Anda di halaman 1dari 334

FATE and

TRANSPORT of
HEAVY METALS
in the
VADOSE ZONE
Edited by
H. Magdi Selim
Iskandar K. Iskandar
Project Editor:
Acquiring Editor
Marketing Managers:
Cover design:
Manufacturing Manager:
Sylvia Wood
Skip DeWall
Bamara Glunn / Jane Stark
Jonathan Pennell
Carol Slatter
Library of Congress Cataloging-in-Publication Data
Fate and transport of heavy metals in the vadose zone / edited by H.M. Selim, 1.K.
Iskandar
p. cm.
Includes bibliographical references and index.
ISBN 0-8493-4112-4 (alk. paper)
1. Soils-Heavy metal content. 2. Heavy metals-Environmental aspects
3. Zone of aeration. 1. Selim, Hussein Magdi Eldin, 1944- .II. Iskandar, 1.K.
(Iskandar Karam), 1938- .
S592.6.H43F37 1999
628.5'5-dc21 98-26915
CIP
This book contains information obtained from authentic and highly regarded sources. Reprinted
material is quoted with permission, and sources are indicated. A wide variety of references are listed.
Reasonable efforts have been made to publish reliable data and information, but the author and the
publisher cannot assume responsibility for the validity of all materials or for the consequences of their use.
Neither this book nor any part may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, microfilming, and recording, or by any information
storage or retrieval system, without prior permission in writing from the publisher.
All rights reserved. Authorization to photocopy items for internal or personal use, or the personal or
internal use of specific clients, may be granted by CRC Press LLC, provided that $.50 per page photo-
copied is paid directly to Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923
USA. The fee code for users of the Transactional Reporting Service is ISBN 0-8493-9469-
4/99/$0.00+$.50. The fee is subject to change without notice. For organizations that have been granted
a photocopy license by the CCC, a separate system of payment has been arranged.
The consent of CRC Press LLC does not extend to copying for general distribution, for promotion,
for creating new works, or for resale. Specific permission must be obtained in writing from CRC Press
LLC for such copying.
Direct all inquiries to CRC Press LLC, 2000 Corporate Blvd., N.W., Boca Raton, Florida 33431.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and
are used only for identification and explanation, without intent to infringe.
1999 by CRC Press LLC
Lewis Publishers is an imprint of CRC Press
No claim to original U.S. Government works
International Standard Book Number 0-8493-4112-4
Library of Congress Card Number 99-26915
Printed in the United States of America 1 2 3 4 5 6 7 8 9 0
Printed on acid-free paper
THE EDITORS
H. Magdi Selim is Professor of Soil Physics at Louisiana State University, Baton
Rouge. He received his M.S. and Ph.D. in Soil Physics from Iowa State University,
Ames, in 1969 and 1971, respectively, and his B.S. in Soil Science from Alexandria
University in Egypt, in 1964. Professor Selim has published numerous papers and book
chapters, and is a coauthor of one book and several monographs. His research interests
concern the modeling of the mobility of dissolved chemicals and their reactivity in soils
and groundwaters. His research interests also include saturated and unsaturated water
flow in multilayered soils. Professor Selim served as associate editor of Water Ruource.J
Ruearch and the SoiL Science Society of America JournaL He is the recipient of several profes-
sional awards including the Phi Kappa Phi, Gamma Sigma Delta Award for Research,
and the Doyle Chambers Career Achievements Award. Professor Selim is a Fellow of
the American Society of Agronomy and the Soil Science Society of America.
Iskandar K. Iskandar received his Ph.D. degree in soil science and water chemistry
at the University of Wisconsin-Madison, in 1972. He is a Research Physical Scientist at
the Cold Regions Research and Engineering Laboratory (CRREL) and a Distinguished
Research Professor at the University of Massachusetts, Lowell. He developed a major
research program on land treatment of municipal wastewater, and coordinated a num-
ber of research areas including transformation and transport of nitrogen, phosphorus,
and heavy metals. His recent research efforts focused on the fate and transformation of
toxic chemicals, development of nondestructive methods for site assessments, and evalu-
ation of in situ and on-site remediation alternatives. Dr. Iskandar has edited several
books and published numerous technical papers. He organized several national and in-
ternational conferences, workshops, and symposia. He received a number of awards
including the Army Science Conference Award, and CRREL Research and Develop-
ment Award. Dr. Iskandar is a Fellow of the Soil Science Society of America.
