ. Monitor. 4: 280–283.
Franke, C.; Studinger; G., Berger, G.; Böhling, S.; Bruckmann, U.; Cohors-Fresenborg, D.; Jöhncke,
U. 1994, The assessment of bioaccumulation. Chemosphere 29: 1501–1514.
Gale, R.W.; Huckins, J.N.; Petty, J.D.; Peterman, P.H.; Williams, L.L.; Morse, D.; Schwartz, T.R.;
Tillitt, D.E. 1997, Comparison of the uptake of dioxin-like compounds by caged channel catfish
and semipermeable membrane devices in the Saginaw River, Michigan. Environ. Sci. Technol. 32:
178–187.
Gatermann, R.; Siselli, S.; Hühnerfuss, H.; Rimkus, G. G.; Hecker, M.; Karbe, L. 2002, Synthetic
musks in the environment. Part 1: Species-dependent bioaccumulation of polycyclic and nitro
musk fragrances in freshwater fish and mussels. Arch. Environ. Con. Tox. 42: 437–446.
Gilek, M.; Björk, M.; NäF, C. 1996, Influence of body size on the uptake, depuration, and bioaccu-
mulation of polychlorinated biphenyl congeners by Baltic Sea blue mussels, Mytilus edulis. Mar.
Biol. 125: 499–510.
164 Chapter 7
Gobas, F.A.P.C.; Opperhuizen, A.; Hutzinger, O. 1986, Bioconcentration of hydrophobic chemicals in

fish: Relationship with membrane permeation. Environ. Toxicol. Chem. 5: 637–646.
Gobas, F.A.P.C.; Zhang, X.; Wells, R. 1993, Gastrointestional magnification: the mechanism of bio-
magnification and food chain accumulation of organic chemicals. Environ. Sci. Technol. 27: 2855–
2863.
Goudreau, S.E.; Neves, R.J.; Sheehan, R. 1993, Effects of wastewater treatment plant effluents of
freshwater mollusks in the upper Clinch River, Virginia, USA. Hydrobiologia 252: 211–230.
Granmo, Å.; Ekelund, R.; Berggren, M.; Brorström-Lunden, E.; Bergqvist, P.-A. 2000, Temporal
trend of organochlorine marine pollution indicated by concentrations in mussels, semipermeable
membrane devices, and sediment. Environ. Sci. Technol. 34: 3323–3329.
Gray, J.S. 2002, Perceived and real risks: Produced water from oil extraction. Mar. Pollut. Bull. 44:
1171–1172.
Hayton, W.L. and Barron, M.G. 1990, Rate-limiting barriers to xenobiotic uptake by the gill. Environ.
Toxicol. Chem. 9: 151–157.
Hebert, C.E. and Keenleyside, K.A. 1995, To normalize or not to normalize? Fat is the question.
Herve, S.; Prest, H.F.; Heinonen, P.; Hyötyläinen, T.; Koistinen, J.; Paasivirta, J. 1995, Lipid-filled
semipermeable membrane devices and mussels as samplers of organochlorine compounds in lake
water. Envrion. Sci. Pollut. R. 2: 24–30.
Hofelt, C.S. and Shea, D. 1997, Accumulation of organochlorine pesticides and PCBs by semipermeable
membrane devices and Mytilus edulis in New Bedford Harbor. Environ. Sci. Technol. 31: 154–159.
Huckins, J.N.; Tubergen, M.W.; Manuweera, G.K. 1990, Semipermeable membrane devices contain-
27: 2489–2496.
Huckins, J.N.; Prest, H.F.; Petty, J.D.; Røe, T.I.; Meadows, J.C.; Echols, K.R.; Lebo, J.A.; Clark, R.C.
1998a, A Overview of the Results of Several Comparisons of Lipid-containing Semipermeable
Membrane Devices (SPMDs) and Biomonitoring Organisms for Assessing Organic Chemical
Exposure. Abstracts of the 19th Annual National Meeting; SETAC; Charlotte, NC; November
15–19, 1998; p. 215.
Huckins, J.N.; Petty, J.D.; Gale, R.W.; Booij, K.; Prest, H.F.; Clark, R.C. 1998b, Observations on
declining SPMD sampling rates for high K ow compounds: Abstracts of the 19th Annual National
Meeting; SETAC; Charlotte, NC; November 15–19, 1998; pp. 241.
Technol. 33: 3918–3923.
Huebner, J.D. and Pynnönen, K.S. 1992, Viability of glochida of two species of Anodonta exposed to
low pH and selected metals. Can. J. Zool. 70: 2348–2354.
Hwang, S-T. and Kammermeyer, K. 1984, Membranes in Separations. Krieger Publishing: Malabar,
FL; pp. 559.
Jimenez, B.D.; Cirmo, C.P.; McCarthy, J. F. 1987, Effects of feeding and temperature on uptake,
elimination and metabolism of benzo[a]pyrene in the bluegill sunfish (Lepomis macrochirus).
Aquat. Toxicol. 10: 41–57.
Kolok, A.S.; Huckins, J.N.; Petty, J.D.; Oris, J.T. 1996, The role of water ventilation and sediment
ingestion in the uptake of benzo[a]pyrene in gizzard shad (Dorosoma cepedianum). Environ.
Toxicol. Chem. 10: 1752–1759.
Leppanen, H. and Kukkonen, V.K. 2000, Effect of sediment-chemical contact time on availability
of sediment-associated pyrene and benzo[a]pyrene to oligochaete worms and semipermeable
membrane devices. Aquat. Toxicol. 49: 227–241.
Lu, Y.; Wang, Z.; Huckins, J.N. 2002, Review of the background of triolein-containing semipermeable
membrane devices in aquatic environmental study. Aquat. Toxicol. 60: 139–153.
Macdonald, R.; Mackay, D.; Hickie, B. 2002, Contaminant amplification in the environment. Environ.
Sci. Technol. 36: 456A–462A.
FL; Vol. V.
Mayer, F.L. 1976, Residue dynamics of di-2-ethylhexylphthalate in fathead minnows (Pimephales
promelas). J. Fish. Res. Board Can. 33: 2610–2613.
McCarthy, K.A. and Gale, R.W. 2001, Evaluation of persistent hydrophobic organic compounds in the
Columbia River Basin using semipermeable membrane devices. Hydrol. Process. 15: 1271–1283.
McKim, J.; Schmieder, P.; Veith, G. 1985, Absorption dynamics of organic chemical transport across
trout gills as related to octanol-water partition coefficient. Toxicol. Appl. Pharm. 77: 1–10.
Meadows, J.C. 2005, USGS Columbia Environmental Research Center, Columbia, MO. USA. Personal
communication.
Opperhuizen, A.; van de Velde, E.W.; Gobas, F.A.P.C.; Liem, A.K.D.; van de Steen, J.M.D.; Hutzinger,
O.; 1985, Relationship between bioconcentration in fish and steric factors of hydrophobic chem-
icals. Chemosphere 14: 1871–1896.
Peterman, P.H. 2005, USGS Columbia Environmental Research Center, Columbia, MO. USA. Personal
communication.
166 Chapter 7
Peven, C.S.; Uhler, A.D.; Querzoli, F.J. 1996, Caged mussels and semipermeable membrane devices
as indicators of organic contaminant uptake in Dorchester and Duxbury Bays, Massachusetts.
Phillips, D.J.H. 1980, Quantitative Aquatic Biological Indicators. Applied Science Publishers: London,
UK.
Phillips, D.J.H. and Rainbow, P.S. 1993, Biomonitoring of Trace Aquatic Contaminants. Elsevier
Applied Science: New York; p. 371.
Prest, H.F.; Huckins, J.N.; Petty, J.D.; Herve, S.; Paasivirta, J.; Heinonen, P. 1995a, A survey of recent
results in passive sampling of water and air by semipermeable membrane devices. Mar. Pollut.
Bull. 31: 306–312.
Prest, H.F.; Richardson, B.J.; Jacobson, L.A.; Vedder, J.; Martin, M. 1995b, Monitoring organochlorines
with semipermeable membrane devices (SPMDs) and mussels (Mytilus edulis) in Corio Bay,
Victoria, Australia. Mar. Pollut. Bull. 30: 543–554.
Prest, H.F.; Hodgins, M.M.; Petty, J.D.; Huckins, J.N.; Lebo, J.A.; Orazio, C.E.; Gibson, V. 1997,
Laboratory Comparisons of Oysters and Semipermeable Membrane Devices (SPMDs). USGS
Midwest Science Center, Columbia, MO; Report to the American Petroleum Institute (API):
Washington; DC; API Report No. DR 360.
Pruell, R.J.; Lake, J.L.; Davis, W.R.; Quinn, J.G. 1986, Uptake and depuration of organic contaminants
by blue mussel (Mytilus edulis) exposed to environmentally contaminated sediment. Mar. Biol.
91: 497–507.
Rand, G.M. and Petrocelli, S.R., Eds. 1985, Fundamentals of Aquatic Toxicology, Methods and Appli-
cations; Hemisphere Publishing: Washington, DC.
Randall, R.C.; Lee, H. II; Ozretich, R.J.; Lake, J.L.; Pruell, J. 1991, Evaluation of selected lipid methods
for normalizing pollutant bioaccumation. Environ. Toxicol. Chem. 10: 1431–1436.
Richardson, B.J.; Zheng, G.J.; Tse, E.S.C.; De Luca-Abbott, S.B.; Siu, S.Y.M, Lam, P.K.S. 2003, A
comparison of polycyclic aromatic hydrocarbon and petroleum hydrocarbon uptake by mussels
(Perna viridis) and semi-permeable membrane devices (SPMDs) in Hong Kong Coastal Waters.
Environ. Pollut. 122: 223–227.
Sabaliūnas, D.; Lazutka, J.; Sabaliūniene, I.; Södergren, A. 1998, Use of semipermeable membrane de-
vices for studying effects of organic pollutants: Comparison of pesticide uptake by semipermeable
membrane devices and mussels. Environ. Toxicol. Chem. 17: 1815–1824.
Schmidt, C. 2004, USGS Columbia Environmental Research Center, Columbia, MO. USA. Personal
communication.
Schneider, R. 1982, Polychlorinated biphenyls (PCBs) in cod tissues from the Western Baltic: Sig-
nificance of equilibrium partitioning and lipid composition in the bioaccumulation of lipophilic
pollutants in gill-breathing animals. Meeresforschung 29: 69–79.
Schüürmann, G. and Klein, W. 1988, Advances in bioconcentration prediction. Chemosphere 17:
1551–1574.
Sericano, J.L.; Wade, T.L.; Brooks, J.M. 1996, Accumulation and depuration of organic contaminants
by the American oyster (Crassostrea virginica). Sci. Total Environ. 179: 149–160.
Södergren, A. 1987, Solvent filled dialysis membranes simulate uptake of pollutants by aquatic organ-
isms. Environ. Sci. Technol. 21: 855–859.
Spacie, A. and Hamelink, J.L. 1985, Bioaccumulation. In Fundamentals of Aquatic Toxicology. Rand,
G.M. and Petrocelli, S.R., Eds.; Hemisphere Publishing: Washington; pp. 495–525.
Stow, C.A. 1995, Factors associated with PCB concentrations in Lake Michigan salmonids. Environ.
Sci. Technol. 29: 522–527.
Thomann, R.V.; Connolly, J.P.; Parkerton, T.F. 1992, Modeling Accumulation of Organic Chemicals
in Aquatic Food Webs. In Chemical Dynamics in Freshwater Ecosystems. Gobas F.A.P.C., Mc-
Corquodale, J.A., Eds.; Lewis Publishers: Boca Raton; FL.; pp. 153–185.
Toro, B.; Navarro, J.M.; Palma-Fleming, H. 2003, Relationship between bioenergetics responses and
organic pollutants in the giant mussel; Chormytilus chorus (Mollusca: Mytilidae). Aquat. Toxicol.
63: 257–269.
Tse, E.; Richardson, B.; Lam, P. 2000, Uptake and release kinetics of selected organochlorines and PAHs
by mussels and SPMDs. Presented at the 6th International SPMD Workshop and Symposium; July
25-27th, 2000; USGS Columbia Environmental Research Center, Columbia, MO.
Utvik, T.I.R.; Durell, G.S.; Johnsen, S. 1999, Determining produced water originating polycyclic
aromatic hydrocarbons in North Sea waters: Comparison of sampling techniques. Mar. Pollut.
Bull. 38: 977–989.
Utvik, T.I.R. and Johnsen, S. 1999, Bioavailability of polycyclic aromatic hydrocarbons in the North
Sea. Environ. Sci. Technol. 33: 1963–1969.
Veith, G.D.; Macek, K.J.; Petrocelli, S.R.; Carroll, J. 1979, An evaluation of using partition coefficients
and water solubility to estimate bioconcentration factors for organic chemicals in fish. Fed. Regist.
44: 15926–15981.
bioavailable PAH, PCB, and OCP concentrations in water using semipermeable membrane devices
(SPMDs), sediments and caged carp. Chemosphere 11: 1675–1689.
Wang, W-X. and Fisher, N.S. 1999, Assimilation efficiencies of chemical contaminants in aquatic
invertebrates: A synthesis. Environ. Toxicol. Chem. 18: 2034–2045.
Wang, Y.; Wang, C.; Wang, Z. 1998, Uptake of moderately hydrophobic chlorophenols from water
by semipermeable membrane devices (SPMDs) and by goldfish (Crassius auratus). Chemosphere
37: 327–339.
Wang, Z.; Wang, Y.; Ma, M.; Lu, Y.; Huckins, J.N. 2002, Use of triolein-SPMD to assess the biocon-
centration processes of hydrophobic organic contaminants in the Huaihe River, China. Environ.
Toxicol. Chem. 21: 2378–2384.
Chapter 8
Selected Case Studies
8.1. REVIEW OF SPMD APPLICATIONS
In the previous chapters, we have shown that SPMD technology is useful for
a variety of applications in surface water, groundwater, sediments, and air. More
specifically, documented applications of SPMD technology include: 1) determi-
nation of the presence, sources, and the transport and fate of hydrophobic organic
chemicals (HOCs); 2) estimation of ambient time-weighted average (TWA) con-
centrations of HOC solutes or vapors; 3) determination of the fluxes of bioavailable
(i.e., dissolved phase) residues in aquatic systems; 4) in situ biomimetic con-
centration of bioavailable chemicals for screening with bioindicator tests, other
bioassays, and immunoassays, and for toxicity identification evaluation investi-
gations; 5) estimation of waterborne chemical exposure to aquatic organisms and
the bioconcentration potential of residues; and 6) the enrichment of HOCs in lipid
extracts. In this chapter we highlight a number of case studies. These studies do
not encompass all of the applications listed above but are intended to provide the
reader with a view of several unique attributes of SPMDs as well as to show how
they complement existing technology.
8.2. AIR-WATER-MICROLAYER EQUILIBRIUM
SPMDs are powerful tools to assess the degree of equilibrium between envi-
ronmental compartments. Booij and van Drooge (2001) exposed SPMDs to water
169
170 Chapter 8
and air at a coastal site during winter. In addition, SPMDs were exposed to the
sea surface microlayer (SSM) using an exposure device that caused the SPMDs
to intermittently emerge from the water and submerge to a depth of 0.5 cm below
the surface.
The SSM conventionally is sampled using rotating drums, glass plates, or
metal mesh screens (Hardy et al., 1988; Hardy and Cleary, 1992; Garabetian
et al., 1993) which sample the water surface in the sub-millimeter range. With
these samplers, a comparison of concentrations in the SSM and in deeper wa-
ter is often difficult, because substantial corrections for dissolved organic carbon
(DOC)-bound contaminants have to be applied, particularly for the SSM, where
DOC concentrations are often elevated. The application of SPMDs for sampling
the SSM circumvents the DOC correction problem, although at the cost of re-
duced depth resolution. For the coastal site studied by Booij and van Drooge
(2001), the ratios of dissolved polychlorinated biphenyls (PCBs) concentrations
between SSM and deeper water layers were 0.8 to 0.9 on average during two
time periods, indicating a minimal departure from equilibrium between bulk water
and SSM.
The authors also assessed the degree of equilibrium between atmospheric and
aqueous PCBs and hexachlorobenzene (HCB). The ratio of absorbed amounts for
SPMDs exposed to air (Na ) and water (Nw ) is given by

Na Ca 1 − exp − kea t
= (8.1)
Nw Cw K aw 1 − exp − kew t
where Ca , Cw are the contaminant concentrations in air and water, kea , kew are the
exchange rate constants for SPMD-air and SPMD-water, and K aw is the air-water
partition coefficient (volume per volume units). This equation has an interest-
ing feature that is worth noting. Because K aw Cw equals the equilibrium atmo-
spheric concentration, the first term at the right-hand side of Eq. 8.1 represents
the relative departure from equilibrium. When this term is larger than 1, the air
is supersaturated relative to the water phase. Using the dissipation of three PRCs,
Booij and Van Drooge (2001) showed that there was reason to believe that kea
and kew were about the same in their study. The authors concluded that the ra-
tio of the amounts in air-exposed and water-exposed SPMDs was equal to the
term Ca /(Cw K aw ), i.e. equal to the relative departure of equilibrium. Their results
indicated that a fair degree of air-water equilibrium existed for PCBs, and that
the atmosphere was oversaturated with HCB by a factor of 6 to 9, relative to the
water phase (Figure 8.1). In this work, the use of SPMDs allowed the authors to
ignore the correction for DOC-bound chemicals and permitted the determination
of the relative departure (if any) from equilibrium of target compound concentra-
tions in the aqueous and atmospheric boundary layers. This type of data is very
useful for modeling the flux of chemicals between the aqueous and atmospheric
phases.
Selected Case Studies 171
FIGURE 8.1 Ratio of amounts absorbed by air-exposed and water-exposed SPMDs at a Dutch
coastal site. Reprinted from Booij and van Drooge (2001), copyright (2001); reproduced with
permission from Elsevier.
8.3. SAMPLING INDOOR AIR
In a joint project between the US Environmental Protection Agency (EPA)

and the US Geological Survey (USGS), SPMDs were deployed at 57 sites indoors
along the border between Arizona (USA) and Mexico (Petty et al., 2002). The
exposure period was 30 d and each sample represented a composite of four 1-mL
triolein SPMDs. Numerous HOC vapors of historic concern were found in ex-
posed SPMDs, including organochlorine pesticides (OCPs), PCBs, and polycyclic
aromatic hydrocarbons (PAHs). In addition, a variety of contaminants of emerg-
ing concern, e.g., diazinon, chlorpyrifos, permethrin, etc., were present at many
sites. Figure 8.2 illustrates the very high amounts of some of these “current use”
pesticides accumulated in SPMD samples.
Also, Petty et al. (2002) performed an in depth analysis of the OCP fraction
of SPMD extracts by gas chromatography-mass spectrometry (GC-MS) resulting
in the tentative identification of about 400 airborne organic chemicals, which were
not present in SPMD field blanks. The OCP fraction represents only one of several
enriched fractions from SPMD samples. Table 8.1 summarizes the various classes
of compounds tentatively identified in SPMDs exposed to indoor air.
The estimated ambient vapor-phase concentrations of chemicals in indoor
air were, in general, below the NIOSH time-weighted average exposure limits
for a 40 hr workweek. However, the authors of this study concluded that the
consequences of long term human exposure to these complex mixtures of airborne
172 Chapter 8
TABLE 8.1 Classes of Compounds Tentatively Identified in SPMD Samples Exposed to

Indoor Air
Polycyclic aromatic compounds (PAHs) (28)a Phthalate esters (5)

benzofluoranthene diethylhexyl phthalate
chrysene dibutyl phthalate
benz[a]anthracene benzylbutyl phthalate
fluoranthene Organochlorine pesticides (27)
phenanthrene chlordanes
fluorene chlordenes
pyrene chlorpyrifos
p,p -DDT
Alkylated PAHs (19)
C1–C5 naphthalenes Fragrance components (20)
C1–C4 phenanthrenes galoxolide
polycyclic musk
C1–C2 pyrenes (9) hexyl cinnamaldehyde
Alkyl biphenyls manoyl oxide
C1–C4 biphenyls menthol
musk xylene
Alkyl benzenes (5) jasmal
lilial
Alkyl terphenyls (3)
Hydrocarbons (10) Terpenoids (30)

C8–C22 alkanes calamenene
terpineol
Cyclic non-aromatic hydrocarbons (9) cedrol
C2–C3 tetrahydronaphthalenes isomethyl ionone
propyl cyclopentane xanthene
Alkyl aldehydes (10) Phenolics (6)

C9–C13 alkyl and alkenyl aldehydes bis(dimethylbenzyl) phenol
bis(dimethylbenzyl)tertiarybutyl phenol
Alkyl ketones (9)
C8–C16 alkyl and alkenyl ketones Phosphates
triphenyl phosphate
Alkyl alcohols (15) cresyl diphenyl phosphate
C8–C22 alkyl and alkenyl alcohols
Unidentified compounds (120)
Alkyl acids and esters (16) possible terpenoid isomers and
alkyl propanoates fragrance-related components
Sterols (5)
cholesterol
sitosterol
stigmasterol, fucosterol, campesterol
a
Parenthetical values correspond to the number of compounds identified in a chemical class.
FIGURE 8.2 Amounts of current use pesticides in SPMD samples at selected sites.
chemicals are unknown, especially with respect to gender and age related vari-
ables. A major concern is the possibility of additive effects arising from prolonged
respiratory exposure. In this study, SPMDs provided a convenient approach for
determining the presence of several recognized vapor-phase contaminants and a
means of defining the presence of a large number of previously unidentified air-
borne chemicals.
8.4. FURTHER COMPARISONS OF BIVALVES AND SPMDs
Bivalves are widely used in monitoring programs to assess the waterborne

contaminant exposure. These organisms are used because they are immobile,
widely available, have very low levels of certain enzyme systems known to mediate
the metabolism of many contaminants, and ventilate large volumes of water when
feeding (see Chapter 7). However, mounting evidence suggests that the levels of
some residues in bivalve tissues often are not proportional to ambient chemical
concentrations. This lack of proportionality may result in the inability to reliably
extrapolate ambient exposure concentrations, insufficient residue concentration
factors for the detection or quantification of target compounds, and higher concen-
trations of high K ow compounds than predicted by equilibrium partition theory. In
combination, these difficulties can lead to the inability to detect the presence of
target compounds and to discern concentration gradients or chemical sources.
174 Chapter 8
FIGURE 8.3 GC-MS comparison of ion chromatograms of extracts from mussels (Mytilus edulis)
and SPMDs, Corio Bay Australia. Peaks labeled ISTD are internal standards. Reprinted with
permission from the American Petroleum Institute, copyright 2002 (Huckins et al., 2002).
The following example illustrates one aspect of this potential problem. Prest
et al. (1995) deployed blue mussels (Mytilus edulis) and SPMDs contiguously
at several sites, including sites near a refinery effluent, in Corio Bay, Victoria,
Australia. This 60 d exposure study was designed to examine the relative abilities
of BMOs and SPMDs to monitor a known gradient of chlorinated contaminants.
Overall, the levels of chlorinated organic chemicals were about the same in both
sample types. However, the GC-MS ion chromatograms from the two matrices
differed markedly, as shown in Figure 8.3.
Analysis of these SPMD samples suggested that lower chlorinated PCBs and
a complex mixture of unknowns (early eluting, relatively low K ow components)
were present at high concentrations in the water column, while data from the mussel
samples implied essentially the reverse. These results are not surprising for several
reasons. Unlike SPMDs, concentrations of some hydrophobic organic contami-
nants in bivalves are not necessarily proportional to ambient water concentrations
(Huckins et al., 2004). Also, early eluting (GC) chlorinated organics are much more
soluble in water than late eluting compounds and would be expected to be present
at higher concentrations in aquatic environments than the higher molecular weight
(later eluting), very hydrophobic components. However, models successfully used
for estimating bioconcentration in aquatic organisms exhibit an inverse relation-
ship between the lipid content of tissues and the related elimination rate constant
(ke ). Large ke s mean that time to equilibrium is short and thus the volume of water
cleared of chemicals is relatively small. The lipid content of SPMDs is at least an
order-of-magnitude greater than bivalves, while the uptake rates for low K ow com-
pounds are similar (Huckins et al., 2004). Thus, bivalves are expected to accumulate
much lower amounts of these compounds than SPMDs. In fact, BCFs of organ-
isms with low lipid contents may be inadequate for the detection or quantification
of trace amounts of these early eluting compounds. The Prest et al. (1995) study
illustrates one other significant difference between BMOs and SPMDs. One of the
study sites was in a refinery effluent stream where bivalves could not survive due
to elevated temperature and turbidity. However, Prest et al. (1995) concluded that
overall bivalves and SPMDs provide complimentary information, such as the abil-
ity to better determine the relative roles of respiratory and dietary routes of uptake.
In another bivalve study, SPMDs and Asian clams, Corbicula fluminea were
deployed at stream sites in the Dallas-Fort Worth Metropolitan Area (Moring and
Rose, 1997) to assess the presence and concentrations of bioavailable, dissolved
PAHs. Deployment sites were White Rock Creek, West Fork Trinity River and
Trinity River below Dallas, and the exposure periods ranged between 30 and 35 d.
Bivalves were depurated for 1 d before being processed for analysis. Twenty-four
PAHs were measured in SPMDs, 20 of which occurred at all sites and only three
PAHs were detected in the co-deployed clams (Figure 8.4). Throughout all sites,
non-alkylated PAHs were found at greater levels in SPMDs than the alkylated
forms. Nine out of the 16 EPA priority pollutant PAHs were detected in SPMDs.
In several cases (i.e., benz[a]anthracene, benzo[a]pyrene, and chrysene), estimated
concentrations in water exceeded the EPA’s human health criteria. This example
illustrates that the occurrence of potentially toxicologically significant residues of
PAHs may not be detected when using BMOs in some environments. It seems
unlikely that the depuration of gut contents prior to analysis would have been
responsible for the small number of PAHs detected in these samples. A much
more likely explanation is chemical stressor induced valve closure, limiting water
exchange only to that necessary for survival (e.g., Goudreau et al., 1993; Hellou
et al., 2004; Huckins et al., 2004).
8.5. ESTIMATION OF EXPOSURE TO DIOXIN-LIKE

COMPOUNDS USING SEDIMENTS, CAGED FISH, AND SPMDs
Gale et al. (1997) conducted a study in the Saginaw River, MI, USA to com-
pare the SPMD method of water sampling with sediment-based and caged fish
based methods for assessing exposure to dioxin-like compounds. More specif-
ically, the targeted compounds were planar halogenated hydrocarbons (PHHs),
which consisted of planar PCBs, polychlorinated dibenzo- p-dioxins (PCDDs),
and polychlorinated dibenzofurans (PCDFs). The list of target compounds was
expanded to include polychlorinated dibenzothiophenes (PCDTs) after finding
PCDT residues at all sites. The log K ow s of target compounds ranged from 6.1 to
8.2. Fish and SPMDs were exposed for 28 d at five sites and sediment samples
(0–10 cm depth) were collected at each site. SPMDs used were of standard design
176 Chapter 8
FIGURE 8.4 Comparison of PAHs detected by GC-PID in SPMDs and clams (Corbicula fuminea)
deployed in the Trinity River, Dallas,Texas, USA. Reprinted from Moring and Rose (1997),
copyright (1997); reproduced with permission from Elsevier.
(see Chapter 4) but 152 cm in length and the fish were juvenile hatchery-reared
channel catfish (Ictalurus punctatus; 8–10 cm).
A number of toxicologically active 2,3,7,8-substituted PCDDs and PCDFs,
planar PCBs and PCDD, PCDF, and PCDT homologs were measured in fish, SP-
MDs, and sediments. Only two target compounds exceeded the detection limits
of 0.2–1 pg g−1 in SPMD field blanks (see definition in Chapter 5). These ex-
ceptions were octachlorodibenzo- p-dioxin (OCDD) and octachlorodibenzofuran
(OCDF) which were present in SPMDs at about 5 pg g−1 . However, negative
control fish contained significant amounts of 1,2,3,4,6,7,8-heptachlorodibenzo-

p-dioxin (≈ 5 pg g−1 ), OCDD (≈ 50 pg g−1 ) and just quantifiable amounts of
Cl4−6 PCDDs (≈1 pg g−1 ).
Analysis of the three test matrices revealed that congener patterns representing
each matrix were constant and distinctive, regardless of sampling site. In general,
only the analyte concentrations varied among sites (Figure 8.5). Figure 8.5 (site 5,
downstream from Bay City) also shows that patterns of congeners in each homolog
group were similar for SPMDs and sediments. More specifically, the number of
congeners found and their relative ratios in each homolog group (Cl3 –Cl7 ) corre-
sponded well between SPMDs and sediments. The patterns of PCDDs and PCDFs
in fish were different from those of either SPMDs or sediments but PCDT patterns
showed very little difference among the three matrices (Figure 8.5). The apparent
metabolism of PCDDs and PCDFs by channel catfish appeared to be the causal
factor in the distinct difference between the residue patterns in the fish and SPMDs
(Figure 8.5). Although not shown in Figure 8.5, Gale et al. (1997) found that the
coplanar PCB congener 77 or 3,3 ,4,4 -tetrachlorobiphenyl was about an order of
magnitude higher in SPMDs than in channel catfish, whereas, SPMD and channel
catfish concentrations of coplanar PCB congener 81 or 3,4,4 ,5-tetrachlorobipheny
were within two fold of each other. Structure-specific metabolism appears to be
the reason for the difference in the concentrations of the two congeners. Because
coplanar PCB congeners are present in technical PCB mixtures at low levels and
because of the apparent high specificity of the degradation pathway, the Gale et al.
(1997) data does not invalidate the good correlation shown between SPMDs and
fish in Figure 7.6.
8.6. USING SPMDs IN CONJUNCTION WITH BIOASSAYS AND

CHEMICAL ANALYSIS TO CHARACTERIZE TOXIC POLLUTANTS
Lake Shkodra/Skadar is the largest lake in the Balkan region and is located
on the border between Albania and Montenegro. Due to the broad array of rare
or endangered plant and animal species it supports, the lake and the extensive
associated wetlands have been designated a Ramsar site. The lake is also the
focus of an international cooperative investigation by a diverse team of researchers
from Albania, Serbia and Montenegro, Germany, Austria and the UK, concerned
with the possible detrimental effects of increasing anthropogenic contaminants.
From the onset of these investigations, several members of the multinational team
expressed concern over the long-term risks posed by increased loads of HOCs on
the lake’s aquatic biota. However, limited access to suitable facilities and equipment
in the Balkan region, and logistical problems associated with the transportation
of large quantities of water for analysis delayed investigations into the identity
and distribution of HOCs in the lake. These problems were overcome by Rastall
et al. (2004a) through the application of SPMD sampling technology. SPMDs
were transported to the region, deployed by local workers with knowledge of
FIGURE 8.5 Ion traces of PCDDs (set A, Cl4 –Cl8 ), PCDFs (set B, Cl4 –Cl8 ), and PCDTs (set C, Cl3 –
Cl7 ) in caged channel catfish, SPMDs, and sediment from site 5, downstream from Bay City, MI,
USA. The matrix identity is given in the upper left of the first chromatogram of the homolog series
and the maximum peak height is given in the upper right of each chromatogram for comparison
of peak heights among contaminant homologues and matrices. Reprinted from Gale et al. (1997),
copyright (1997); reproduced with permission from American Chemical Society.
FIGURE 8.6 Induction of significant activity in the ethoxyresorufin-O-deethylase (EROD) screen

by an extract of an SPMD sample from Lake Shkodra/Skadar. Reprinted from Rastall et al. (2004a),
copyright (2004); reproduced with permission from Environmental Science and Pollution
Research.
anthropogenic inputs and influences in the lake, and then returned to the laboratory
for analysis. Moreover, SPMDs enabled sequestration of sufficient amounts of
chemicals for multiple rounds of bioassay and chemical analysis.
Analysis of the SPMD sample extracts from five of the six sampling sites
in this study revealed the presence of a wide variety of waterborne contaminants,
including EPA priority pollutants PAHs and their alkylated analogues, atrazine
and other agricultural chemicals, and a variety of sterols and sterol derivatives.
A total of 39 HOCs were tentatively identified in SPMD extracts but numerous
compounds remain unidentified. Extracts from all but one sampling site were
found to induce significant concentration dependent ethoxyresorufin-O-deethylase
(EROD) activity, which is a biomarker of exposure to dioxin-like compounds
(Figure 8.6; also see Section 6.3.). Although PAHs were present at relatively high
levels in these samples, it appears that EPA priority pollutants represented only
0.06% of the total EROD-inducing potential. Similarly, five of the six sample sites
induced significant concentration-dependent activity in the yeast estrogen screen
(YES) assay (Figure 8.7; see Section 6.5.). Also, the results of the YES assay from
two sites were indicative of an inhibitory or toxic response, i.e., the yeast cells
either did not grow or growth was delayed over that observed with controls. No
significant activity was observed in any of the blanks or control sample extracts.
These data indicate that hydrophobic chemicals with estrogenic and EROD-
inducing potential are widespread in this ecosystem and are readily bioavailable
to aquatic organisms living in the system. By incorporating the YES, the recently
developed yeast androgen screen, and the EROD assays (Rastall, 2004b) into
exposure assessment approaches, a more complete picture of the potential toxicity
180 Chapter 8
FIGURE 8.7 Induction of significant activity in yeast estrogen screen (YES) assay by an extract
of an SPMD sample from Lake Shkodra/Skadar. Reprinted from Rastall et al. (2004a), copyright
(2004); reproduced with permission from Environmental Science and Pollution Research.
of readily bioavailable HOCs is available. Furthermore, as anthropogenic impacts

on Lake Shkodra-Skadar increase in the future, SPMD-based sampling is expected
to play a central role in surveillance and investigative monitoring, as well as to
facilitate research into the effects of HOCs on Lake Shkodra-Skadar’s aquatic biota.
8.7. REFERENCES
and semipermeable membrane devices in the Saginaw River, Michigan. Environ. Sci. Technol. 32:
178–187.
Garabetian, F.; Romano, J.-C.; Paul, R.; Sigoillot, J.-C. 1993, Organic matter composition and pollutant
enrichment of sea surface microlayer inside and outside slicks. Mar. Environ. Res. 35: 323–339.
Goudreau, S.E.; Neves, R.J.; Sheehan, R. 1993, Effects of wastewater treatment plant effluents of
freshwater mollusks in the upper Clinch River, Virginia, USA. Hydrobiologia 252: 211–230.
Hardy, J.T.; Coley, J.A.; Antrim, L.D.; Kiesser, S.L. 1988, A hydrophobic large-volume sampler for
collecting aquatic surface microlayers: Characterization and comparison with the glass plate
method. Can. J. Fish. Aquat. Sci. 45: 822–826.
Hardy, J.T. and Cleary, J. 1992, Surface microlayer contamination and toxicity in the German Bight.
Mar. Ecol. Prog. Ser. 91: 203–210.
Hellou, J.; Steller, S.; Langille, M.A.; Leonard, J.; Tremblay, D. 2004, Partitioning of polycyclic
aromatic hydrocarbons between water and particles and bioaccumulation in mussels: A harbour
case. Mar. Environ. Res. 59: 101–117.
Petty, J.D.; Huckins, J.N.; Robertson, G.L.; Cranor, W.L.; Gale, R.W.; Alvarez, D.A.; Clark, R.C.
2002, The application of semipermeable membrane devices (SPMDs) as samplers of airborne
contaminants in indoor air. Presented at the 53rd Pittsburgh Conference on Analytical Chemistry
and Applied Spectroscopy, March 16–22, 2002; New Orleans, LA.
Prest, H.F.; Richardson, B.J.; Jacobson, L.A.; Vedder, J.; Martin, M. 1995, Monitoring organochlorines
Rastall, A.C.; Neziri, A.; Vukovic, Z.; Jung, C.; Mijovic, S.; Hollert, H.; Nikcevic, S.; Erdinger, L. 2004a,
The identification of readily bioavailable pollutants in Lake Shkodra/Skadar using semipermeable
membrane devices (SPMDs), bioassays and chemical analysis. Environ. Sci. Pollut. R. 11: 240–
253.
Rastall, A.C. 2004b, University of Heidelberg, Germany; Personal communication.
Appendix A
SPMD Calibration Data
A.1. SPMD-WATER PARTITION COEFFICIENTS
Literature values of SPMD-water partition coefficients (K sw s) should be used

with caution. Different units of K sw are generated when the concentrations in
SPMD and water are expressed on a mass basis (e.g., ng g−1 ) or a volume basis
(e.g., ng mL−1 ). Depending on the choice of concentration units, K sw values will
have units of g g−1 , mL mL−1 , or mL g−1 , or K sw values are given as dimensionless
numbers in the former two cases. Care should be taken to distinguish between the
different versions of K sw . Preferably, mL mL−1 units should be used, because most
of the equations for SPMD uptake kinetics use the volume of an SPMD rather than
its mass. The two most frequently used versions of K sw are interrelated by
K sw (mL mL−1 units) = K sw (mLg−1 units) · ρ s (A.1)
where ρs is the SPMD density (0.91 g mL−1 ). Hence the difference in the numerical
values of the two K sw definitions amounts to a factor of 1/0.91 = 1.10, which
corresponds to 0.04 log units. This value is low, when compared to the magnitude
of errors associated with environmental-analytical chemistry, but any data scatter
originating from incompatible units should be avoided.
In Tables A.1 through A.3, all K sw values were converted to mL mL−1 units
where applicable. In the literature sources where the units were not explicitly
mentioned, these were inferred from the model equations used. To avoid possible
ambiguity in the future, authors are urged to specify the units of their reported K sw
values.
183
184 Appendix A
TABLE A.1 SPMD-Water Partition Coefficients (Ksw s, mL mL−1 ) of Mono-Aromatic

and Polycyclic Aromatic Hydrocarbons
Compound M log K ow a log K sw ◦C ref. b
benzene 78.1 2.13 1.78 25 1