PREFACE
During the past decades, phenomenal progress has been made in several areas of
biology, ecology, health, and environmental geochemistry of heavy metals in soils. Prior
to the 1960s, research was focused on enhancing the plant uptake or availability of se-
lected heavy metals or minor elements (also referred to as micro nutrients) from the soil.
More recently, concerns regarding heavy metal contamination in the environment af-
fecting all ecosystem components including aquatic and terrestrial systems have been
identified with increasing efforts on limiting their bioavailability in the vadose zone.
Moreover, several mathematical models for predicting the forms of metals in soils and
the mechanisms of transformations and transport have been developed and evaluated.
Because of the concerns regarding the role of heavy metals in the environment, a series
of international conferences was held to explore the emerging issues of the biogeochem-
istry of heavy metals in the environment. In June 1997, the Fourth International Con-
ference on the Biogeochemistry of Trace Elements was held in Berkeley, California. The
contributions in this book were presented in part in the special symposium focusing on
the fate and transport of heavy metals in the vadose zone as part of this international
conference.
The first four chapters of this book are devoted to sorption-desorption processes of
selected heavy metals in the vadose zone. Kinetics of trace metal sorption-desorption
with soil and soil components is the focus of Chapter 1. Importance of slow reactions and
sorption mechanisms are also emphasized. In Chapter 2, adsorption of nickel by various
soils and their isotherms are discussed. Moreover, a general isotherm approach based on
intrinsic soil properties such as cation exchange characteristics and specific surface area
is developed. Chapter 3 provides an overview of the sorption-desorption, precipitation,
as well as complexation processes for cadmium reactions in soils. A discussion of sorp-
tion nonequilibrium during cadmium transport and reversibility of sorption processes
are highlighted. Chapter 4 provides a comprehensive treatment of single and multiple
retention mechanisms of the linear and nonlinear type which are commonly used to de-
scribe sorption-desorption of heavy several heavy metals in soils. Examples include hys-
teresis, reversibility and ion exchange retention kinetics during transport in soils.
In the next three chapters, complexation and speciation processes and their influence
on heavy metal mobility are discussed in detail. In Chapter 5, factors influencing com-
plex formation of copper are emphasized. The effect of humic and fulvic acids on the
retention of copper by soils and minerals is also presented. In Chapter 6, two sorption
models that describe heavy metal binding of copper with solid and dissolved organic
matter are presented. The applicability of such models to describe copper retention dur-
ing transport is assessed. Also, the bioavailability (accumulation and excretion rates and
toxicity) of copper for earthworms is discussed in Chapter 6. The effect of dissolved
selenium species including metal selenium complexes and other dissolved organic car-
bon on selenium forms and their retention behavior in soils is presented in Chapter 7.
Bioavailability and retention of heavy metals and their mobility in the vadose zone are
presented in Chapters 8 through 11. The bioavailability and mobility of several divalent
heavy metals as affected by pH and redox conditions are the focus of Chapter 8. Meth-
ods for quantif}ring and predicting the influence on mobility are also illustrated. In Chapter
9, the mobility of lead in calcareous mined soils is presented. The effect of various re-
agents on the mobility of lead in the vadose zone under different pH and redox condi-
tions is also evaluated.
In Chapter 10, an overview of modeling of heavy metal retention is given, along with
factors influencing their mobilization/immobilization when organic residue/sewage sludge
amendments are incorporated in the vadose zone. The use of a multiple reaction model-
ing approach is illustrated and the effect on retention parameters when organic waste is
incorporated to the soil was also presented. The significance of the rhizosphere and its
role on trace element interactions in the soil-plant system is the focus of Chapter 11. In
addition, a conceptual model describing the dynamics of trace element processes be-
tween plant roots and the soil in the vadose zone is presented. In Chapter 12, plant-
available concentration levels for selected heavy metals in the vadose zone based on
several extraction methods is discussed. The potential of various extraction methodolo-
gies is also evaluated. In Chapter 13, the authors discuss case studies of metal contami-
nation from emission sources and old abandoned sites. The site investigations, monitoring,
and alternative methods for remediation are given.