toluene 92.1 2.69 2.34 25 1
ethylbenzene 106.2 3.13 2.80 25 1
naphthalene 128.2 3.37 3.20 17 5
naphthalene 128.2 3.37 3.36 24 2
1-methylnaphthalene 142.2 3.87 4.20 17 5
acenaphthylene 152.2 4.00 4.20 17 5
acenaphthylene 152.2 4.00 3.63 24 2
acenaphthene 154.2 3.92 4.00 2 4
acenaphthene 154.2 3.92 4.13 13 4
acenaphthene 154.2 3.92 4.50 17 5
acenaphthene 154.2 3.92 4.05 24 2
acenaphthene 154.2 3.92 3.96 30 4
biphenyl 154.2 3.90 4.30 17 5
1,3-dimethylnaphthalene 156.2 4.42 5.10 17 5
fluorene 166.2 4.18 4.80 17 5
fluorene 166.2 4.18 4.21 24 2
anthracene 178.2 4.54 4.67 24 2
phenanthrene 178.2 4.57 4.49 2 4
phenanthrene 178.2 4.57 4.70 8 3
phenanthrene 178.2 4.57 4.66 13 4
phenanthrene 178.2 4.57 4.53 18 3
phenanthrene 178.2 4.57 4.47 24 2
phenanthrene 178.2 4.57 4.29 30 3
phenanthrene 178.2 4.57 4.59 30 4
fluoranthene 202.3 5.22 5.22 2 4
fluoranthene 202.3 5.22 5.22 13 4
fluoranthene 202.3 5.22 4.68 24 2
fluoranthene 202.3 5.22 5.00 30 4
pyrene 202.3 5.18 5.28 2 4
pyrene 202.3 5.18 5.24 13 4
pyrene 202.3 5.18 4.79 24 2
pyrene 202.3 5.18 5.09 30 4
benzo[a]anthracene 228.3 5.91 5.45 2 4
benzo[a]anthracene 228.3 5.91 5.46 30 4
chrysene 228.3 5.86 5.52 2 4
chrysene 228.3 5.86 5.51 30 4
a
Adopted from the original reference, from Mackay et al. (1992a), or from US EPA (2003), in order of availability.
b
1. Calculated from LDPE-water partition coefficients (Reynolds et al., 1990) and triolein-water partition coeffi-
cients (Chiou, 1985); 2. Huckins et al. (1999); 3. Huckins et al. (2002a); 4. Booij et al. (2003); 5. Huckins et al.
(2004).
SPMD Calibration Data 185
TABLE A.2 SPMD-Water Partition Coefficients (Ksw s, mL mL−1 ) of Chlorobenzenes

and PCBs
Compound M log K ow a log K sw ◦C ref. b
1,2,3,4-tetrachlorobenzene 215.9 4.50 4.36 2 3

pentachlorobenzene 250.3 5.00 5.00 30 3
PCB 1 188.7 4.30 4.18 25 1
PCB 4/5/8 223.1 5.13 4.65 25 1
PCB 28 257.5 5.80 5.41 2 3
PCB 28/29 257.5 5.70 5.14 25 1
PCB 28 257.5 5.80 5.50 30 3
PCB 52 292.0 6.10 5.35 2 3
PCB 52 292.0 6.10 5.55 8 2
PCB 52 292.0 6.10 5.66 18 2
PCB 52/47 292.0 6.00 5.35 25 1
PCB 52 292.0 6.10 5.53 30 2
PCB 52 292.0 6.10 5.47 30 3
PCB 98/101 326.4 6.36 5.44 25 1
PCB 153/154 360.9 6.97 5.86 25 1
a
Adopted from the original reference or from Mackay et al. (1992b), in order of availability.
b
1. Calculated from LDPE-water partition coefficients (Lefkovitz et al., 1996) and triolein-water partition coefficients
(Chiou, 1985); 2. Huckins et al. (2002a); 3. Booij et al. (2003).
A.2. WATER SAMPLING RATES
SPMD uptake kinetics are commonly reported in terms of sampling rates

(Rs s) or uptake rate constants (ku s). Care should be taken to identify the SPMD
design and exposure condition for which these parameters are given. The surface
area of the SPMD is one of the most important parameters, because Rs values are
linearly proportional to the surface area. For this reason, we urge authors to specify
the surface area of SPMDs along with Rs values. Sometimes, Rs values have been
normalized to the 460 cm2 surface area of the original (standard) SPMD design
(2.54 cm wide, 91.4 cm long LDPE layflat tubing with a wall thickness of 85 µm,
containing 1-mL triolein).
The accumulation of chemicals can also be expressed in terms of ku s. Unfor-
tunately, ku s are expressed in several different ways, contributing to uncertainty
in data comparability. The symbol “ku ” was initially used in the SPMD related
literature as a parameter that describes the equilibration rate constant (Huckins et
al., 1993), whereas in later work, the symbol “ke ” was adopted for this parameter
(Booij et al., 1998; Huckins et al., 1999). Depending on whether contaminant con-
centrations in SPMDs are expressed on a mass or a volume basis, ku s may have
units of or mL mL−1 d−1 (i.e., d−1 ) or mL g−1 d−1 . These different unit-forms of
ku are interrelated by
ku (d −1 units) = ku (mLg−1 d−1 units) · ρs (A.2)

186 Appendix A
TABLE A.3 SPMD-Water Partition Coefficients (Ksw s, mL mL−1 ) of Organochlorine

Pesticides
Compound M log K ow a log K sw ◦C ref.b
α-HCH 290.8 3.80 3.29 19 2

α-HCH 290.8 3.80 3.37 19 2
α-HCH 290.8 3.80 3.32 19 2
α-HCH 290.8 3.80 3.31 19 2
β-HCH 290.8 3.80 2.91 19 2
β-HCH 290.8 3.80 2.87 19 2
β-HCH 290.8 3.80 3.17 19 2
β-HCH 290.8 3.80 3.26 19 2
γ -HCH 290.8 3.70 3.01 19 2
γ -HCH 290.8 3.70 3.22 19 2
γ -HCH 290.8 3.70 3.33 19 2
γ -HCH 290.8 3.70 3.22 19 2
δ-HCH 290.8 4.10 3.09 19 2
δ-HCH 290.8 4.10 3.34 19 2
δ-HCH 290.8 4.10 3.56 19 2
p,p -DDE 318.0 6.14 5.92 8 3
p,p -DDE 318.0 6.14 6.04 18 3
p,p -DDE 318.0 6.14 5.93 30 3
trifluralin 335.3 5.10 4.50 c 1
chlorpyrifos 350.6 5.00 4.64 c 1
heptachlor 373.3 6.10 4.82 c 1
dieldrin 380.9 5.40 4.81 c 1
a
Adopted from the original reference or from US EPA (2003), in order of availability.
b
1. Sabaliūnas and Södergren (1997); 2. Vrana and Schüürmann (2002); 3. Huckins et al. (2002a).
c
Not specified.
It may be tempting to use SPMD mass or volume to correct ku values for

differences in SPMD design. However, this approach is only partly valid. For
example, by adding more triolein to a standard sized SPMD, the mass of the SPMD
is increased while the surface area remains constant. This change results in a de-
crease of the mass or volume based ku , but does not change the water sampling rate
(L d−1 ). Therefore, ku values can only be directly compared among studies when
SPMD surface-area-to-volume ratios (AV −1 ) are similar. Acknowledging that the
LDPE thickness for standard-design SPMDs may range between 70 and 95 µm
and that the triolein mass fraction should be kept at exactly 20%, standard-design
SPMDs may vary in AV −1 ratios between 84 and 114 cm2 cm−3 (cm−1 ), corre-
sponding to a difference of 1.4 fold. It is therefore important that authors clearly
specify the SPMD design for which ku values are reported, i.e., either specify the
surface area and the volume or mass, or the ratio between the two (e.g., cm−1 ).
In the Tables A.4 through A.12, Rs and ku values are listed for a large number
of compounds (one table per study). When necessary, values for 460 cm2 SPMDs
with a mass of 4.5 g (V = 4.9 cm3 ) were calculated from the reported literature
values. The Rs values that were derived from calculated aqueous concentrations
were excluded.
TABLE A.4 Water Sampling Rates (Rs s) and Uptake Rate Constants (Ku s) of 4.5 g,
460 cm2 SPMDs at a Temperature of 10, 18, and 26 ◦ C, and a Flow Rate of 0.004
cm s−1 (Huckins et al., 1999)
Compounds M (g mol−1 ) log K ow a VLebas (cm3 mol−1 ) R s (L d−1 ) ku (L g−1 d−1 )
0.004 cm s−1 , 10 ◦ C
naphthalene 128.2 3.45 147.6 1.9 0.42
acenaphthylene 152.2 4.08 165.7 2.3 0.51
acenaphthene 154.2 4.22 173.0 2.7 0.60
fluorene 166.2 4.38 188.0 3.0 0.67
anthracene 178.2 4.54 197.0 2.9 0.64
phenanthrene 178.2 4.46 199.0 3.8 0.84
fluoranthene 202.3 5.20 217.0 3.6 0.80
pyrene 202.3 5.30 214.0 4.5 1.00
benzo[a]anthracene 228.3 5.91 248.0 3.2 0.71
chrysene 228.3 5.61 251.0 3.7 0.82
benzo[a]pyrene 252.3 6.35 263.0 3.2 0.71
benzo[b]fluoranthene 252.3 5.78 268.9 2.8 0.62
benzo[k]fluoranthene 252.3 6.20 268.9 2.9 0.64
benzo[g,h,i]perylene 276.3 6.90 277.5 1.8 0.40
indeno[1,2,3-c,d]pyrene 276.3 6.51 283.5 2.0 0.44
dibenz[a,h]anthracene 278.4 6.75 300.0 3.0 0.67
0.004 cm s−1 , 18 ◦ C
acenaphthylene 152.2 4.08 165.7 1.4 0.31
acenaphthene 154.2 4.22 173.0 2.3 0.51
fluorene 166.2 4.38 188.0 1.7 0.38
anthracene 178.2 4.54 197.0 3.6 0.80
phenanthrene 178.2 4.46 199.0 3.6 0.80
fluoranthene 202.3 5.20 217.0 4.5 1.00
pyrene 202.3 5.30 214.0 5.2 1.16
chrysene 228.3 5.61 251.0 4.8 1.07
benzo[a]pyrene 252.3 6.35 263.0 3.7 0.82
indeno[1,2,3-c,d]pyrene 276.3 6.51 283.5 3.0 0.67
0.004 cm s−1 , 26 ◦ C
acenaphthylene 152.2 4.08 165.7 1.7 0.38
acenaphthene 154.2 4.22 173.0 2.4 0.53
fluorene 166.2 4.38 188.0 2.8 0.62
anthracene 178.2 4.54 197.0 4.6 1.02
phenanthrene 178.2 4.46 199.0 5.0 1.11
fluoranthene 202.3 5.20 217.0 6.8 1.51
pyrene 202.3 5.30 214.0 7.6 1.69
chrysene 228.3 5.61 251.0 7.6 1.69
benzo[a]pyrene 252.3 6.35 263.0 5.4 1.20
(Continued)
188 Appendix A
TABLE A.4 (Continued )
Compounds M (g mol−1 ) log K ow a VLebas (cm3 mol−1 ) R s (L d−1 ) ku (L g−1 d−1 )

indeno[1,2,3-c,d]pyrene 276.3 6.51 283.5 3.4 0.76
a
Adopted from Huckins et al. (1999).
TABLE A.5 Water Sampling Rates (R s s) and Uptake Rate Constants (ku s) of 4.5 g,
460 cm2 SPMDs at a Temperature of 25 ◦ C, and Flow Rates of 0.01 and 50 cm s−1
(Luellen and Shea, 2002)
Compound M(g mol−1 ) log K ow a VLebas (cm3 mol−1 ) R s (L d−1 ) ku (L g−1 d−1 )
0.01 cm s−1 , 25 ◦ C
naphthalene 128.2 3.37 147.6 3.0 0.68
C1-naphthalenes 142.2 3.86 169.8 5.4 1.20
biphenyl 154.2 3.90 184.6 4.4 0.97
C2-naphthalenes 156.2 4.37 192.0 7.3 1.62
fluorene 166.2 4.18 188.0 5.1 1.14
dibenzofuran 168.2 4.12 184.6 5.2 1.16
C3-naphthalenes 170.3 4.90 214.2 6.3 1.39
phenanthrene 178.2 4.46 199.0 4.5 0.99
C1-fluorenes 180.3 4.97 210.0 5.4 1.20
dibenzothiophene 184.3 4.38 191.3 3.0 0.66
C4-naphthalenes 184.3 5.30 236.4 4.3 0.96
C1-phenanthrenes/anthracenes 192.3 5.14 220.0 5.6 1.23
C2-fluorenes 194.3 5.20 232.0 5.5 1.22
C1-dibenzothiophene 198.3 4.80 213.5 3.4 0.74
fluoranthene 202.3 5.22 217.0 4.0 0.89
pyrene 202.3 5.18 214.0 5.8 1.28
C3-fluorenes 208.3 5.50 254.0 4.5 0.99
C1-fluoranthenes-pyrenes 216.3 5.70 237.5 3.9 0.87
chrysene 228.3 5.61 251.0 3.5 0.77
C1-chrysenes 242.3 6.20 273.2 1.7 0.38
benzo[e]pyrene 252.3 6.40 263.0 1.5 0.33
C2-chrysenes 256.4 6.50 295.4 1.0 0.22
C3-chrysenes 270.4 6.80 317.6 0.8 0.17
(Continued)

C4-chrysenes 284.4 7.10b 339.8 0.4 0.08
50 cm s−1 , 25 ◦ C
naphthalene 128.2 3.37 147.6 3.5 0.77
1-methyl naphthalene 142.2 3.86 169.8 5.0 1.12
C1-naphthalenes 142.2 3.86 169.8 5.1 1.14
acenaphthylene 152.2 4.07 173.0 3.7 0.81
acenaphthene 154.2 3.92 165.7 3.9 0.87
biphenyl 154.2 3.90 184.6 4.5 1.00
2,6-dimethyl naphthalene 156.2 4.37 192.0 5.7 1.26
C2-naphthalenes 156.2 4.37 192.0 4.8 1.06
50 cm s−1 , 25 ◦ C
fluorene 166.2 4.18 188.0 4.6 1.02
dibenzofuran 168.2 4.12 184.6 4.8 1.06
2,3,5-trimethyl naphthalene 170.3 4.90 214.2 5.9 1.32
C3-naphthalenes 170.3 4.90 214.2 5.7 1.27
anthracene 178.2 4.54 197.0 4.7 1.05
phenanthrene 178.2 4.46 199.0 5.0 1.10
1-methyl fluorene 180.3 4.97 210.0 5.6 1.24
C1-fluorenes 180.3 4.97 210.0 5.8 1.29
C4-naphthalenes 184.3 5.30 236.4 4.5 0.99
1-methyl phenanthrene 192.3 5.14 218.7 5.6 1.25
C2-fluorenes 194.3 5.20 232.0 6.0 1.34
fluoranthene 202.3 5.22 217.0 5.5 1.22
pyrene 202.3 5.18 214.0 6.2 1.38
C3-fluorenes 208.3 5.50 254.0 5.7 1.26
C1-fluoranthenes-pyrenes 216.3 5.70 237.5 4.5 0.99
chrysene 228.3 5.61 251.0 4.7 1.04
C1-chrysenes 242.3 6.20 273.2 4.3 0.95
benzo[a]pyrene 252.3 6.04 263.0 3.1 0.69
benzo[e]pyrene 252.3 6.40 263.0 2.4 0.53
perylene 252.3 6.50 263.0 2.5 0.55
C2-chrysenes 256.4 6.50 295.4 3.2 0.70
C3-chrysenes 270.4 6.80 317.6 2.4 0.52
(Continued)
benzo[g, h, i]perylene 276.3 7.23 277.0 1.3 0.29

indeno[1,2,3-c, d]pyrene 276.3 7.00 283.5 2.2 0.49
dibenz[a, h]anthracene 278.4 6.75 300.0 1.9 0.41
C4-chrysenes 284.4 7.10b 339.8 0.8 0.17
coronene 300.4 7.64 292.0 1.0 0.23
a
Adopted from Luellen and Shea (2002).
b
The value listed by the authors (log K ow = 8) seems to be unrealistic compared to the values for C1-, C2-, and
C3-chrysenes.
460 cm2 SPMDs at a Temperature of 16.5 ◦ C and a Flow Rate of 0.1 cm s−1 (Huckins
et al., 2004)
Compound M(g mol−1 ) log K ow

a
VLebas (cm3 mol−1 ) R s (L d−1 ) ku (L g−1 d−1 )
0.1 cm s−1 , 16.5 ◦ C

naphthalene 128.2 3.40 147.6 0.5 0.11
benzo[b]thiophene 134.2 3.10 139.7 1.0 0.22
acenaphthylene 152.2 4.10 165.7 1.8 0.40
acenaphthene 154.2 4.20 173.0 2.0 0.45
biphenyl 154.2 3.90 184.6 1.9 0.43
1,3-dimethylnaphthalene 156.2 4.30 192.0 2.3 0.50
1-ethyl naphthalene 156.2 4.40 192.0 2.2 0.48
fluorene 166.2 4.40 188.0 2.1 0.46
4-methyl biphenyl 168.2 4.63 206.8 2.2 0.48
2,3,5-trimethyl naphthalene 170.3 4.80 214.2 2.3 0.51
anthracene 178.2 4.50 197.0 2.3 0.52
phenanthrene 178.2 4.50 199.0 2.3 0.51
1-methyl fluorene 180.3 4.97 210.0 2.2 0.49
2-methyl phenanthrene 192.3 5.15 218.7 2.5 0.56
fluoranthene 202.3 5.20 217.0 2.4 0.54
pyrene 202.3 5.30 214.0 2.5 0.55
3,6-dimethyl phenanthrene 206.3 5.25 240.9 2.7 0.60
2-methyl fluoranthene 216.3 5.30 236.0 2.7 0.59
chrysene 228.3 5.60 251.0 1.5 0.34
benzo[b]naphtho[2,1-d]thiophene 234.3 5.19 242.9 2.4 0.53
benzo[a]pyrene 252.3 6.40 263.0 1.6 0.36
benzo[e]pyrene 252.3 6.04 263.0 1.6 0.36
perylene 252.3 6.25 263.0 1.7 0.38
indeno[1,2,3-c,d]pyrene 276.3 6.50 283.5 1.8 0.41
a
Adopted from Huckins et al. (2004).
460 cm2 SPMDs at Temperatures of 2, 13, and 30 ◦ C, and a Flow Rate of 90 cm s−1
(Booij et al., 2003)
Compound M(g mol−1 ) log K ow a VLebas (cm3 mol−1 ) Rs (L d−1 ) ku (L g−1 d−1 )
90 cm s−1 , 2 ◦ C
acenaphthene 154.2 3.92 173.0 21.2 4.72
phenanthrene 178.2 4.57 199.0 40.2 8.93
fluoranthene 202.3 5.22 217.0 44.3 9.85
pyrene 202.3 5.18 214.0 53.1 11.81
chrysene 228.3 5.86 251.0 51.5 11.44
benzo[a]pyrene 252.3 6.04 263.0 47.4 10.54
1,2,3,4-tetrachlorobenzene 215.9 4.64 180.0 25.8 5.74
pentachlorobenzene 250.3 5.18 200.0 54.8 12.18
hexachlorobenzene 284.8 5.73 221.4 73.6 16.36
PCB 28 257.5 5.80 247.3 84.4 18.75
PCB 52 292.0 6.10 268.2 72.9 16.20
PCB 118 326.4 6.40 289.1 44.1 9.80
PCB 153 360.9 6.90 310.0 42.9 9.54
PCB 170 395.3 6.90 330.9 46.2 10.27
PCB 194 429.8 7.40 351.8 67.3 14.96
90 cm s−1 , 13 ◦ C
acenaphthene 154.2 3.92 173.0 52.1 11.58
phenanthrene 178.2 4.57 199.0 71.5 15.88
fluoranthene 202.3 5.22 217.0 101.7 22.59
pyrene 202.3 5.18 214.0 108.7 24.17
chrysene 228.3 5.86 251.0 99.8 22.17
benzo[a]pyrene 252.3 6.04 263.0 100.4 22.30
benzo[e]pyrene 252.3 6.04 263.0 100.0 22.22
PCB 52 292.0 6.10 268.2 116.8 25.95
PCB 118 326.4 6.40 289.1 97.4 21.65
PCB 153 360.9 6.90 310.0 97.6 21.69
PCB 170 395.3 6.90 330.9 83.7 18.59
PCB 194 429.8 7.40 351.8 78.4 17.42
90 cm s−1 , 30 ◦ C
acenaphthene 154.2 3.92 173.0 68.0 15.12
phenanthrene 178.2 4.57 199.0 104.9 23.30
fluoranthene 202.3 5.22 217.0 125.2 27.83
pyrene 202.3 5.18 214.0 130.1 28.92
chrysene 228.3 5.86 251.0 181.2 40.26
benzo[a]pyrene 252.3 6.04 263.0 197.4 43.86
benzo[e]pyrene 252.3 6.04 263.0 182.0 40.44
90 cm s−1 , 30 ◦ C
1,2,3,4-tetrachlorobenzene 215.9 4.64 180.0 145.3 32.29
(Continued)
192 Appendix A

PCB 28 257.5 5.80 247.3 210.4 46.75
PCB 52 292.0 6.10 268.2 187.5 41.66
PCB 118 326.4 6.40 289.1 165.5 36.79
PCB 153 360.9 6.90 310.0 142.5 31.67
PCB 170 395.3 6.90 330.9 131.0 29.11
PCB 194 429.8 7.40 351.8 121.5 27.01
a
Adopted from Booij et al. (2003).
460 cm2 SPMDs at a Temperature of 12 ◦ C and a Flow Rate of 0.004 cm s−1
(Meadows et al., 1998)
0.004 cm s−1 , 12 ◦ C
PCB 6 223.1 5.10 226.4 13.2 2.93
PCB 18 257.5 5.20 247.3 9.6 2.12
PCB 19 257.5 5.00 247.3 5.5 1.21
PCB 22 257.5 5.60 247.3 5.9 1.31
PCB 25 257.5 5.70 247.3 5.9 1.31
PCB 26 257.5 5.70 247.3 5.9 1.31
PCB 28 257.5 5.70 247.3 8.6 1.92
PCB 31 257.5 5.70 247.3 7.3 1.62
PCB 40 292.0 5.70 268.2 6.8 1.52
PCB 41 292.0 5.70 268.2 6.4 1.42
PCB 42 292.0 5.80 268.2 6.4 1.42
PCB 43 292.0 5.80 268.2 6.4 1.42
PCB 44 292.0 5.80 268.2 7.7 1.72
PCB 45 292.0 5.50 268.2 8.2 1.82
PCB 46 292.0 5.50 268.2 4.6 1.01
PCB 47 292.0 5.80 268.2 7.7 1.72
PCB 48 292.0 5.80 268.2 3.6 0.81
PCB 49 292.0 5.80 268.2 5.5 1.21
PCB 51 292.0 5.60 268.2 5.0 1.11
PCB 52 292.0 5.80 268.2 6.4 1.42
PCB 53 292.0 5.60 268.2 5.0 1.11
PCB 63 292.0 6.20 268.2 5.5 1.21
PCB 64 292.0 6.00 268.2 7.7 1.72
PCB 66 292.0 6.20 268.2 5.5 1.21
PCB 67 292.0 6.20 268.2 5.5 1.21
PCB 70 292.0 6.20 268.2 7.3 1.62
PCB 74 292.0 6.20 268.2 6.4 1.42
PCB 81 292.0 6.40 268.2 5.0 1.11
PCB 82 326.4 6.20 289.1 4.6 1.01
(Continued)
PCB 83 326.4 6.30 289.1 5.0 1.11

PCB 84 326.4 6.00 289.1 4.6 1.01
PCB 85 326.4 6.30 289.1 5.0 1.11
PCB 87 326.4 6.30 289.1 5.5 1.21
PCB 90 326.4 6.40 289.1 6.4 1.42
PCB 91 326.4 6.10 289.1 4.6 1.01
PCB 92 326.4 6.40 289.1 5.5 1.21
PCB 95 326.4 6.10 289.1 6.4 1.42
PCB 97 326.4 6.30 289.1 4.6 1.01
PCB 99 326.4 6.40 289.1 4.6 1.01
0.004 cm s−1 , 12 ◦ C
PCB 101 326.4 6.40 289.1 6.4 1.42
PCB 105 326.4 6.60 289.1 4.1 0.91
PCB 107 326.4 6.70 289.1 5.5 1.21
PCB 110 326.4 6.50 289.1 5.9 1.31
PCB 114 326.4 6.60 289.1 4.6 1.01
PCB 118 326.4 6.70 289.1 5.0 1.11
PCB 119 326.4 6.60 289.1 4.6 1.01
PCB 128 360.9 6.70 310.0 4.6 1.01
PCB 129 360.9 6.70 310.0 3.6 0.81
PCB 130 360.9 6.80 310.0 4.1 0.91
PCB 134 360.9 6.60 310.0 5.0 1.11
PCB 136 360.9 6.20 310.0 5.5 1.21
PCB 137 360.9 6.80 310.0 3.6 0.81
PCB 138 360.9 6.80 310.0 5.0 1.11
PCB 141 360.9 6.80 310.0 5.0 1.11
PCB 146 360.9 6.90 310.0 5.0 1.11
PCB 149 360.9 6.70 310.0 5.9 1.31
PCB 151 360.9 6.60 310.0 5.5 1.21
PCB 153 360.9 6.90 310.0 3.6 0.81
PCB 156 360.9 7.20 310.0 2.7 0.61
PCB 157 360.9 7.20 310.0 2.7 0.61
PCB 158 360.9 7.00 310.0 3.6 0.81
PCB 172 395.3 7.30 330.9 1.4 0.30
PCB 174 395.3 7.10 330.9 3.2 0.71
PCB 176 395.3 6.80 330.9 2.3 0.51
PCB 178 395.3 7.10 330.9 3.2 0.71
PCB 179 395.3 6.70 330.9 2.3 0.51
PCB 180 395.3 7.40 330.9 2.7 0.61
PCB 183 395.3 7.20 330.9 3.2 0.71
PCB 187 395.3 7.20 330.9 3.6 0.81
PCB 194 429.8 7.80 351.8 1.4 0.30
PCB 199 429.8 7.60 351.8 1.8 0.40
PCB 201 429.8 7.30 351.8 1.8 0.40
PCB 207 464.2 7.70 372.7 0.3 0.06
a
Adopted from Meadows et al. (1998).
194 Appendix A
460 cm2 SPMDs at a Temperature of 11 and 19 ◦ C and a Flow Rate of 8 cm s−1
(Rantalainen et al., 2000)
8 cm s−1 , 11 ◦ C
PCB 77 292.0 6.36 268.2 2.5 0.56
PCB 78 292.0 6.35 268.2 3.8 0.85
PCB 79 292.0 6.42 268.2 4.4 0.98
PCB 81 292.0 6.36 268.2 3.2 0.71
PCB 126 326.4 6.89 289.1 1.9 0.43
PCB 127 326.4 6.95 289.1 1.4 0.31
PCB 169 360.9 7.42 310.0 1.8 0.40
2,3,7,8-TCDF 306.0 6.53 268.2 2.2 0.48

1,2,3,7,8-PeCDF 340.4 6.79 289.1 1.7 0.39
2,3,4,7,8-PeCDF 340.4 6.92 289.1 1.6 0.36
1,2,3,4,7,8-HxCDF 374.9 7.00 310.0 1.1 0.24
1,2,3,6,7,8-HxCDF 374.9 7.00 310.0 1.1 0.24
1,2,3,7,8,9-HxCDF 374.9 7.00 310.0 0.8 0.19
2,3,4,6,7,8-HxCDF 374.9 7.00 310.0 1.1 0.24
1,2,3,4,6,7,8-HpCDF 409.3 7.92 330.9 0.6 0.14
OCDF 443.8 7.97 351.8 0.4 0.09
2,3,7,8-TCDD 322.0 6.42 275.6 2.2 0.48

1,2,3,7,8-PeCDD 356.4 6.64 296.5 1.6 0.35
1,2,3,4,7,8-HxCDD 390.9 7.80 317.4 1.2 0.28
1,2,3,6,7,8-HxCDD 390.9 7.80 317.4 1.2 0.26
1,2,3,7,8,9-HxCDD 390.9 7.80 317.4 1.1 0.24
1,2,3,4,6,7,8-HpCDD 425.2 8.00 338.3 0.6 0.14
OCDD 459.8 8.20 359.2 1.1 0.25
8 cm s−1 , 19 ◦ C
PCB 77 292.0 6.36 268.2 3.8 0.85
PCB 78 292.0 6.35 268.2 4.5 0.99
PCB 79 292.0 6.42 268.2 4.5 1.01
PCB 81 292.0 6.36 268.2 4.3 0.95
PCB 126 326.4 6.89 289.1 3.7 0.81
PCB 127 326.4 6.95 289.1 3.6 0.79
PCB 169 360.9 7.42 310.0 5.1 1.13
OCDF 443.8 7.97 351.8 1.6 0.35
8 cm s−1 , 19 ◦ C
2,3,7,8-TCDF 306.0 6.53 268.2 3.2 0.71
1,2,3,7,8-PeCDF 340.4 6.79 289.1 3.3 0.74
2,3,4,7,8-PeCDF 340.4 6.92 289.1 3.7 0.81
1,2,3,4,7,8-HxCDF 374.9 7.00 310.0 2.3 0.51
1,2,3,6,7,8-HxCDF 374.9 7.00 310.0 2.5 0.56
1,2,3,7,8,9-HxCDF 374.9 7.00 310.0 2.0 0.44
2,3,4,6,7,8-HxCDF 374.9 7.00 310.0 2.6 0.57
1,2,3,4,6,7,8-HpCDF 409.3 7.92 330.9 2.3 0.51
(Continued)
Compound M(g mol−1 ) log Kow a VLebas (cm3 mol−1 ) R s (L d−1 ) ku (L g−1 d−1 )
2,3,7,8-TCDD 322.0 6.42 275.6 3.3 0.74

1,2,3,7,8-PeCDD 356.4 6.64 296.5 2.9 0.65
1,2,3,4,7,8-HxCDD 390.9 7.80 317.4 3.5 0.78
1,2,3,6,7,8-HxCDD 390.9 7.80 317.4 2.7 0.61
1,2,3,7,8,9-HxCDD 390.9 7.80 317.4 2.5 0.56
1,2,3,4,6,7,8-HpCDD 425.2 8.00 338.3 1.9 0.43
OCDD 459.8 8.20 359.2 2.6 0.58
a
Adopted from Rantalainen et al. (2000).
460 cm2 SPMDs at a Temperature of 19 ◦ C and Flow Rates of 0.06, 0.28, and 1.14
cm s−1 (Vrana and Schüürmann, 2002)
0.06 cm s−1 , 19 ◦ C
α-HCH 290.8 3.80 243.6 2.3 0.52
β-HCH 290.8 3.80 243.6 3.6 0.79
γ-HCH 290.8 3.70 243.6 2.5 0.56
δ-HCH 290.8 4.10 243.6 3.1 0.69
0.28 cm s−1 , 19 ◦ C
α-HCH 290.8 3.80 243.6 2.4 0.54
β-HCH 290.8 3.80 243.6 2.0 0.44
γ-HCH 290.8 3.70 243.6 2.3 0.52
δ-HCH 290.8 4.10 243.6 3.8 0.84
1.14 cm s−1 , 19 ◦ C
α-HCH 290.8 3.80 243.6 1.9 0.42
β-HCH 290.8 3.80 243.6 0.9 0.20
γ-HCH 290.8 3.70 243.6 1.3 0.28
δ-HCH 290.8 4.10 243.6 1.8 0.41
a
Adopted from Vrana and Schüürmann (2002).
196 Appendix A
460 cm2 SPMDs at a Temperature of 10, 18, and 26 ◦ C and a Flow Rate of
0.004 cm s−1 (Huckins et al., 2002b)
0.004 cm s−1 , 10 ◦ C
pentachloroanisole 280.4 5.48 231.8 2.9 0.64
α-HCH 290.8 3.86 243.6 0.9 0.20
γ -HCH 290.8 3.71 243.6 0.7 0.16
o,p -DDE 318.0 5.56 305.2 2.3 0.51
p,p -DDE 318.0 6.14 305.2 2.8 0.62
o,p -DDD 320.1 6.08 312.6 2.5 0.56
p,p -DDD 320.1 5.75 312.6 2.3 0.51
dacthal 332.0 4.26 301.6 0.6 0.13
p,p -methoxychlor 345.7 4.61 354.3 1.2 0.27
o,p -DDT 354.5 5.59 333.5 2.0 0.44
p,p -DDT 354.5 5.47 333.5 2.0 0.44
heptachlor 373.3 5.19 308.2 2.6 0.58
endrin 380.9 4.63 318.2 2.3 0.51
dieldrin 380.9 4.60 318.2 1.3 0.29
heptachlor epoxide 389.3 4.51 309.6 1.3 0.29
cis-chlordane 409.8 5.38 336.5 2.6 0.58
trans-chlordane 409.8 5.38 336.5 2.4 0.53
oxychlordane 423.8 5.48 334.1 2.3 0.51
cis-nonachlor 444.2 6.20 357.4 2.2 0.49
trans-nonachlor 444.2 6.35 357.4 2.7 0.60
mirex 545.6 6.89 403.2 3.0 0.67
0.004 cm s−1 , 18 ◦ C
α-HCH 290.8 3.86 243.6 1.4 0.31
γ -HCH 290.8 3.71 243.6 1.1 0.24
o,p -DDE 318.0 5.56 305.2 2.4 0.53
p,p -DDE 318.0 6.14 305.2 2.7 0.60
o,p -DDD 320.1 6.08 312.6 2.3 0.51
p,p -DDD 320.1 5.75 312.6 2.5 0.56
dacthal 332.0 4.26 301.6 1.8 0.40
o,p -DDT 354.5 5.59 333.5 3.3 0.73
p,p -DDT 354.5 5.47 333.5 3.7 0.82
endrin 380.9 4.63 318.2 3.2 0.71
dieldrin 380.9 4.60 318.2 2.6 0.58
cis-chlordane 409.8 5.38 336.5 1.7 0.38
trans-chlordane 409.8 5.38 336.5 2.0 0.44
oxychlordane 423.8 5.48 334.1 1.9 0.42
cis-nonachlor 444.2 6.20 357.4 2.0 0.44
trans-nonachlor 444.2 6.35 357.4 1.9 0.42
mirex 545.6 6.89 403.2 2.4 0.53
(Continued)
0.004 cm s−1 , 26 ◦ C
α-HCH 290.8 3.86 243.6 1.8 0.40
β-HCH 290.8 3.86 243.6 1.6 0.36
γ -HCH 290.8 3.71 243.6 2.3 0.51
o,p -DDE 318.0 5.56 305.2 6.0 1.33
p,p -DDE 318.0 6.14 305.2 6.8 1.51
o,p -DDD 320.1 6.08 312.6 5.5 1.22
p,p -DDD 320.1 5.75 312.6 6.1 1.36
dacthal 332.0 4.26 301.6 2.0 0.44
p,p -methoxychlor 345.7 4.61 354.3 2.5 0.56
o,p -DDT 354.5 5.59 333.5 4.0 0.89
p,p -DDT 354.5 5.47 333.5 4.1 0.91
heptachlor 373.3 5.19 308.2 6.8 1.51
endrin 380.9 4.63 318.2 7.6 1.69
dieldrin 380.9 4.60 318.2 4.6 1.02
cis-chlordane 409.8 5.38 336.5 6.0 1.33
trans-chlordane 409.8 5.38 336.5 6.0 1.33
oxychlordane 423.8 5.48 334.1 5.6 1.24
cis-nonachlor 444.2 6.20 357.4 4.9 1.09
trans-nonachlor 444.2 6.35 357.4 6.0 1.33
mirex 545.6 6.89 403.2 5.8 1.29
a
Adopted from Huckins et al. (2002b) or from US EPA (2003).
460 cm2 SPMDs at an Unknown Temperature and a Flow Rate of 0.006 cm s−1
(Sabaliūnas and Södergren, 1997)
0.006 cm s−1 , ◦ C unknown

trifluralin 335.3 5.34 295.9 3.6 0.80
chlorpyrifos 350.6 4.96 298.8 5.2 1.16
heptachlor 373.4 6.10 308.2 5.0 1.11
dieldrin 380.9 5.40 318.2 6.8 1.51
fenvalerate 419.9 6.20 479.6 2.6 0.58
deltamethrin 505.2 6.20 462.5 2.5 0.56
a
Adopted from Sabaliūnas and Södergren (1997) or from US EPA (2003).
198 Appendix A
A.3. AIR SAMPLING RATES
Air sampling rates are available from laboratory exposures and field deploy-
ments. In the latter case, vapor phase concentrations (Ca ) were determined by
traditional high-volume sampling (HiVol), using a glass fiber filter in combina-
tion with a solid-phase adsorbent. Vapor-phase concentrations in the field may be
highly variable, and some uncertainty exists to what extent the intermittently deter-
mined Ca values reflect the average concentration during the SPMD exposure. The
Rs s from studies that fail to simultaneously determine vapor phase concentrations
are excluded from this appendix. As discussed in Chapter 3, it should be noted
that air-exposed SPMDs may sample particle-bound contaminants such as PAHs
≥ 5-rings. In Tables A.13 through A.16, Rs values are listed for those studies in
which vapor phase concentrations were determined and particulate contributions
could be neglected.
TABLE A.13 Air Sampling Rates (R s s) and Uptake Rate Constants (ku s) of 4.5 g,
460 cm2 SPMDs at Various Flow Rates and Temperatures (Ockenden et al.,
1998)
Compound M(g mol−1 ) log K oa a VLebas (cm3 mol−1 ) R s (m3 d−1 ) ku (m3 g−1 d−1 )
460 cm s−1 , 4 ◦ C
PCB 28 257.5 9.0 247.3 4.5 1.00
PCB 52 292.0 9.3 268.2 6.2 1.38
PCB 101 326.4 10.1 289.1 11.0 2.44
PCB 118 326.4 11.3 289.1 14.0 3.11
PCB 138 360.9 10.9 310.0 18.0 4.00
PCB 153 360.9 10.9 310.0 15.0 3.33
PCB 180 395.3 11.4 330.9 13.0 2.89
unknown flow rate, 14 ◦ C