We wish to thank the authors for their contributions to this book. Weare most grate-
ful for their valuable time and effort in critiquing the various chapters and in keeping our
focus on the main theme of our topic on heavy metals in the vadose zone. Special thanks
are due to Drs. C. Hinz (University of Goettingen) and Giles Marion (U.S. Army
CRREL) for their help in reviewing Chapters 3 and 11. Without the support of the
Louisiana State University and the U.S. Army CRREL, this project could not have been
achieved. Finally, we wish to express our appreciation to Ann Arbor Press for their help.
H. Magdi Selim
I.K. Iskandar
CONTRIBUTORS
K. Bajracharya
Resource Sciences Center
Department of Natural Resources
Block C, Gate 2, 80 Meiers Road
Indooroopilly
Queensland 4068
Australia
D.A. Barry
Department of Civil and Environmental
Engineering
University of Edinburgh
Edinburgh EH9 3JN
United Kingdom
Klara Bujtas
Research Institute for Soil Science and
Agricultural Chemistry of the
Hungarian Academy of Sciences
(RISSAC)
P.O. Box 35
H-1525 Budapest
Hungary
Philippe Cambier
INRA
Science du Sol
Route de St Cyr
F -78026, Versailles
France
Rayna Charlatchka
INRA
Science du Sol
Route de St Cyr
F -78026, Versailles
France
Stephen Clegg
Department of Ecology and
Environmental Research
Swedish University of Agricultural
Sciences
Box 7072
S-750 07 Uppsala
Sweden
Francois Courchesne
Departement de Geographie
Universite de Montreal
C.P. 6128 Succursale
Centre-Ville
Montreal, Quebec H3C 3J7
Canada
J. Csillag
Research Institute for Soil Science and
Agricultural Chemistry of the
Hungarian Academy of Sciences
(RISSAC)
P.O. Box 35
H-1525 Budapest
Hungary
Carolina Garcia-Rizo
University of Murcia
Department of Agricultural Chemistry,
Geology and Pedology
Faculty of Chemistry
E-30071 Murcia
Spain
George R. Gobran
Department of Ecology and
Environmental Research
Swedish University of Agricultural
Sciences
Box 7072
S-750 07 Uppsala
Sweden
Antonius A.F. Kettrup
Institute of Ecological Chemistry
GSF-National Research Center for
Environment and Health
N euherberg/Munich
Postfach 1129
D-85764 Oberschleissheim
Germany
R.S. Kookana
Cooperative Research Center for Soil
and Land Management
CSIRO Land and Water, PMB No.2
Glen Osmond SA 5064
Australia
A. Lukacs
Research Institute for Soil Science and
Agricultural Chemistry of the
Hungarian Academy of Sciences
(RISSAC)
P.O. Box 35
H-1525 Budapest
Hungary
Luis Madrid
Instituto de Recursos Naturales y
Agrobiologia (CSIC)
Apartado 1052
E-41080 Seville
Spain
Mari P.J.C. Marinussen
Wageningen Agricultural University
Sub-Department Soil Science and Plant
Nutrition
P.O. Box 8005
6700 Wageningen
The Netherlands
Josefa Martinez-Sanchez
University of Murcia
Department of Agricultural Chemistry,
Geology and Pedology
Faculty of Chemistry
E-30071 Murcia
Spain
E.V. Mironenko
Institute of Soil Science and
Photosynthesis
Academy of Sciences of Russia
Push chino, Moscow Region
142292 Russia
R. Naidu
Cooperative Research Center for Soil
and Land Management
CSIRO Land and Water, PMB No.2
Glen Osmond SA 5064
Australia
T. Nemeth
Research Institute for Soil Science and
Agricultural Chemistry of the
Hungarian Academy of Sciences
(RISSAC)
P.O. Box 35
H-1525 Budapest
Hungary
G. Partay
Research Institute for Soil Science and
Agricultural Chemistry of the
Hungarian Academy of Sciences
(RISSAC)
P.O. Box 35
H-1525 Budapest
Hungary
Carmen Perez-Sirvent
University of Murcia
Department of Agricultural Chemistry,
Geology and Pedology
Faculty of Chemistry
E-30071 Murcia
Spain
Alexander A. Ponizovsky
Institute of Soil Science and
Photosynthesis
Academy of Sciences of Russia
Pushchino, Moscow Region
142292 Russia
Katta J. Reddy
Department of Renewable Resources
P.O. Box 3354
University of Wyoming
Laramie, WY S2071
S. Schulte-Hostede
Institute of Ecological Chemistry
GSF-National Research Center for
Environment and Health
N euherberg/Munich
Postfach 1129
D-S5764 Oberschleissheim
Germany
H. Magdi Selim
Sturgis Hall
Agronomy Department
Louisiana State University
Baton Rouge, LA 70S03
Donald L. Sparks
Department of Plant and Soil Sciences
University of Delaware
Newark, DE 19717-1303
Daniel G. Strawn
Department of Plant and Soil Sciences
University of Delaware
Newark, DE 19717-1303
T.A. Studenikina
Institute of Soil Science and
Photosynthesis
Academy of Sciences of Russia
Push chino, Moscow Region
142292 Russia
Erwin J.M. Temminghoff
Wageningen Agricultural University
Sub-Department Soil Science and Plant
Nutrition
P.O. Box S005
6700 Wageningen
The Netherlands
Y.T. Tran
Department of Environmental
Engineering
University of Western Australia
Nedlands W A 6907
Australia
Irena Twardowska
Institute of Environmental Engineering
Polish Academy of Sciences
34 M. Sklodowska-Curie Street
41-S19 Zabrze
Poland
Sjoerd E.A.T.M. Van der Zee
Wageningen Agricultural University
Sub-Department of Soil Science and
Plant Nutrition
P.O. Box S005
6700 Wageningen
The Netherlands
M. Th. van Genuchten
U.S. Salinity Laboratory
USDA, ARS
450 W. Big Springs Road
Riverside, CA 92507
Walter W. Wenzel
University of Agriculture
Institute of Soil Science
Gregor-Mendel Strasse 33
A-lISO Vienna
Austria
Franz Zehetner
University of Agriculture
Institute of Soil Science
Gregor-Mendel Strasse 33
A-lISO Vienna
Austria
CO.'\'TEl'\TS
Chapter 1. Sorption Kinetics of Trace Elements in Soils and Soil Materials ............... 1
DanieL G. Strawn and Dona{J L. SparlcJ
Introduction ....................................................................................................................... 1
Evidence for Slow Sorption and Desorption Reactions ................................................. 3
Diffusion-Controlled Kinetic Reactions ........................................................................... 8
Kinetics and Mechanisms of Adsorption Processes .................................................. 10
Kinetics and Mechanisms of Surface Precipitation .................................................. 18
Summary .......................................................................................................................... 24
References ........................................................................................................................ 25
Chapter 2. Adsorption Isotherms of Nickel in Acid Forest Soils ................................ 29