PCB 28 257.5 8.5 247.3 2.8 0.63
PCB 52 292.0 8.8 268.2 4.5 1.01
PCB 101 326.4 9.5 289.1 7.2 1.61
PCB 118 326.4 10.7 289.1 8.5 1.89
PCB 138 360.9 10.4 310.0 7.2 1.60
PCB 153 360.9 10.3 310.0 6.3 1.41
PCB 180 395.3 10.9 330.9 7.0 1.56
270 cm s−1 , 18 ◦ C
PCB 52 292.0 8.6 268.2 1.3 0.29
PCB 101 326.4 9.3 289.1 3.3 0.73
PCB 118 326.4 10.5 289.1 6.9 1.53
PCB 138 360.9 10.2 310.0 4.3 0.96
PCB 153 360.9 10.1 310.0 4.5 1.00
PCB 180 395.3 10.7 330.9 4.9 1.09
a
Adopted from Kömp and MacLachlan (1997), evaluated at the exposure temperature.
TABLE A.14 Air Sampling Rates (Rs s) and Uptake Rate Constants (ku s) of 4.5 g,
460 cm2 SPMDs at Unknown Flow Rates and Temperatures of 25 and 26 ◦ C
(Huckins et al., 1994; Petty et al., 1993)

PCB 7 223.1 7.4 226.4 0.9 0.20
PCB 17 223.1 7.6 226.4 0.9 0.19
PCB 18 257.5 7.6 247.3 0.9 0.21
PCB 22 257.5 8.0 247.3 2.6 0.58
PCB 52 292.0 8.4 268.2 3.0 0.66
PCB 74 292.0 9.1 268.2 5.7 1.28
PCB 95 326.4 9.1 289.1 3.6 0.80
PCB 101 326.4 9.3 289.1 5.3 1.19
PCB 110 326.4 9.0 289.1 5.7 1.27
PCB 138 360.9 10.1 310.0 7.1 1.59
PCB 153 360.9 10.0 310.0 7.0 1.56
PCB 174 395.3 10.3 330.9 7.7 1.71
PCB 183 395.3 10.3 330.9 5.6 1.25
PCB 201 429.8 10.9 351.8 9.6 2.14

PCB 52 8.2 3.0 0.67
a
Adopted from Kömp and MacLachlan (1997) or from Harner and Bidleman (1996).
460 cm2 SPMDs at a Flow Rate of 0.001 cm s−1 and a Temperature of 23 ◦ C (Shoeib
and Harner, 2002)
0.001 cm s−1 , 23 ◦ C
PCB 28 257.5 8.12 247.3 3.5 0.78
PCB 44 292.0 8.61 268.2 3.3 0.72
PCB 52 292.0 8.43 268.2 3.6 0.81
PCB 77 292.0 9.75 268.2 3.3 0.72
PCB 81 292.0 9.67 268.2 3.4 0.76
PCB 99 326.4 9.24 289.1 3.1 0.68
PCB 101 326.4 9.19 289.1 3.2 0.70
PCB 105 326.4 10.2 289.1 3.7 0.83
PCB 114 326.4 10.0 289.1 3.9 0.87
PCB 118 326.4 9.96 289.1 4.0 0.89
PCB 126 326.4 10.4 289.1 9.2 2.04
PCB 128 360.9 10.4 310.0 4.8 1.07
PCB 137/138 360.9 10.1 310.0 5.0 1.12
PCB 153 360.9 9.91 310.0 3.8 0.85
PCB 156 360.9 10.8 310.0 4.6 1.01
PCB 180 395.3 10.7 330.9 7.3 1.63
tetrachloronaphthalenes 266.0 8.4 231.2 2.8 0.62

pentachloronaphthalenes 300.4 9.1 252.1 3.0 0.66
hexachloronaphthalenes 334.8 10.1 273.0 4.7 1.05
a
PCBs: Shoeib and Harner (2002); PCNs: Harner and Bidleman (1998).
200 Appendix A
460 cm2 SPMDs at a Flow Rate of 3.4 cm s−1 and a Temperature of 25 ◦ C
(Robertson, 2004)
3.4 cm s−1 , 25 ◦ C
diazinon 304.4 9.1 349.0 2.6 0.58
p, p -DDE 318.0 9.3 305.2 6.9 1.53
p, p -DDD 320.1 9.6 312.6 6.9 1.53
chlorpyrifos 350.6 8.9 325.8 5.1 1.13
p, p -DDT 354.5 10.4 333.5 5.5 1.22
cis-permethrin 391.3 10.6 427.4 3.4 0.76
trans-permethrin 391.3 10.6 427.4 5.5 1.22
trans-chlordane 409.8 8.9 336.5 5.1 1.13
cis-chlordane 409.8 8.9 336.5 5.5 1.22
cyfluthrin 434.3 10.4 459.1 4.1 0.91
cypermethrin 434.3 10.8 454.1 0.7 0.16
fipronil 437.2 11.5 358.5 2.3 0.51
fluorene 166.2 6.8 188.0 2.4 0.53