Franz Zehetner and WaLter W. wenzeL
Introduction ..................................................................................................................... 29
Adsorption ....................................................................................................................... 29
Definition ..................................................................................................................... 29
The Diffuse Double-Layer ......................................................................................... 30
Adsorption Mechanisms ............................................................................................. 30
Adsorption Isotherms ...................................................................................................... 33
Classification ............................................................................................................... 33
The Langmuir Equation ............................................................................................. 34
The van Bemmelen-Freundlich Equation .................................................................. 39
Case Study ....................................................................................................................... 40
Adsorption versus Precipitation ................................................................................. 42
Langmuir and van Bemmelen-Freundlich Isotherms ............................................... 42
Effect of Soil:Solution Ratio on Quantity-Intensity Relationships ......................... 46
Fractionation of Adsorbed Nickel .............................................................................. 50
Adsorption Density and General Adsorption Density Isotherms ............................ 52
Summary .......................................................................................................................... 54
References ........................................................................................................................ 55
Chapter 3. Sorption-Desorption Equilibria and Dynamics of
Cadmium During Transport in Soil ......................................................................... 59
R.S. Kookana, R. NaiJu, D.A. Barry, Y.T. Tran, and K Bajracharya
Introduction ..................................................................................................................... 59
Processes Governing Fate of Cadmium in the Soil Profile ........................................... 60
Sorption ....................................................................................................................... 60
Factors Affecting Cd Sorption in Soils ................................................................... 61
Precipitation ................................................................................................................ 69
Kinetics of Cd Sorption .............................................................................................. 70
Sorption Behavior of Cd During Transport Through Soil Columns ....................... 71
Batch versus Flow-Through Systems ..................................................................... 71
Evidence of Sorption Nonequilibrium During Cd Transport Through Soil ........... 74
Asymmetrical Breakthrough Curves ...................................................................... 74
Flow-Interruption as a Test for Sorption Nonequilibrium .................................... 75
Model Fitting ........................................................................................................... 76
Mass Balance Check for Complete BTCs .............................................................. 77
Causes of Sorption Nonequilibrium During Transport ............................................ 78
Cd Transport Under Field Conditions and Its Modeling ......................................... 78
Desorption and Reversibility of Cd Sorption ............................................................ 80
Desorption of Specifically Sorbed Cd .................................................................... 80
Partial Reversibility of Cd Sorption from Calcite and Calcareous Soils .............. 82
Cd Desorption Kinetics ........................................................................................... 82
Sorption Reversibility in Flow-Through Experiments .......................................... 83
Summary .......................................................................................................................... 83
References ........................................................................................................................ 85
Chapter 4. Modeling the Kinetics of Heavy Metals Reactivity in Soils ...................... 91
H. Magdi SeLim
Introduction ..................................................................................................................... 91
Linear Retention .............................................................................................................. 92
Nonlinear Retention ........................................................................................................ 93
Langmuir or Second-Order Kinetics ............................................................................. 94
Hysteresis ......................................................................................................................... 96
Irreversible Reactions ..................................................................................................... 96
Specific Sorption ............................................................................................................. 96
Multiple Retention .......................................................................................................... 98
Ion Exchange Retention ................................................................................................ 100
Kinetic Ion Exchange ............................................................................................... 102
Case Study ................................................................................................................. 102
References ...................................................................................................................... 105
Chapter 5. Copper Retention as Mfected by Complex Formation with
Tartaric and Fulvic Acids ....................................................................................... 107
Alexander A. PonimIJcflcy, T.A. StUdenilcina, and E. V. Mironenlco
Introduction ................................................................................................................... 107
Copper(II) Retention by Soils, Oxides, and Clays ................................................. 107
Solution Complex Formation and Cu(II) Adsorption ............................................ 109
Complexes of Cu(II) with Fulvic Acids .................................................................. 110
Influence of FA and Humic Acids (HA) on the Retention of
Cu(II) by Solid Phases ......................................................................................... III
Copper Retention by Soil (A Case Study) ................................................................... III
Kinetics of Cu(II) Retention .................................................................................... III
Cu(II) Retention Isotherms and Cation Balance .................................................... 112
Evaluation of Na
2
EDTA Ability to Extract Retained Copper ............................... 115
Effect of Tartrate and Fulvic Acid on Cu(II) Retention Isotherms ....................... 115
Modeling of Cu(II) Retention (Exchange) by Soil.. ............................................... 119
Summary ........................................................................................................................ 121
References ...................................................................................................................... 122
Chapter 6. Copper Mobility and Bioavailability in Relation with
Chemical Speciation in Sandy Soil ........................................................................ 127
E.J.M Temminghoffi MP.J.G. MarinLMden, and S.E.A.T.M Van der Zee
Introduction ................................................................................................................... 127
Sorption Models ............................................................................................................ 128
Parameter Assessment Sorption Models ...................................................................... 129
Copper Speciation in a Copper Contaminated SoiL ................................................... 130
Mobility .......................................................................................................................... 133
DOC Mobility Enhanced Copper Mobility ............................................................ 133
Field Site Accumulation in Soil ................................................................................ 136
Bioavailability ................................................................................................................ 136
Bioavailability for Soil Organisms ........................................................................... 136
Field Site Accumulation by Earthworms ................................................................ 139
Summary ........................................................................................................................ 143
References ...................................................................................................................... 145
Chapter 7. Selenium Speciation in Soil Water:
Experimental and Model Predictions .................................................................... 147
Katta J. Reddy
Introduction ................................................................................................................... 147
Speciation of Dissolved Se ............................................................................................ 148
Experimental and Model Predictions ........................................................................... 149
Dissolved Se Speciation with CuO .......................................................................... 149
Dissolved Se Speciation with GEOCHEM ............................................................ 149
Comparison ............................................................................................................... 150
Future Research ............................................................................................................ 156
References ...................................................................................................................... 156
Chapter 8. Influence of Reducing Conditions on the Mobility of
Divalent Trace Metals in Soils ............................................................................... 159
Philippe Cambier and Rayna Charlatchlca
Introduction ................................................................................................................... 159
Controversial Studies on Soil-Plant Systems .............................................................. 160
Formation of Insoluble Sulfides and Other Solubility Equilibria .............................. 161
Role of Fe and Mn Oxides as Trace Metal Sorbents .................................................. 162
Reducing Processes Change pH ................................................................................... 164
Role of Soluble Organic Ligands ................................................................................. 168
Transformation of Insoluble Organics ......................................................................... 170
Summary ........................................................................................................................ 171
References ...................................................................................................................... 172
Chapter 9. Lead Mobilization in Calcareous Agricultural Soils ................................ 177
Carmen Pirez-Sirvent, JOde/a Martinez-Sanchez, and Carolina Garda-Rizo
Introduction ................................................................................................................... 177