phenanthrene 178.2 7.6 199.0 4.3 0.96
anthracene 178.2 7.3 197.0 3.7 0.82
fluoranthene 202.3 8.9 217.0 6.4 1.42
pyrene 202.3 8.8 214.0 6.9 1.53
benz[a]anthracene 228.3 9.5 248.0 15.0 3.33
chrysene 228.3 10.4 251.0 13.0 2.89
benzo[a]pyrene 252.3 10.8 263.0 6.0 1.33
a
fluorene, phenanthrene, fluoranthene, pyrene: Harner and Bidleman (1998); other PAHs: calculated from log K ow s
and Henry’s law constants (Mackay et al. 1992b); pesticides: calculated from log K ow s and Henry’s law constants
(US EPA, 2003).
A.4. REFERENCES
Booij, K.; Hofmans, H.E.; Fischer, C.V.; van Weerlee, E.M. 2003, Temperature-dependent uptake
rates of non-polar organic compounds by semipermeable membrane devices and low-density
polyethylene membranes. Environ. Sci. Technol. 37: 361–366.
Harner, T. and Bidleman, T.F. 1996, Measurements of octanol-air partition coefficients for polychlori-
nated biphenyls. J. Chem. Eng. Data 41: 895–899.
Harner, T. and Bidleman, T.F. 1998, Measurement of octanol-air partition coefficients for polycyclic
aromatic hydrocarbons and polychlorinated naphthalenes. J. Chem. Eng. Data 43: 40–46.
27: 2489–2496.
Huckins, J.N.; Petty, J.D.; Orazio, C.E.; Zajicek, J.L.; Gibson, V.L.; Clark, R.C.; Echols, K.R. 1994,
Semipermeable Membrane Device (SPMD) Sampling Rates for Trace Organic Contaminants in
Air and Water. 15th Annual meeting of Society of Environmental Toxicology and Chemistry;
Denver, CO.; Abstract MB01.
Technol. 33: 3918–3923.
approach for in situ calibration of semipermeable membrane devices. Environ. Sci. Technol. 36:
85–91.
W.R.; Steen, A.; Gale, R.W.; and Ingersoll, C.G. 2004, Overview and comparison of lipid-
containing semipermeable membrane devices (SPMDs) and oysters (Crassostrea gigas) for as-
sessing organic chemical exposure. Environ. Toxicol. Chem. 23: 1617–1628.
Kömp, P. and McLachlan, M.S. 1997, Octanol/air partitioning of polychlorinated biphenyls. Environ.
Toxicol. Chem. 16: 2433–2437.
Lefkovitz, L.; Crecelius, E.; McElroy, N. 1996, The use of Polyethylene Alone to Predict Dissolved-
phase Organics in the Columbia River. 17th Annual Meeting of Society of Environmental Toxicol-
ogy Chemistry; November 17–21, 1996; Washington, D.C.; Abstract PO523.
Luellen, D.R. and Shea, D. 2002, Calibration and field verification of semipermeable membrane de-
vices for measuring polycyclic aromatic hydrocarbons in water. Environ. Sci. Technol. 36: 1791–
1797.
Mackay, D.; Shiu, W.Y.; Ma, K.C. 1992b,Illustrated Handbook of Physical-Chemical Properties and
Rantalainen, A.-L.; Cretney, W.; Ikonomou, M.G. 2000, Uptake rates of semipermeable membrane
devices (SPMDs) for PCDDs, PCDFs and PCBs in water and sediment. Chemosphere 40: 147–
158.
Reynolds, G.W.; Hoff, J.T.; Gillham, G.W. 1990, Sampling bias caused by materials used to monitor
halocarbons in groundwater. Environ. Sci. Technol. 24: 135–140.
Robertson, G, 2004, U.S. EPA, National Exposure Research Laboratory: Las Vegas, NV;. Personal
communication.
202 Appendix A
Sabalinas, D. and Södergren, A. 1997, Use of semi-permeable membrane devices to monitor pollutants
in water and assess their effects: A laboratory test and field verification. Environ. Pollut. 96:
195–206.
US EPA 2003, Estimation Program Interface (EPI) Suite. Version 3.11; U.S. Environmental Protection
Agency; http://www.epa.gov/opptintr/exposure/docs/episuite.htm (accessed on November, 2004).
Appendix B
SPMD Bibliography
Over the last decade, the growth in SPMD research and applications has been
remarkable. Reports on this work have been in the form of graduate degree (masters
and Ph.D.) theses, abstracts from presentations, laboratory reports, journal articles
and book chapters. Herein, we provide the reader with a list of SPMD related
peer-reviewed journal articles and book chapters. However, we do not claim that
this list is complete and apologize in advance to authors of articles not included.
In light of the current rate of the publication of SPMD related articles, additional
papers will undoubtedly be available by the time this book is published.
Anderson, K.A. and Johnson, E. 2001, Bioavailable organochlorine pesticides in a semi-arid region of
Eastern Oregon, USA, as determined by gas chromatography with electron capture detection. J.
AOAC 84: 1371–1382.
Axelman, J.; Næs, K.; Näf, C.; Broman, D. 1999, Accumulation of polycyclic aromatic hydrocarbons
in semipermeable membrane devices and caged mussels (Mytilus edulis) in relation to water
column phase distribution. Environ. Toxicol. Chem. 18: 2454–2461.
Balmer, M.E.; Poiger, T.; Droz, C.; Romanin, K.; Bergqvist, P.-A.; Müller, M.D.; Buser, H.-R. 2004,
Occurrence of methyl triclosan, a transformation product of the bactericide triclosan in fish from
various lakes in Switzerland. Environ. Sci. Technol. 38: 390–395.
38: 5983–5900.
devices during chronic exposure to dispersed crude oil.Environ. Toxicol. Chem. 20: 1175–1184.
Bennett, E.R.; Metcalfe, T.L.; Metcalfe, C.D. 1996, Semipermeable membrane devices (SPMDs)
for monitoring organic contaminants in the Otanabee River, Ontario. Chemosphere 33:
363–375.
203
204 Appendix B
Bennett, E.R. and Metcalfe, C.D. 2000, Distribution of degradation products of alkylphenol ethoxylates
near sewage treatment plants in the lower Great Lakes, North America. Environ. Toxicol. Chem.
19: 784–792.
Bergqvist, P.-A.; Strandberg, B.; Hjelt, M.; Rappe, C. 1990, Nondestructive steps for the lipid reduction
in PCDD and PCDF analyses. Organohal. Comp. 2: 103–106.
Bergqvist, P.-A.; Strandberg, B.; Ekelund, R.; Rappe, C.; Granmo, Å. 1998a, Temporal monitoring
of organochlorine compounds in seawater by semipermeable membranes following a flooding
episode in Western Europe. Environ. Sci. Technol. 32: 3887–3892.
Bergqvist, P.-A.; Strandberg, B.; Rappe, C. 1998b, Lipid removal using semipermeable membranes in
(Mytilus edulis). Organohal. Comp. 47: 89–92.
Booij, K.; Smedes, F.; van Weerlee, E.M. 2002a, Spiking of performance reference compounds in low
density polyethylene and silicone passive water samplers. Chemosphere 46: 1157–1161.
Booij, K.; Zegers, B.N.; Boon, J.P. 2002b, Levels of some polybrominated diphenyl ether (PBDE)
flame retardants along the Dutch coast as derived from their accumulation in SPMDs and blue
mussels (Mytilus edulis).Chemosphere 46: 683–688.
Booij, K.; Hofmans, H.E.; Fischer, C.V.; van Weerlee, E.M. 2003, Temperature-dependent uptake
rates of nonpolar organic compounds by semipermeable membrane devices and low-density
polyethylene membranes.Environ. Sci. Technol. 37: 361–366.
Bridges, C.M. and Little, E.E. 2003, Using Semipermeable Membrane Devices (SPMDs) to Assess the
Toxicity and Teratogenicity of Aquatic Amphibian Habitats. ASTM Special Technical Publication
No. 1443; pp. 159–168.
Bridges, C.M.; Little, E.E.; Gardiner, D.M.; Petty, J.D.; Huckins, J.D. 2004, Assessing the toxicity of
teratogenicity of pond water in North-Central Minnesota to amphibians. Environ. Sci. Pollut. R.
11: 233–239.
Čáslavskŷ, J.; Zdrahal, Z.; Vytopilova, M. 2000, Application of SPMDs for PAH sampling in the DEZA
chemical factory. Polycyclic Aromat. Compd.20: 123–141.
Čáslavskŷ, J. and Kotlařı́ková, P. 2003, High-molecular-weight polycyclic aromatic hydrocarbons in
the area and vicinity of the DEZA chemical plant, Czech Republic. Polycyclic Aromat. Compd.
23: 327–352.
Čáslavskŷ, J.; Kotlařı́ková, P.; Benešová, K. 2004, Sampling of airborne polycyclic aromatic hydro-
carbons with semipermeable membrane devices. Environ. Chem. Letters 2: 89–92.
Chambers, D.B. 1999, Semipermeable membrane devices used to estimate bioconcentration of poly-
chlorinated biphenyls. J. Am. Water Res. Assoc. 33: 143–154.
Cleveland, L.; Little, E.E.; Petty, J.D.; Johnson, B.T.; Lebo, J.A.; Orazio, C.E.; Dionne. J.; Crockett,
A. 1997, Toxicological and chemical screening of Antarctica sediments: Use of whole sediment
toxicity tests, Microtox, Mutatox, and semipermeable membrane devices (SPMDs). Mar. Pollut.
Bull. 34: 194–202.
Corrigan, B.P. and Jones, K.C. 2002, Passive sampling survey of PBDEs and PCBs in UK air.
Organohal. Comp. 56: 449–452.
Crunkilton, R.L. and DeVita, W.M. 1997, Determination of aqueous concentrations of polycyclic
aromatic hydrocarbons (PAHs) in an urban stream. Chemosphere 35: 1447–1463.
De la Torre, A.I.; Fernandez, C.; Tarazona, J.V.; Muñoz, M.J. 1995, Detection of aroclor, DDT,
malathion, and HCB using semipermeable membranes as concentration method. Chemosphere
31: 2727–2737.
SPMD Bibliography 205
De la Torre, A.I.; Fernandez, C.; Tarazona, J.V.; Muñoz, M.J. 1998, Combination of semipermeable
membranes and toxicity bioassays for the detection of lipophylic toxic chemicals in environ-
mental samples. Ecotoxicol. Environ. Restoration 1: 78–84.
De la Torre, A.I.; Jimenez, J.A.; Carballo, M.; Fernandez, C.; Roset, J.; Muñoz, M.J. 2000, Ecotoxi-
cological evaluation of a pig slurry. Chemosphere 41: 1629–1635.
DeVita, W.M. and Crunkilton, R.L. 1998, Quality Control Associated with Use of Semipermeable
Membrane Devices. In Environmental Toxicology & Risk Assessment, Seventh Volume; Little,
E.E., Delonay, A.J., Greenberg, B.M., Eds.; ASTM STP 1333; American Society for Testing and
Materials: West Conshocken, PA; pp 237–245.
Echols, K.R.; Gale, R.W.; Schwartz, T.R.; Huckins, J.N.; Williams, L.L.; Meadows, J.C.; Morse, D.;
Petty, J.D.; Orazio, C.E.; Tillitt, D.E. 2000, Comparing polychlorinated biphenyl concentrations
and patterns in the Saginaw River using sediment, caged fish, and semipermeable membrane
devices. Environ. Sci. Technol. 34: 4095–4102.
Følsvik, N.; Brevik, E.M.; Berge, J.A. 2000, Monitoring of organotin compounds in seawater using
semipermeable membrane devices (SPMDs)—tentative results. J. Environ. Monitor. 2: 281–284.
Følsvik, N.; Brevik, E.M.; Berge, J.A. 2002, Organotin compounds in a Norwegian fjord. A comparison
of concentration levels in semipermeable membrane devices (SPMDs), blue mussels (Mytilus
edulis) and water samples. J. Environ. Monitor. 4: 280–283.
Franke, S.; Meyer, C.; Heinzel, N.; Gaterman, R.; Huehnerfuss, H.; Rimkus, G.; Koenig, W.A.; Francke,
W. 1999, Enantiomeric composition of the polycyclic musks HHCB and AHTN in different
aquatic species. Chirality 11: 795–801.
and semipermeable membrane devices in the Saginaw River, Michigan. Environ. Sci. Technol.
32: 178–187.
Gatermann, R.; Siselli, S.; Hühnerfuss, H.; Rimkus, G.G.; Franke, S.; Hecker, M.; Kallenborn, R.;
Karbe, L.; König, W.A. 2002a, Synthetic musks in the environment. Part 2: Enantioselective
transformation of the polycyclic musk fragrances HHCB, AHTN, AHDI, and ATII in freshwater
fish. Arch. Environ. Con. Tox. 42: 447–453.
Gatermann, R.; Siselli, S.; Hühnerfuss, H.; Rimkus, G. G.; Hecker, M.; Karbe, L. 2002b, Synthetic
musks in the environment. Part 1: Species-dependent bioaccumulation of polycyclic and nitro
musk fragrances in freshwater fish and mussels. Arch. Environ. Con. Tox. 42: 437–446.
Granmo, Å.; Ekelund, R.; Berggren, M.; Brorström-Lunden, E.; Bergqvist, P.-A. 2000, Temporal
trend of organochlorine marine pollution indicated by concentrations in mussels, semipermeable
membrane devices, and sediment. Environ. Sci. Technol. 34: 3323–3329.
Gustavson, K.E.; DeVita, W.; Revis, A.; Harkin, J.M. 2000a, A novel use of a dual-zone restricted access
sorbent: Normal phase separations of methyl oleate and polynuclear aromatic hydrocarbons
stemming from semipermeable membrane devices. J. Chromatogr. A 883: 143–149.
Gustavson, K.E. and Harkin, J.M. 2000b; Comparison of sampling techniques and evaluation of
semipermeable membrane devices (SPMDs) for monitoring polynuclear aromatic hydrocarbons
(PAHs) in groundwater. Environ. Sci. Technol. 34: 4445–4451.
Hajšlova, J. and Šetkova, L. 2004, Synthetic Musks in Bioindicators: Monitoring Data of Fish and
Human Milk Samples from the Czech Republic. The Handbook of Environmental Chemistry,
Springer-Verlag: Heidelberg, Germany; Vol. 3X; pp. 151–188.
Harner, T.; Ikonomou, M.; Shoeib, M.; Stern, G.; Diamond, M. 2002, Passive air sampling re-
sults for polybrominated diphenyl ethers along an urban-rural transect. Organohal. Comp. 57:
33–36.
206 Appendix B
Harner, T.; Shoeib, M.; Diamond, M.; Stern, G.; Rosenberg, B. 2004, Using passive air samplers to
assess urban-rural trends for persistent organic pollutants. 1. Organochlorine pesticides. Environ.
Sci. Technol. 38: 4474–4483.
Herve, S.; Paukku, R.; Paasivirta, J.; Heinonen, P.; Sodergren, A. 1991, Uptake of organochlo-
rines from lake water by hexane-filled dialysis membranes and by mussels. Chemosphere 22:
997–1001.
Herve, S.; Prest, H.F.; Heinonen, P.; Hyötyläinen, T.; Koistinen, J.; Paasivirta, J. 1995, Lipid-filled
semipermeable membrane devices and mussels as samplers of organochlorine compounds in
lake water. Environ. Sci. Pollut. R. 2: 24–30.
Hewitt, L.M.; Pryce, A.C.; Parrott, J.L.; Marlatt, V.; Wood, C.; Oakes, K.; Van der Kraak, G.J. 2003,
Accumulation of ligands for aryl hydrocarbon and sex steroid receptors in fish exposed to treated
effluent from a bleached sulfite/groundwood pulp and paper mill. Environ. Toxicol. Chem. 22:
2890–2897.
Hofelt, C.S. and Shea, D. 1997, Accumulation of organochlorine pesticides and PCBs by semipermeable
membrane devices and Mytilus edulis in New Bedford Harbor. Environ. Sci. Technol. 31: 154–
159.
Huckins, J.N.; Tubergen, M.W.; Lebo, J.A.; Gale, R.W.; Schwartz, T.R. 1990a, Polymeric film dialysis
Huckins, J.N.; Tubergen, M.W.; Manuweera, G.K. 1990b, Semipermeable membrane devices contain-
27: 2489–2496.
membrane devices (SPMDs) for polycyclic aromatic hydrocarbons (PAHs) in water. Environ.
Sci. Technol. 33: 3918–3923.
R.C.; Mogensen, B.B. 2002, Development of the permeability/performance reference compound
Hühnerfuss, H.; Biselli, S.; Gaterman, R. 2004, Enantioselective Analysis of Polycyclic Musks as a Ver-
satile Tool for the Understanding of Environmental Processes. The Handbook of Environmental
Chemistry, Springer-Verlag: Heidelberg, Germany; Vol. 3X, pp. 213–231.
Ikonomou, M.G.; Rayne, S.; Addison, R.F. 2002a, Exponential increases of the brominated flame
retardants, polybrominated diphenyl ethers, in the Canadian Arctic from 1981 to 2000. Environ.
Sci. Technol. 36: 1886–1892.
Ikonomou, M.G.; Rayne, S.; Fischer, M.; Fernandez, M.P.; Cretney, W. 2002b, Occurrence and con-
gener profiles of polybrominated diphenyl ethers in environmental samples from Coastal British
Columbia, Canada. Chemosphere 46: 649–663.
Isidora, M.; Ferrara, M.; Lavorgna, M.; Nardelli, A.; Parrella, A. 2003, In situ monitoring of urban air
in Southern Italy with the Tradescantia micronucleus bioassay and semipermeable membrane
Jaward, F.M.; Meijer, S.N.; Steinnes, E.; Thomas, G.O.; Jones, K.C. 2004, Further studies on the lati-
tudinal and temporal trends of persistent organic pollutants in Norwegian and U.K. background
air. Environ. Sci. Technol. 38: 2523–2530.
Jegorova, I. 2002, Monitoring with Semipermeable Membrane Device (SPMD) Technology for Per-
sistent Organic Pollutants (POPs) Analysis. In Principles of Instrumental Analysis; Leary, J,
Skoog, D, Eds.; 4th Ed.; Harcourt Brace Jovanovich: Orlando, FL.
Johnson, B.T.; Petty, J.D.; Huckins, J.N. 2000, Collection and detection of lipophilic chemical con-
taminants in water, sediment, soil, and air. Environ. Toxicol. 15: 248–252.
Johnson, B.T.; Petty, J.D.; Huckins, J.N.; Lee, K. 2004, Hazard assessment of simulated oil spill on
intertidal areas of the St. Lawrence River with SPMD-TOX. Environ. Toxicol. 19: 329–335.
Kǒcı́, V.; Ocelka, T.; Kochánková, L. 2001a, SPMD–a modern approach to passive sampling. Vodni
Hospodarstvi 51: 331–332.
Kǒcı́, V.; Ocelka, T.; Kochánková, L. 2001b, SPMD–description of perspective method of POP moni-
toring in water bodies. Vodni Hospodarstvi 51: 345–348.
Kǒcı́, V.;Mlejnek, M.; Kochánková, L. 2003, Toxicological evaluations of exposed SPMD membranes.
Cen. Eur. J. Chem. 1: 28–34.
Kǒci, V.; Lukavsky, J.; Mlejnek, M.; Kochánková, L.; Grabic, R.; Ocelka, T. 2004a, Application of a
semipermeable membrane device (SPMD) for assessment of organic toxicants dangerous to the
green microalga Scenedesmus subspicatus. Arch. Hydrobiol. 150: 173–186.
Kǒcı́, V.; Ocelka, T.; Mlejnek, M.; Grabic, R. 2004b, Efficiency assessment of wastewater treatment
plant based on SPMD sampling. Cen. Eur. J. Chem. 2: 91–112.
Koistinen, J.; Lehtonen, M.; Tukia, K.; Soimasuo, M.; Lahtipera, M.; Oikari, A. 1998a, Identification of
lipophilic pollutants discharged from a Finnish pulp and paper mill. Chemosphere 37: 219–235.
Koistinen, J.; Soimasuo, M.; Tukia, K.; Oikari, A.; Blankenship, A.; Giesy, J.P. 1998b, Induction
of EROD activity in HEPA-1 mouse hepatoma cells and estrogenicity in MC-7 human breast
cancer cells by extracts of pulp mill effluents, sludge, and sediment exposed to effluents. Environ.
Toxicol. Chem. 17: 1499–1507.
Kolok, A.S.; Huckins, J.N.; Petty, J.D.; Oris, J.T. 1996, The role of water ventilation and sediment
ingestion in the uptake of benzo[a]pyrene in gizzard shad (Dorosoma cepedianum). Environ.
Toxicol. Chem. 10: 1752–1759.
Kot, A.; Zabiegata, B.; Namieśnik, J. 2000, Passive sampling for long-term monitoring of organic
pollutants in water. Trends Anal. Chem. 9: 446–459.
Kukkonen, J.V.K.; Rantalainen, A.-L.; Mikkelson, P.; Lyytikäinen, M.; Hamäläinen, H.; Passivirta,
J. 1998, Bioavailability of polychlorinated diphenylethers (PDPEs) in contaminated sediments:
Comparison of benthic organisms and semipermeable membrane devices (SPMDs). Organohal.
Comp. 39: 5–8.
K.R.; Schroeder, D.J.; Inmon, L.E. 1995, Use of the semipermeable membrane device (SPMD)
as an in situ sampler of waterborne bioavailable PCDD and PCDF residues at sub-part-per-
quadrillion concentrations. Environ. Sci. Technol. 29: 2886–2892.
Lebo, J.A.; Zajicek, J.L.; Orazio, C.E.; Petty, J.D.; Huckins, J.N.; Douglas, E.H. 1996, Use of the
semipermeable membrane device (SPMD) to sample polycyclic aromatic hydrocarbon pollution
in a lotic system. Polycyclic Aromat. Compd.8: 53–65.
Lebo, J.A.; Huckins, J.N.; Petty, J.D.; Ho, K.T.; Stern, E.A. 2000, Selective removal of organic con-
taminants from sediments: A methodology for toxicity identification evaluations (TIEs). Chemo-
sphere 40: 811–819.
208 Appendix B
Leppanen, H. and Kukkonen, V.K. 2000, Effect of sediment-chemical contact time on availability
of sediment-associated pyrene and benzo[a]pyrene to oligochaete worms and semipermeable
membrane devices. Aquat. Toxicol. 49: 227–241.
Lindström, A.; Buerge, I.J.; Poiger, T.; Bergqvist, P.A.; Müller, M.D.; Buser, H.R. 2002, Occurrence
and environmental behavior of the bactericide triclosan and its methyl derivative in surface waters
and in wastewater. Environ. Sci. Technol. 36: 2322–2329.
Lohmann, R.; Corrigan, B.P.; Howsak, M.; Jones, K.; Ockenden, W.A. 2001, Further developments
in the use of semipermeable membrane devices (SPMDs) as passive air samplers for persistent
organic pollutants: Field application in a spatial survey of PCDD/Fs and PAHs. Environ. Sci.
Technol. 35: 2576–2582.
Lu, Y.; Wang, Z.; Huckins, J.N. 2002, Review of the background of triolein-containing semipermeable
membrane devices in aquatic environmental study. Aquat. Toxicol. 60: 139–153.
Lu, Y. and Wang, Z. 2003, Accumulation of organochlorinated pesticides by triolein-containing
semipermeable membrane device (triolein-SPMD) and rainbow trout. Water Res. 37: 2419–
2425.
Luellen, D.A. and Shea, D. 2002, Calibration and field verification of semipermeable membrane devices
for measuring polycyclic aromatic hydrocarbons in water. Environ. Sci. Technol. 36: 1791–1797.
Luellen, D.A. and Shea, D. 2003, Semipermeable membrane devices accumulate conserved ratios of
sterane and hopane petroleum biomarkers. Chemosphere 53: 705–713.
Lyytikäinen, M.; Hirva, P.; Minkkinen, P.; Hämäläinen, H.; Rantalainen, A.-L.; Mikkelson, P.; Paa-
sivirta, J.; Kukkonen, J.V.K. 2003a, Bioavailability of sediment-associated PCDD/Fs and PCDEs:
Relative importance of contaminant and sediment characteristics and biological factors. Environ.
Sci. Technol. 37: 3926–3934.
Lyytikäinen, M.; Rantalainen, A.-L.; Mikkelson, P.; Hämäläinen, H.; Paasivirta, J.; Kukkonen, J.V.K.
2003b, Similarities in bioaccumulation patterns of polychlorinated dibenzo-p-dioxins and fu-
rans and polychlorinated diphenyl ethers in laboratory-exposed oligochaetes and semipermeable
membrane devices and in field-collected chironomids. Environ. Toxicol. Chem. 22: 2405–2415.
Ma, M.; Wang, Y.; Wang; Z. 1999, Sample pretreatment for toxic assessment of organic pollutants in
river water. Huanjing Kexue/Environ. Sci. 20: 80–83.
MacRae, J.D. and Hall, K.J. 1998, Comparison of methods used to determine the availability of
polycyclic aromatic hydrocarbons in marine sediment. Environ. Sci. Technol. 32: 3809–3815.
McCarthy, J.F.; Southworth, G.R.; Palmer, J.A.; Ham, K.D. 2000, Time-integrated, flux-based monitor-
ing using semipermeable membrane devices to estimate the contribution of industrial facilities
to regional polychlorinated biphenyl budgets. Environ. Toxicol. Chem. 19: 352–359.
McCarthy, K.A. and Gale, R.W. 2001, Evaluation of persistent hydrophobic organic compounds in
the Columbia River Basin using semipermeable membrane devices. Hydrological Processes 15:
1271–1283.
Meadows, J.C.; Tillitt, D.E.; Schwartz, T.R.; Schroeder, D.J.; Echols, K.R.; Gale, R.W.; Powell, D.C.;
Bursian, S.J. 1996, Organochlorine contaminants in double-crested cormorants from Green Bay,
Wisconsin: 1. Large-scale extraction and isolation from eggs using semi-permeable membrane
dialysis. Arch. Environ. Con. Tox. 31: 218–224.
Meijer, S.N.; Ockenden, W.A.; Steinnes, E.; Corrigan, B.P.; Jones, K.C. 2003, Spatial and temporal
trends of POPs in Norwegian and UK background air: Implications for global cycling. Environ.
Sci. Technol. 37: 454–461.
Metcalfe, T.L.; Metcalfe, C.D.; Bennett, E.R.; Haffner, G.D. 2000, Distribution of toxic organic con-
taminants in water and sediments in the Detroit River. J. Great Lakes Res. 26: 55–64.
Miege, C.; Durand, S.; Garric, J.; Gourlay, C.; Wang, D.; Mouchel, J.M.; Tusseau-Vuillemin,
M.H. 2004, Semipermeable membrane device-availability of polycyclic aromatic hydrocar-
bons in river waters and wastewater treatment plant effluents. Polycyclic Aromat. Compd.24:
805–825.
Norrgren, L.; Pettersson, U.; Örn, S.; Bergqvist, P.A. 2000, Environmental monitoring of the Kafue
River, located in the Copperbelt, Zambia. Arch. Environ. Con. Tox. 38: 334–341.
Ocelka, T.; Prazak, J.; Kǒcı́, J.; Raclavska, H.; Grabic, R. 2003, Application of the SPMD method for
POPs monitoring and toxicity assessment in waste waters. Vodni Hospodarstvi. 53: 211–214.
Ockenden, W.A.; Prest, H.F.; Thomas, G.O.; Sweetman, A.; Jones, K.C. 1998a, Passive air sampling
Ockenden, W.A.; Sweetman, A.J.; Prest, H.F.; Steinnes, E.; Jones, K.C. 1998b, Toward an under-
standing of the global atmospheric distribution of persistent organic pollutants: The use of
semipermeable membrane devices as time-integrated passive samplers. Environ. Sci. Technol. 32:
2795–2803.
Ockenden, W.A.; Corrigan, B.P.; Howsam, M.; Jones, K.C. 2001, Further developments in the use of
semipermeable membrane devices as passive air samplers: Application to PCBs. Environ. Sci.
Technol. 35: 4536–4543.
Pablos, M.V.; Fernandez, C.; Garcia-Hortiguela, P.; Valdovinos, C.; Tarazona, J.V. 1999, Comparison
of different extraction procedures for organic-fraction toxicity testing of urban sewages. Toxicol.
Environ. Chem. 70: 115–127.
Palowitch, A.W. 1996, Semipermeable membrane devices. Sea Technology 37: 73–79.
Parrott, J.L. and Tillitt, D.E. 1997, The use of Semipermeable Membrane Devices (SPMDs) to Concen-
trate Inducers of Fish Hepatic Mixed Function Oxygenase (MFO). In Ecotoxicology: Responses,
Biomarkers, and Risk Assessment. An OECD Workshop; Zelikoff, J.T., Ed.; SOS Publications:
Fair Haven, NJ.; pp. 185–196.
Parrott, J.L.; Backus, S.M.; Borgmann, A.I.; Swyripa, M. 1999, The use of semipermeable membrane
devices to concentrate chemicals in oil refinery effluent on the Mackenzie River. Arctic 52:
125–138.
Petty, J.D.; Huckins, J.N.; Martin, D.B.; Adornato, T.G. 1995a, Use of semipermeable membrane de-
vices (SPMDs) to determine bioavailable organochlorine pesticide residues in streams receiving
irrigation drainwater. Chemosphere 30: 1891–1903.
Kaiser, E.M. 1995b, Determination of bioavailable organochlorine pesticide residues in the
Lower Missouri River. Environ. Sci. Technol. 29: 2561–2566.
1998, Determination of bioavailable contaminants in the Lower Missouri River following the
Petty, J.D.; Jones, S.B.; Huckins, J.N.; Cranor, W.L.; Parris , J.T.; McTague, T.B.; Boyle, T.P. 2000a,
An approach for assessment of water quality using semipermeable membrane devices (SPMDs)
and bioindicator tests. Chemosphere 41: 311–321.
Petty, J.D.; Orazio, C.E.; Huckins, J.N.; Gale, R.W.; Lebo, J.A.; Meadows, J.C.; Echols, K.R.; Cranor,
W.I. 2000b, Considerations involved with the use of semipermeable membrane devices for
monitoring environmental contaminants. J. Chromatogr. A 879: 83–95.
210 Appendix B
contaminants.Chemosphere 54: 695–705.
Peven, C.S.; Uhler, A.D.; Querzoli, F.J. 1996, Caged mussels and semipermeable membrane devices
as indicators of organic contaminant uptake in Dorchester and Duxbury Bays, Massachusetts.
Poiger, T.; Buser, H.-R.; Balmer, M.E.; Bergqvist, P.-A.; Müller, M.D. 2004, Occurrence of UV fil-
ter compounds from sunscreens in surface waters: Regional mass balance in two Swiss lakes.
Prest, H.F.; Huckins, J.N.; Petty, J.D.; Herve, S.; Paasivirta, J.; Heinonen, P. 1995a, A survey of recent
results in passive sampling of water and air by semipermeable membrane devices. Mar. Pollut.
Bull. 31: 306–312.
Prest, H.F.; Jacobson, L.A.; Huckins, J.N. 1995b, Passive sampling of water and coastal air via semiper-
meable membrane devices. Chemosphere 30: 1351–1361.
Prest, H.F, Richardson, B.J.; Jacobson, L.A.; Vedder, J.; Martin, M. 1995c, Monitoring organochlorines
Prest, H.F.; Jacobson, L.A.; Wilson, M. 1997, Passive water sampling for polynuclear aromatic hy-
drocarbons using lipid-containing semipermeable membrane devices (SPMDs): Application to
contaminant residence times. Chemosphere 35: 3047–3063.
Rantalainen, A.-L.; Ikonomou, M.G.; Rogers, I.H. 1998a, Lipid-containing semipermeable-membrane
Rantalainen, A-.L.; Paasivirta, J.; Herve, S. 1998b, Uptake of chlorohydrocarbons from soil
by lipid-containing semipermeable membrane devices (SPMDs).Chemosphere 36: 1415–
1437.
Rantalainen, A.-L.; Hyötyläinen, T.; Saramo, M.; Niskanen, I. 1999, Passive sampling of PAHs in
indoor air in Nepal. Toxicol. Environ. Chem. 68: 335–348.
Rantalainen, A.-L.; Cretney, W.; Ikonomou, M.G. 2000a, Uptake rates of semipermeable membrane
devices (SPMDs) for PCDDs, PCDFs, and PCBs in water and sediment. Chemosphere 40: 147–
158.
Rantalainen, A.-L.; Crewe, N.F.; Ikonomou, M.G. 2000b, Comparison of three techniques for lipid
removal from seal blubber: Gel permeation, acid treatment, and dialysis with semipermeable
membrane. Int. J. Environ. An. Ch. 76: 31–47.
Rastall, A.C.; Neziri, A.; Vukovic, Z.; Jung, C.; Mijovic, S.; Hollert, H.; Nikcevic, S.; Erdinger,
L. 2004, The identification of readily bioavailable pollutants in Lake Shkodra/Skadar using
semipermeable membrane devices (SPMDs), bioassays and chemical analysis. Environ. Sci.
Pollut. R. 11: 240–253.
Rayne, S. and Ikonomou, M.G. 2002, Reconstructing source polybrominated diphenyl ether congener
patterns from semipermeable membrane devices in the Fraser River, British Columbia, Canada:
Comparison to commercial mixtures. Environ. Toxicol. Chem. 21: 2292–2300.
Richardson, B.J.; Zheng, G.J.; Tse, E.S.C.; Lam, P.K.S. 2001, A comparison of mussels (Perna viridis)
and semi-permeable membrane devices (SPMDs) for monitoring chlorinated trace organic con-
taminants in Hong Kong coastal waters. Chemosphere 45: 1201–1208.
Richardson, B.J.; Lam, P.K.S.; Zheng, G.J.; McClellan, K.E.; De Luca-Abbott, S.B. 2002, Biofouling
confounds the uptake of trace organic contaminants by semi-permeable membrane devices. Mar.
Pollut. Bull. 44: 1372–1379.
Richardson, B.J.; Zheng, G.J.; Tse, E.S.C.; De Luca-Abbott, S.B.; Siu, S.Y.M, Lam, P.K.S. 2003, A
comparison of polycyclic aromatic hydrocarbon and petroleum hydrocarbon uptake by mussels
(Perna viridis) and semi-permeable membrane devices (SPMDs) in Hong Kong Coastal Waters.
Environ. Pollut. 122: 223–227.
Robertson, G.; Lebowitz, M.; Needham, L.; O’Rourke, M.K.; Rogan, S.; Petty, J.D.; Huckins, J.N.
2002, Distribution of Residential Organochlorine Pesticide Residuals along the Arizona/Mexico
Border. Proceedings: 9th International Conference on Indoor Air Quality and Climate; Monterey,
CA, June 2002; pp. 63–68.
Rogers, H.R. 1997, Influence of suspended solids and back diffusion on organic contaminant uptake
by semi-permeable membranes (SPMDs). Chemosphere 35: 1651–1658.
Rohr, A.C.; Hall, E.R.; Hall, K.J. 1996, Use of semipermeable membrane devices for monitoring pulp
mill effluents: A preliminary assessment. Water Qual. Res. J. Can. 31: 85–100.
Sabaliūnas, D. and Sodergren, A. 1996, Uptake of organochlorine pesticides by solvent-filled cellulose
and polyethylene membranes. Ecotox. Environ. Safe. 35: 150–155.
Sabaliūnas, D. and Sodergren, A. 1997, Use of semipermeable membrane devices to monitor pollutants
in water and assess their effects: A laboratory test and field verification.Environ. Pollut. 96: 195–
205.
Sabaliūnas, D.; Ellington, J.; Neuman, J. 1998a, Semipermeable Membrane Devices for Monitoring
Pollutants in Aquatic Ecosystems of Lithuania. CRC Critical Reviews in Analytical Chemistry,
Vol. 2812; CRC Press: Boca Raton, FL.; p. 50.
Sabaliūnas, D.; Lazutka, J.; Sabaliūniene, I.; Sodergren, A. 1998b, Use of semipermeable membrane
devices for studying effects of organic pollutants: Comparison of pesticide uptake by semiper-
meable membrane devices and mussels. Environ. Toxicol. Chem. 17: 1815–1824.
Sabaliūnas, D.; Ellington, J.; Sabaliūniene, I. 1999, Screening bioavailable hydrophobic toxicants in
surface waters with semipermeable membrane devices: Role of inherent oleic acid in toxicity
evaluations. Ecotox. Environ. Safe. 44: 160–167.
Sabaliūnas, D.; Lazutka, J.R.; Sabaliūniene, I. 2000, Acute toxicity and genotoxicity of aquatic hy-
drophobic pollutants sampled with semipermeable membrane devices. Environ. Pollut. 109:
251–265.
Sabaliūnas, D.; Webb, S.F.; Hauk, A.; Jacob, M.; Eickhoff, W.S. 2003, Environmental fate of triclosan
in the River Aire Basin, UK.Water Res. 37: 3145–3154.
Scott, G.I.; Fulton, M.H.; Wirth, E.F.; Chandler, G.T.; Key, P.B.; Daugomah, J.W.; Bearden, D.; Chung,
K.W.; Strozier, E.D.; DeLorenzo, M.; Sivertsen, S.; Dias, A.; Sanders, M.; Macauley, J.M.;
Goodman, L.R.; LaCroix, M.W.; Thayer, G.W.; Kuclick, J. 2002, Toxicological studies in tropical
ecosystems: An ecotoxicological risk assessment of pesticide runoff in South Florida estuarine
ecosystems. J. Agr. Food Chem. 50: 4400–4408.
Shaw, M.; Tibbetts, I.R.; Müller, J.F. 2004, Monitoring PAHs in the Brisbane River and Moreton
Bay, Australia, using semipermeable membrane devices and EROD activity in yellowfin bream,
Acanthopagrus australis. Chemosphere 3: 237–246.
Shigenaka, G. and Henry, C.B. Jr. 1995, Use of Mussels and Semipermeable Membrane Devices To
Assess Bioavailability of Residual Polynuclear Aromatic Hydrocarbons Three Years After the
Exxon Valdez Oil Spill. In Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters; Wells,
P.G., Butler, J.N., Hughes, J.S, Eds.; American Society for Testing and Materials: Philadelphia,
PA.; pp. 239–260.
Shim, S.M.; Santerre, C.R.; Dorworth, L.E.; Miller, B.K.; Stahl, J.R.; Deardorff, D.C. 2004, Prediction
of PCB content in sportfish using semipermeable membrane devices (SPMDs). J. Environ. Sci.
Health, Part B 39: 263–271.
Söderström, H.S. and Bergqvist, P.-A. 2003, Polycyclic aromatic hydrocarbons in a semiaquatic plant
and semipermeable membrane devices exposed to air in Thailand. Environ. Sci. Technol. 37:
47–52.
Söderström, H.S. and Bergqvist, P.-A. 2004, Passive air sampling using semipermeable membrane
devices at different wind-speeds in situ calibrated by performance reference compounds. Environ.
Sci. Technol. 38: 4828–4834.
212 Appendix B
Strandberg, B.; Wagman, N.; Bergqvist, P.-A.; Haglund, P.; Rappe, C. 1997, Semipermeable membrane
devices as passive samplers to determine organochlorine pollutants in compost. Environ. Sci.
Technol. 31: 2960–2965.
526–533.
Stuer-Lauridsen, F. and Dahl, B. 1995, Tributyltin transport at a marine water/sediment interface-A
field study. Chemosphere 30: 831–845.
Stuer-Lauridsen, F. and Kjlholt, J. 2000, Identification of selected hydrophobic organic contaminants
in wastewater with semipermeable membrane devices (SPMDs). Water Res. 34: 3478–3482.
Sztamberek-Gola, I.; Grochowalski, A.; Chrzaszcz, R. 2003, Monitoring of PCDDs, PCDFs, and PAHs
in wastewater with use of semipermeable membrane devices (SPMD). Organohal. Comp. 60:
45–48.
Tandlich, R.; Vrana, B.; Balaz, S. 2003, Monitoring of Polychlorinated Biphenyls in Slovakian Freshwa-
ter Sediments: Use of Semipermeable Membrane Devices. In The Utilization of Bioremediation
to Reduce Soil Contamination: Problems and Solutions; Sasek, V., Glaser, J., Baveye, P. Eds.;
Kluwer Academic Publishers: Dordrecht, The Netherlands; 221–226.
Utvik, T.I.R.; Durell, G.S.; Johnsen, S. 1999a, Determining produced water originating polycyclic
aromatic hydrocarbons in North Sea waters: Comparison of sampling techniques. Mar. Pollut.
Bull. 38: 977–989.
Utvik, T.I.R. and Johnsen, S. 1999b, Bioavailability of polycyclic aromatic hydrocarbons in the North
Sea. Environ. Sci. Technol. 33: 1963–1969.
Van der Oost, R.; Beyer, J.; Vermeulen, N.P.E. 2003, Fish bioaccumulation and biomarkers in environ-
mental risk assessment: A review. Environ. Toxicol. Phar. 13: 57–149.
Van Stee, L.L.P.; Leonards, P.E.G.; van Loon, W.M.G.M.; Hendriks, A.J.; Maas, J.L.; Struijs, J.;
Brinkman, U.A.Th. 2002, Use of semi-permeable membrane devices and solid-phase extraction
for the wide-range screening of microcontaminants in surface waters by GC-AED/MS. Water
Res. 36: 4455–4470.
bioavailable PAH, PCB, and OCP concentrations in water using semipermeable membrane de-
vices (SPMDs), sediments and caged carp. Chemosphere 11: 1675–1689.
Villeneuve, D.L.; Crunkilton, R.L.; DeVita, W.M. 1997, Aryl hydrocarbon receptor-mediated toxic
potency of dissolved lipophilic organic contaminants collected from Lincoln Creek, Milwaukee,
Wisconsin, USA, to PLHC-1 (Poeciliopsis lucida) fish hepatoma cells. Environ. Toxicol. Chem.
16: 977–984.
Villeneuve, J.Y.; Niimi, A.J.; Metcalfe, C.D. 2000, Distribution and bioaccumulation of chlorinated
diphenyl ethers in a contaminated embayment of Lake Ontario. J. Great Lakes Res. 25: 760–771.
Voie, Ø.A.; Johnsen, A.; Rossland, H.K. 2002, Why biota still accumulate high levels of PCB after
removal of PCB contaminated sediments in a Norwegian Fjord Chemosphere 46: 1367–1372.
Vrana, B.; Paschke, A.; Popp, P. 2001a, Polyaromatic hydrocarbon concentrations and patterns in
sediments and surface water of the Mansfield Region, Saxony-Anhalt, Germany. J. Environ.
Monitor. 3: 602–609.
Vrana, B.; Paschke, A.; Popp, P.; Schüürmann, G. 2001b, Use of semipermeable membrane devices
(SPMDs): Determination of bioavailable organic contaminants in the industrial region of Bitter-
feld, Saxony-Anhalt, Germany. Environ. Sci. Pollut. R. 8: 27–34.
Wågman N, Strandberg B, Tysklind M. 1998, Passive sampling of PCBs from in—and outdoor air with
semipermeable membrane devices (SPMDs). Organohal. Comp. 35: 209–212.
Wang, C.; Wang, Y.; Kiefer, F.; Yediler, A.; Wang, Z.; Kettrup, A. 2003, Ecotoxicological and chemical
characterization of selected treatment process effluents of municipal sewage treatment plant.
Ecotox. Environ. Safe. 56: 211–217.
Wang, Y.; Wang, C.; Wang, Z. 1998, Uptake of moderately hydrophobic chlorophenols from water by
semipermeable membrane devices (SPMDs) and by goldfish (Crassius auratus). Chemosphere
37: 327–339.
Wang, Y.; Wang, Z.; Liu, J.; Ma, M.; Belzile, N. 1999a, Monitoring priority pollutants in the Yangtze
River by dichloromethane extraction and semipermeable membrane device (SPMD). Chemo-
sphere 39: 113–131.
Wang, Y.; Wang, Z.; Wang, C.; Wang, W. 1999b, Uptake of moderately hydrophobic nitroaromatics
from water by semipermeable membrane devices (SPMDs) and by goldfish (Crassius auratus).
Wang, Y.; Wang, Z.; Liu, J.; Ma, M.; Wang, C; Mo, Z. 2001, Monitoring priority pollutants in a sewage
treatment process by dichloromethane extraction and triolein-semipermeable membrane devices
(SPMDs). Chemosphere 43: 339–346.
Wang, Y.; Huang, Y.; Huckins, J.N.; Petty, J.D. 2004, Compound-specific carbon and hydrogen isotope
analysis of sub-parts per billion level waterborne petroleum hydrocarbons. Environ. Sci. Technol.
38: 3689–3697.
Wang, Z.; Lu, Y. Wang, Y.; Ma, M. 2002a, Assessing the ecological risk of substituted benzenes in
Huaihe River, China. Acta Sci. Circumstantiae 22: 300–304.
Wang, Z.; Wang, Y.; Ma, M.; Lu, Y.; Huckins, J.N. 2002b, Use of triolein-SPMD to assess the biocon-
centration processes of hydrophobic organic contaminants in the Huaihe River, China. Environ.
Toxicol. Chem. 21: 2378–2384.
Whyte, J.J.; Karrow, N.A.; Boermans, H.J.; Dixon, D.G.; Bols, N.C. 2000, Combined methodolo-
gies for measuring exposure of rainbow trout (Oncorhynchus mykiss) to polycyclic aromatic
hydrocarbons (PAHs) in creosote contaminated microcosms. Polycyclic Aromat. Compd. 18:
71–98.
Williamson, K.S.; Petty, J.D.; Huckins, J.N.; Lebo, J.A.; Kaiser, E.M. 2002a, HPLC-PFD determi-
nation of priority pollutant PAHs in water, sediment, and semipermeable membrane devices.
Williamson, K.S.; Petty, J.D.; Huckins, J.N.; Lebo, J.A.; Kaiser, E.M. 2002b, Sequestration of pri-
ority pollutant PAHs from sediment pore water employing semipermeable membrane devices.
Zabiegala, B.; Kot, A.; Namieśnik, J. 2000, Long-term monitoring of organic pollutants in water–
Application of passive dosimetry. Chem. Anal-Warsaw 45: 645–657.
Zajicek, J.L.; Tillitt, D.E.; Huckins, J.N.; Petty, J.D.; Potts, M.E.; Nardone, D.A. 1996, Application of
Enzyme-Linked Immunosorbent Assay (ELISA) for Measurement of Polychlorinated Biphenyls
(PCBs) from Hydrophobic Solutions: Extracts of Fish and Dialysates of Semipermeable Mem-
brane Devices (SPMDs). In Environmental Immunochemical Methods; ACS Symposium Series
646; American Chemical Society: Washington, D.C.; Chapter 26, pp. 307–325.
Zhang, J. and Zhang, Y. 2001, Development and application of SPMD technology to environmental
research. Mar. Environ. Sci. 20: 67–74.
Zhu, Y.-X.; Zhang, Y.; Zhang, J.; Zhuang, Y.-T.; Poon, K.-F.; Lam, M.H.W.; Hong, H.-S.; Wu, R.S.S.
2001, Monitoring polynuclear aromatic hydrocarbons in sediment porewater by SPMD. Chin. J.
Oceanol. Limnol. 19: 382–388.
Zimmerman, L.R.; Thurman, E.M.; Bastian, K.C. 2000, Detection of persistent organic pollutants in
the Mississippi Delta using semipermeable membrane devices. Sci.Total Environ. 248: 169–179.
Glossary
The units specified for symbols in the table below are only recommendations.
Other units may (and sometimes should) be used, as long as the units on the
right-hand sides of the relevant equation matches the units on the left-hand sides.
For example, if time is measured in days in Eq. 3.22, and Rs is given in L d−1 ,
then the SPMD volume should be given in L, rather than in cm3 , because of the
argument that the exponent (Rs t/Vs K sw ) should be dimensionless. Similarly, when
the absorbed amount is given in ng, and the SPMD volume is given in L, then the
aqueous concentration calculated from Eq. 3.22 has units of ng L−1 .
Recommended
Symbol Quantity unitsa
A Surface area cm2