Soil Formation Factors ................................................................................................. 178
Environmental Conditions ........................................................................................ 178
Nature of the Materials ............................................................................................ 178
Transport ........................................................................................................................ 180
Dissolved Load .......................................................................................................... 182
Particulate Forms: Suspended and Bed Loads ....................................................... 186
Geochemical Processes ................................................................................................. 186
Mobilization-Physical Weathering-Hydration Relations ....................................... 187
Soluble Pb-Adsorbent Precursor Ratio ................................................................ 187
Bicarbonated-Acidic Water Interaction ................................................................... 188
Acid Water-Mineralized Particulate Material-0
2
-C0
2
Interaction ................... 189
Acid Water-Carbonated Particulate Material-0
2
-C0
2
Interaction .................... 189
Pb Sorption-Desorption ........................................................................................... 190
Mobility .......................................................................................................................... 191
Provoked Pb Mobility: Speciation Study ................................................................ 191
Mobility in the Vadose Zone .................................................................................... 193
Pb Assimilation by Plants ......................................................................................... 195
Conclusion ..................................................................................................................... 196
References ...................................................................................................................... 197
Chapter 10. Metal Retention and Mobility as Influenced by
Some Organic Residues Added to Soils: A Case Study ....................................... 201
LUM Madrw
Introduction ................................................................................................................... 201
Soil as a Sink for Trace Metals ................................................................................. 201
Modeling Approaches for Retention of Metals by Soils ......................................... 202
Metal Concentrations in the Soil Solution ................................................................... 204
Factors Causing a Reversal of Immobilization ............................................................ 205
Interaction with Natural Organic Matter .................................................................... 207
Effect of Organic Residues on Metal Solubility .......................................................... 209
The Case of Sewage Sludge .......................................................................................... 210
A Mediterranean Concern: Olive Mill Wastewater .................................................... 211
Setting Up the Problem ............................................................................................ 211
Effect of OMW on Metal Retention Properties of Soils ........................................ 212
OMW in the Aqueous Phase as a Mobilizing Agent of
Insoluble Metal Forms ......................................................................................... 215
Summary ........................................................................................................................ 218
References ...................................................................................................................... 219
Chapter 11. The Rhizosphere and Trace Element Acquisition in Soils .................... 225
George R. Gobran, Stephen Clegg, and FrancoM Courchune
Introduction ................................................................................................................... 225
History ....................................................................................................................... 226
I
Rhizosphere - Defmitions ........................................................................................ 226
Methods of Rhizospheric Study ............................................................................... 226
Rhizodeposition ............................................................................................................. 228
Root Distribution and Longevity ............................................................................. 228
Belowground Carbon Flux ....................................................................................... 229
Exudates in the Rhizosphere .................................................................................... 229
Acid-Base Changes in the Rhizosphere ................................................................... 229
Rhizospheric Feedback Loops ...................................................................................... 232
Regulating Processes ................................................................................................ 232
Element Supply and Mobility in the Rhizosphere .................................................. 233
Microbial Activity and Element Accumulation in the Rhizosphere ....................... 234
Case Studies ................................................................................................................... 235
The Conceptual Model ............................................................................................. 236
Field Site and Treatments ......................................................................................... 236
Soil Fractionation ...................................................................................................... 236