Bm Antilog of the log km − log K ow intercept cm d−1
Bw Antilog of the log kw − logK ow intercept cm d−1
CB Concentration in a biomonitoring organism g g−1
Cc Concentration of concern ng cm−3
Cf Concentration in food g g−1
CL Concentration in the triolein ng cm−3
Cs Concentration in SPMD ng cm−3
Ct Total (dissolved + DOC-bound) concentration in the water ng cm−3
Cw Aqueous concentration ng cm−3
CF Concentration factor cm3 cm−3
C.V. Coefficient of variation %
d Characteristic length scale cm
(Continued )
215
216 Glossary
Glossary (Continued)
Recommended
Da Diffusion coefficient in air cm2 d−1

Db Diffusion coefficient in the biofilm cm2 d−1
Di Diffusion coefficient in the rate-limiting barrier cm2 d−1
Dm Diffusion coefficient in the membrane cm2 d−1
Dw Diffusion coefficient in water cm2 d−1
[DOC] Concentration of dissolved organic carbon g cm−3
E a Activation energy J mol−1
Ev Volume of sample medium cleared of chemical mL
Et Fraction of total sample injected —
F Feeding rate g g−1 d−1
fr Faction of quickly equilibrating particle-bound contaminants —
Fw Groundwater flow rate m3 d−1
H Henry’s law constant Pa m3 mol−1
h Hydraulic head m
I Transport resistance (1/k) d cm−1
jb Flux through the biofilm ng cm−2 d−1
jm Flux through the membrane ng cm−2 d−1
jw Flux through the water boundary layer ng cm−2 d−1
jx Flux in the x direction ng cm−2 d−1
K Bulk sediment-water partition coefficient cm3 cm−3
k2 Sediment desorption rate constant d−1
ka Mass transfer coefficient for the air boundary layer cm d−1
kb Mass transfer coefficient for the biofilm cm d−1
K bw Biofilm-water partition coefficient cm3 cm−3
K DOC DOC-water partition coefficient cm3 g−1
ke Equilibration rate constant or release rate constant d−1
kea Equilibration rate constant for air d−1
kem Rate constant for HOC metabolism d−1
kew Equilibration rate constant for water d−1
kLd Mass transfer coefficient for the lipid-derived film cm d−1
K Lda “Lipid derived film” to air partition coefficient cm3 cm−3
K Lm Lipid-membrane partition coefficient cm3 cm−3
km Mass transfer coefficient for the membrane cm d−1
K ma Membrane-air partition coefficient cm3 cm−3
K mL Membrane-lipid partition coefficient cm3 cm−3
K mw Membrane-water partition coefficient cm3 cm−3
ko Overall mass transfer coefficient cm d−1
K oa Octanol-air partition coefficient cm3 cm−3
K ow Octanol-water partition coefficient cm3 cm−3
kp Mass transfer coefficient for a polymer film cm d−1
K pa Polymer-air partition coefficient cm3 cm−3
K pw Polymer-water partition coefficient cm3 cm−3
K sa SPMD-air partition coefficient cm3 cm−3
K sw SPMD-water partition coefficient cm3 cm−3
ku Uptake rate constant (CRK model) cm3 cm−3 d−1
kw Mass transfer coefficient for the water boundary layer cm d−1
(Continued)
Glossary 217
(Continued)
Recommended
Lf Length of water-bearing strata m

M Molar mass g mol−1
mL Mass of lipid g
m Leq Lipid-equivalent mass g
mm Mass of membrane g
m oc Mass of sedimentary organic carbon g
Mwb Whole body tissue mass g
N Absorbed amount ng
n Number of observations —
N0 Amount prior to exposure ng
Phc Hydraulic permeability m d−1
Pr Procedural recovery of analyte %
Q DOC quality (K DOC /K ow ratio) cm3 g−1
R Gas constant J mol−1 K−1
r Correlation coefficient (regression models) —
Re Reynolds number —
Rs Sampling rate cm3 d−1
Rv Ventilation rate across gills cm3 d−1
S Fractional assimilation efficiency across gut g g−1
s Standard deviation same as the
dependent variable
Sa Solubility coefficient for an amorphous polymer cm3 cm−3
Sc Schmidt number —
Sh Sherwood number —
sn Number of standard SPMDs per mL carrier solvent SPMD mL−1
Sp Solubility coefficient cm3 cm−3
T Absolute temperature K
t Time d or s
t1/2 First order half-life d
t95 Time to 95% of equilibrium d
tL Lag time d or s
tm Residence time d or s
tr Response time d or s
V LeBas molar volume or unit volume cm3 mol−1
ν Kinematic viscosity cm2 s−1
Vc Volume of carrier solvent used in a bioassay mL
Vgw Well water volume cm3
Vi Volume of standard injected µL
VL Lipid or triolein volume cm3
Vm LeBas molar volume cm3 mol−1
Vp Polymer volume cm3
Vs SPMD volume cm3
VT Tissue volume cm3
Vtox Volume of SPMD per day eliciting a toxicological response mL d−1
Vw Water volume cm3
Vw−tox Volume of water extracted eliciting a toxicological response cm3
α Compound-specific effect on the sampling rate —
(Continued)
218 Glossary
Glossary (Continued)
Recommended
β Exposure-specific effect on the sampling rate —

δi Thickness of rate-limiting barrier cm
δb Biofilm thickness cm
δm Membrane thickness cm
δp Polymer film thickness cm
δw Effective water boundary layer thickness cm
φ Porosity —
φa Amorphous volume fraction of polymer cm3 cm−3
θ Tortuosity —
π 3.14159 . . . —
a As discussed above other units may be used.

Index
Activation energies, mass transfer, 55, 56 Bioassays of SPMD (Cont.)

Active Sampling, definition of and methods, ethoxyresorufin-O-deethylase (EROD)
3–6 activity, 121, 127, 128, 151, 179
Air boundary layer (ABL): see Boundary layer Microtox, 121, 124, 125, 126
Air diffusion coefficient: see Diffusion Mutatox, 91, 121, 124, 125, 127
coefficients sister chromatid exchange, 121
Air flow, effects on sampling: see Exposure vitellogenin production (VGT), 121, 131
conditions yeast androgen screen, 179
Air sampling rates: see Sampling rates (Rs ), yeast estrogen screen (YES), 113, 121, 135,
SPMD 179, 180
Air-water partition coefficient: see Partition Bioavailability, definition of, 3
coefficients Bioconcentration, definition of, 2
Ames mutagenicity test: see Bioassay of SPMD Bioconcentration factor (BCF), 2, 34, 35, 55,
extracts 122, 141–143, 158, 159, 161
Amorphous region: see Rubbery region, polymer Biofilm or periphyton, 21, 22, 45, 47, 69, 71, 72,
Arrhenius equation, 55 108
Aqueous boundary layer: see Boundary layer Biofouling, effects on sampler performance and
Artifacts, procedural, 81, 117 control of, 14, 21, 22, 72, 75
Biomagnification, 116, 153, 160
Bioaccumulation, defination of and parameters, Biomagnification factor (BMF), 160
2, 122, 143, 158 Biomarkers or bioindicators: see Bioassay of
Bioaccumulation factor (BAF), 6, 140–144, 153, SPMD extracts
158, 160, 161 Biomimetic, 32, 35, 160, 161
Bioassays of SPMD extracts (also see Toxicity Biomonitoring organisms (BMOs), 6, 139–162
endpoints): Blanks, QC samples for SPMD analysis:
Ames mutagenicity assay, 121 fabrication, 104, 106, 111
Daphtoxkit F, 121 field, 97, 105, 111, 114, 135
enzyme-linked immunosorbent assay process, 104, 106, 111, 135
(ELISA), 121, 134 reagent, 105
219
220 Index
Blanks, QC samples for SPMD analysis: (Cont.) Fabrication Blank: see Blanks
trip, 105 Fick’s law, 38, 40, 49
Boundary layer, passive sampler: Field blank: see Blanks
air (ABL), 9, 23, 77–79, 148 First-order kinetics: see Kinetics
water (WBL), 15, 16, 19, 21, 39, 45–48, 60, Fish lipid: see SPMD liquid phases
63–65, 69, 72, 73, 141, 146, 148, 149 Flow-turbulence: see Exposure conditions;
effects on sampling
Calibration data: see SPMD Flux, equations for, 47, 71, 72
Chemical reaction kinetic (CRK) models: see Food chain, 116, 154, 160
Mathematical models Free volume of polymer, 11, 15, 21, 30, 63
Clearance capacity, 8, 13 Fugacity, 2, 7, 8
Concentration factor (CF), 18, 153, 154
Conductivity, definition of and models, 48, 69, Gas constant, 56
72 Genotoxicity: see Toxicity endpoints
Crystallinity, polymer, 14 Gill extraction efficiencies, 143, 148
Curvilinear uptake phase: see Uptake phases Gills or respiratory lamellae, fish, 3, 143, 145,
147, 148, 161
Daphtoxkit F: see Bioassays of SPMD extracts Goulden large sample extractor, 5
Darcy’s Law, 74 Grab sampling, 3–5
Deployment devices or canisters: see SPMD Groundwater sampling, 34, 35, 74, 75
Deployment precautions: see SPMD Gut assimilation (also see dietary uptake), 143,
Depuration rate constant: see Rate constants 148
Dialysis, organic solvent (OSD): see SPMD
Dietary uptake, 143, 154, 160, 162 Half-life (t1/2 ), 36, 91, 155, 159
Diffusion coefficients: Hayduk-Laudie equation, 65
air, 72, 75–77, 79 Henry’s law constants (H ), 75–79
biofilm, 72 High-volume air sampler (HiVol), 6, 10, 76, 81,
membrane or LDPE or polymer, 41, 63, 68 198
water, 71 Hildebrand and Hansen solubility parameters,
Dissipation rate constant: see Rate constants 11, 20
Dissolved organic carbon (DOC), 4, 51–53, 66,
68, 69 Immunoassays, 121, 134
DOC-water partition coefficient: see Partition Impedance (Io ), 48
coefficients Impurities (SPMD): see Triolein impurities and
Polyethylene waxes
Elimination rate constant: see Rate constants Infiltrix sampler, 5, 6
Empore extraction disk, 4, 12 Integrative sampling, definition of, 6, 8, 37, 122
Encounter-volume, 3, 32, 33, 39 Interferences, bioassays, 134, 135
Endocrine disruption: see Toxicity endpoints Isotropic exchange, 21
Enzyme-linked immunosorbent assay (ELISA):
see Bioassay of SPMDs Kinematic viscosity, 65
Episodic exposure events, 122, 157 Kinetics:
Equilibrium partition (EP) theory, 116, 140–142, first-order, 14, 19, 36, 40, 143, 153
153, 161 Michaelis-Menten, 143
Equilibrium samplers, 4, 8, 15, 49, 50
Estimation of ambient concentration, 7, 14, 21, Lag time, membrane diffusion, 41
29, 34, 35, 51, 62, 69, 70, 73, 76, 78 LDPE strips: see Passive samplers
Exposure adjustment factor (EAF), 59 LDPE-water partition coefficient: see Partition
Exposure conditions, effects on sampling rates: coefficients
biofouling, 12, 14, 19, 21, 22, 69, 72, 75, 96 LeBas molar volume, 56, 65
flow-turbulence, 10, 14, 19, 21, 58, 60, 63–65, Lipid-derived or surficial film, SPMD, 75, 80
76, 78, 96, 98, 147, 148 LDPE-water partition coefficient: see Partition
temperature, 14, 21, 55, 57, 79, 114, 140 coefficients
Index 221
Low-density polyethylene (LDPE) membrane, Partition coefficients (Cont.)

9, 11, 12, 18–20, 30, 63, 88, 89, 91 polymer-air (also see Partition coefficients,
membrane-air), 8
Mass transfer coefficient (MTC) models: see polymer-water (also see Partition coefficients,
Mathematical models membrane-water), 16
Mathematical models: sediment organic carbon-water (K oc ), 15, 16,
chemical Reaction Kinetics (CRK), 46, 49 74
mass Transfer Coefficient (MTC), 47, 49 sediment-water (K ), 71, 73
Membrane-air partition coefficient: see Partition SPMD-air (K sa ), 40, 75, 93
coefficients SPMD-water (K sw ), 35, 37, 39, 47, 48, 50,
Membrane-lipid partition coefficient: see 53–55, 74, 76, 93, 142, 153, 158, 159,
Partition coefficients 161, 183, 184–186
Membrane rate control: see Rate control triolein-membrane (also see Partition
Membrane-water partition coefficient: see coefficients, membrane-lipid), 19, 49
Partition coefficients Passive samplers:
Method detection limits (MDL), 114 PES (LDPE or LDPE strips), 12–14, 16–20,
Method quantitation limit (MQL), 5, 90, 114 35
Methyl oleate: see Triolein impurities PDM (personal diffusional monitors), 8, 9, 15
Michaelis-Menten kinetics: see Kinetics PISCES (passive in situ concentration and
Microtox: see Bioassay of SPMDs extraction sampler), 12, 13, 19
Mixed function oxygenase (MFO), 121, 127 POCIS (polar organic chemical integrative
Molecular diffusion coefficient: see Diffusion sampler), 12, 13, 33
coefficients POG (polymer coated glass), 8, 9, 13, 15, 16
Mussel-SPMD comparisons, 148, 150–153, 174 PSS (passive sampling system), 12, 13
Mutatox: see Bioassay of SPMDs PUF (polyurethane foam) disk, 10
Silastic SPMD, 14, 18–20
Narcosis: see Toxicity endpoints SPME (solid-phase micro extraction), 4, 13,
Neurotoxicity: see Toxicity endpoints 15–17
Nonporous polymers, definition of, 10 ST (silastic or silicone tubing/sheets of
PDMS), 11, 14, 15
Octanol-air partition coefficient: see Partition Passive sampling, definition of, 7
coefficients PDMs: see Passive samplers
Octanol-water partition coefficient: see Partition Performance reference compounds (PRCs) for
coefficients passive sampling, 14, 15, 21, 22, 35, 39,
Oleic acid: see Triolein impurities 50, 51, 59, 60, 68, 69, 72, 81, 92, 93, 98,
Organic solvent dialysis (OSD), see SPMD 105, 114
dialysis Periphyton: see Biofilm
Oyster-SPMD comparisons, 149, 153–155, Permeability, polymer and groundwater, 10, 35,
157 39, 74, 80
Pervaporation, 9, 87
Particulate organic carbon (POC), 4, 16, 51 Photolysis, effects of sunlight or solar radiation
Partition coefficients: on chemicals, 90, 91, 97
air-water (H/RT), 76, 170 PISCES: see Passive samplers
biofilm-water, 47, 72 POCIS: see Passive samplers
DOC-water, 52 POG: see Passive samplers
lipid- or triolein-water (K Lw ), 19, 29, 48, 142 Polar organic chemical integrative sampler
membrane-air (K ma ), 14, 23, 79 (POCIS): see Passive samplers
membrane (LDPE)-water (K mw ), 14, 39, 60, Polydimethyl siloxane (PDMS) samplers: see
61, 68, 79, 146 Passive samplers
membrane-lipid (K mL ), 115, 116 Polymer-air partition coefficient: see Partition
octanol-air (K oa ), 2, 33, 76, 115 coefficients
octanol-water (K ow ), 2, 29, 51, 52, 122 Polymer coated glass (POG): see Passive
particle-water, 51 samplers
222 Index
Polymer diffusion coefficient: see Diffusion Sediment organic carbon-water partition

coefficients coefficient: see Partition coefficients
Polymer-water partition coefficient: see Partition Sediment-water partition coefficient: see
coefficients Partition coefficients
Pore water sampling, 16, 34, 35, 72–74 Semipermeable membrane device (SPMD): see
Principle components analysis (PCA), SPMD SPMD
samples, 151, 152 Semi-volatile organic compounds (SVOCs), 5,
Process Blank: see Blanks 8, 9, 36
PSS: see Passive samplers Sherwood number (Sh), 64
PUF disk: see Passive samplers Shielding from sunlight or solar radiation,
importance of, 91
Quality control (QC), SPMDs, 87, 93, 98, Silicone fluid as a liquid phase for SPMDs, 20,
103–107 29
Quantitation limit: see Method quantitation limit Size exclusion chromatography (SEC), 88, 101,
Quantitative-structure-activity relationships 109–113, 127, 129
(QSARs), 1, 2 Solid-phase microextraction (SPME): see
Quantitative-structure-property relationships Passive samplers
(QSPRs), 2, 3 Solubility coefficient, polymer (Sp ), 14
SPMD:
Rate constants: analysis, 103–113, 171, 174–177
depuration or elimination (ke ; organisms), applications, 33–35, 98, 169
141–144, 156 assembly or preparation, 87–89
desorption (sediment), 74 biofouling: see Exposure conditions and
release or elimination (ke ; passive samplers), Biofilm
9, 37, 39, 46, 51, 114, 155, 156 blanks: see Blanks
uptake (ku ; passive samplers and organisms), calibration data, 50, 51, 57, 66, 81, 183–
22, 46, 49, 117, 141, 142, 144, 146, 148, 200
151, 161, 185–200 cleaning, external membrane surface, 88, 108,
Rate control or limiting (a step in or barrier to 109, 111
chemical exchange): cleanup, extracts or dialysates, 102–104,
air boundary layer (ABL), 9, 77–79 109–112, 127, 135
membrane, 8, 12, 15, 39, 41, 60–63, 77–79, comparison to BMOs, 139, 141, 144, 147,
146, 148 150, 153, 154, 158, 161, 173, 175, 177
sediment desorption, 73 comparison to other passive samplers, 13, 14,
water boundary layer (WBL), 39, 40, 63, 69, 17, 19, 20
146, 148 contamination, sample (also see Quality
Reagent blank: see Blanks control), 87, 96–98 104, 105, 114
Release rate constant: see Rate constants data quality and format, 34, 107, 113–117
Residence time (tm ), 40, 157 density, 49, 183
Resistance to mass transfer, equations of, 48 deployment devices or canisters, 10, 79,
Response time (tr ), 40 90–96
Reynolds number (Re), 14, 64 deployment considerations/precautions (also
Rubbery regions, polymer, 10 see Quality control), 87, 89–97
dialysis, organic solvent (OSD), 20, 21, 30,
Sample preservation, 88, 93 34, 107–109, 123, 127
Sample set, 104 handling precautions (also see Quality
Sampling rates (Rs ), SPMD: control), 95, 97
air, 23, 75, 81, 198–200 liquid phase, 18, 20, 29
water, 39, 50, 51, 57, 59–61, 63, 65, 68, 72, partition coefficients: see Partition
146, 159, 185–197 coefficients
Schmidt number (Sc), 64 processing, sample, 103, 104, 106, 107
Sea surface microlayer (SSM), sampling, 170 sampling rates: see Sampling rates (Rs )
Index 223
SPMD: (Cont.) Triolein impurities

standard design specifications, 19, 20, oleic acid, 20, 23, 80, 111, 112, 134, 135
89 methyl oleate, 20, 23, 80, 111, 112, 135
storage (stored, also see Sample preservation), Triolein-membrane partition coefficient: see
88, 93, 94 Partition coefficients
tether loops, 89, 96 Triolein-water partition coefficient: see Partition
transport, 93, 94 coefficients
types of chemicals sampled, 30, 32, 35, 96, Trip blank: see Blanks
101, 171, 172
SPME: see Passive samplers Uptake rate constant: see Rate constants
ST: see Passive samplers Uptake kinetics: see Kinetics
Steric impedance or size discrimination, Uptake phases
polymer, 30, 32, 147 curvilinear, 36, 37, 40, 93
Surface-area-to-sorbent-volume ratios (AV 1 ), equilibrium, 36, 37, 40, 49, 50
9–13, 16, 18, 89 linear, 36, 37, 40, 49, 50
Temperature effects: see Exposure Ventilation rate, gills, 143, 147, 148
conditions Vitellogenin (VGT): see Bioassay of SPMDs)
Time-weighted average (TWA) concentration, Volatile organic compound (VOC), 5, 8, 9, 17
definition, 6
Toxicity endpoints (also see Bioassays of SPMD Water or aqueous boundary layer (WBL); see
extracts) Boundary layer
deformities, frogs, 132–134 Water quality, effects on sampling, 30
endocrine disruption, 131, 180
genotoxicity, 125–127 XAD resins, 5, 9, 52
narcosis, 126
neurotoxicity, 130, 131 Yeast androgen screen: see Bioassay of SPMD
Toxicity identification evaluation (TIE), 34, extracts
132 Yeast estrogen screen (YES): see Bioassay of
Triolein cleanup method, 109, 112 SPMD extracts

Moni OC Env

Uploaded by

Copyright:

Available Formats

Moni OC Env

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Moni OC Env

Uploaded by

Copyright:

Available Formats

Monitors of Organic Chemicals

Kees Booij, Ph.D.

Cover photos: (Left) Deployment of SPMDs under Antarctica ice, courtesy of

Library of Congress Control Number: 2005933715

Printed on acid-free paper.

Printed in the United States of America. (TB/MVP)

Modern, industrialized societies depend on a wide range of chemical substances

partitioned quantities can be challenging but possible when appropriate methods

The authors have been intimately involved in conducting research to address

1. Introduction to Passive Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

3. Theory and Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.8 Groundwater Samplng . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

5. Analytical Chemistry Related to SPMDs . . . . . . . . . . . . . . . . . . . . . . 101

6. Bioassay of SPMD Extracts or Diluents . . . . . . . . . . . . . . . . . . . . . . . 121

7. Comparisons to Biomonitoring Organisms . . . . . . . . . . . . . . . . . . . . 139

7.4 Relative Amounts Accumulated . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

8. Selected Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Appendix A:. SPMD Calibration Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

Appendix B:. SPMD Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203

1.1. THE NEED FOR PASSIVE IN SITU SAMPLERS

Tens of thousands of chemicals are in commercial production throughout the

characteristics or measured physiochemical properties to endpoints such as bio-

by indirect measurements (e.g., via sorbents). At this juncture, a deﬁnition of

1.1.1. Active Sampling Methods for Water and Air

in all waterborne phases, which includes microorganisms and algae, particulate

requires labor-intensive multiple sample collections. Generally, the volume of col-

1.1.2. Biomonitoring Organisms for Water and Air

1.2. PASSIVE SAMPLER DEVELOPMENT

In this work, we deﬁne “passive” samplers as human-made devices where

site-speciﬁc exposure conditions must be taken into account. Unfortunately, ef-

where K pa is the polymer-air partition coefﬁcient, A is the surface area, δ p is ﬁlm

vapor-phase chemical for a speciﬁed volume of sorbent or partitioning phase, and

SPMDs SPMEs POGs LDPE strips PISCES POCIS PSSs

integrative equilibrium equilibrium integrative integrative integrative integrative

Chemicals HOCs HOCs HOCs HOCs HOCs HPOCsa HOCs,

sampled (SVOCs) (VOCs, SVOCs) (SVOCs) (SVOCs) (SVOCs) HPOCs

Sampling rate EBLb EBL EBL EBL EBL EBL diffusive

Sorbent-liquid 5,000 µL 0.28 µL 29 µL 2,000 µL 2.0 × 105 µL 240 µL 145 µL C18

According to the polymer literature (Flynn and Yalkowsky, 1972; Rogers,

carbon-water partition coefﬁcient, p represents the SPME polymeric phase, and

1.2.3. Lipid-Containing SPMDs and Closely Related Devices

1.2.3.1. Aqueous Phase Sampling

uptake of phenanthrene by replicate SPMDs was within 25% of the ke determined

1.2.3.2. Atmospheric Sampling

1.3. TOPICS COVERED IN SUBSEQUENT CHAPTERS

In subsequent chapters, we provide an overview of SPMD fundamentals and

2.1. SPMD DESCRIPTION AND RATIONALE

relatively high surface-area-to-sorbent-volume (AV −1 ) ratios, and solutes in the

Conceptually, SPMD data ﬁlls a gap between exposure assessments based on

2.2. APPLICABILITY OF SPMDs

Although SPMDs concentrate a very wide range of hydrophobic organic

TABLE 2.1 Classes or Speciﬁc Chemicals Known to Concentrate in SPMDs

priority pollutant PAHs and alkylated PAHs chlorinated dibenzodioxins, including

FIGURE 2.2 Various applications of SPMDs reported in the literature.

biomarkers), immunoassay, and toxicity identiﬁcation evaluation; 5) estimation

2.3. ACCUMULATION OF CHEMICALS BY SPMDs

Although “Theory and Modeling” is more extensively discussed in Chapter 3,

chemicals, and “rate constants” that are independent of chemical concentration.

2.4. PASSIVE SAMPLER FUNDAMENTALS AND TERMINOLOGY