Chemical Properties of the Soil Fractions ............................................................... 237
Weathering in Bulk and Rhizosphere Soil .............................................................. 239
Tree Growth and Rhizosphere Chemistry ............................................................... 242
Implications and Future Research ................................................................................ 242
References ...................................................................................................................... 245
Chapter 12. Distribution of Ecologically Significant Fractions of
Selected Heavy Metals in the Soil Profile ............................................................. 251
T. Nemetb, K. Bujtd.1, J. CdUlag, G. Pdrtay, A. Lukdcd, and M. Tb. van Genucbten
Introduction ................................................................................................................... 251
Sludge Application .................................................................................................... 252
Adsorption and Mobility .......................................................................................... 252
Extractions and Bioavailability ................................................................................ 253
Case Study ..................................................................................................................... 254
Nitric Acid Extraction ................................................................................................... 256
.-\AAc-EDTA Extraction .............................................................................................. 260
Concentrations in Soil Solution .................................................................................... 262
Movement ...................................................................................................................... 266
Summary ........................................................................................................................ 260
References ...................................................................................................................... 270
Chapter 13. Heavy Metal Contamination in Industrial Areas and
Old Deserted Sites: Investigation, Monitoring, Evaluation,
and Remedial Concepts .......................................................................................... 273
Irena Twardowdka, S. ScbuLte-Hodtede, and AntoniUd A.E Kettrup
Introduction ................................................................................................................... 273
Impact of Long-Term Stack Emission ......................................................................... 274
Site Characteristics ................................................................................................... 274
Site I: Nowa Huta n/Cracow, Area Adjacent to the
Sendzimir Steelwork Complex, Poland ........................................................... 274
Site II: Irena Glasswork, Inowroclaw, Poland ..................................................... 281
Soil Enrichment with Heavy Metals in the Areas Impacted by a
Long-Term Stack Emission .............................................................................. 287
Screening Survey and Methods ............................................................................ 287
Metal Distribution in Soil vs. the Duration and Extent of Emission ................. 291
Barrier Capacity of a Surface Soil Layer ............................................................. 292
Heavy Metal Binding Strength and Mobility in Soils ......................................... 293
Monitoring Program Requirements for Risk Assessment from Large-Area
Soil Contamination by Trace Metals from Anthropogenic Sources ............... 298
Evaluation of a Large-Area Deserted Industrial Site ................................................. 300
Site Characteristics ................................................................................................... 300
Sources of Heavy Metal Contamination in the Area ........................................... 300
Monitoring Strategy ................................................................................................. 302
Survey of Transfer Pathways and Risk Receptors .............................................. 302
Human Risk Potential Assessment .......................................................................... 304
Approach to Human Risk Potential Assessment ................................................. 304
Applied Model: Quantitative Exposure Assessment (QEA) .............................. 306
Remedial Concepts ........................................... ~ ............................................................ 316
Summary ........................................................................................................................ 319
References ...................................................................................................................... 319
Index .............................................................................................................................. 323
CHAPTER I
Sorption Kinetics of Trace Elements in
Soils and Soil Materials
Daniel G. Strawn and Donald L. Sparks
INTRODUCTION
Environmental contamination resulting from the extensive use of metals and semimetals
in industry, agriculture, and in manufactured products has magnified the threat of toxic-
ity for plants, animals, and society. Since soils and sediments have a large capacity for
sorbing trace elements, an understanding of metal reaction mechanisms with natural
materials is critical. Many studies have appeared in the literature on various aspects of
metal sorption. Results from these studies have been used to develop government regu-
lations, devise cleanup strategies, and develop models that predict the fate of trace ele-
ments in the environment. However, in conducting these studies researchers often
overlook two important aspects: (1) the length of time soils are exposed to a contaminant
(residence time) in the laboratory is relatively short compared with the much longer
residence times that exist in field contaminated soils, and (2) the kinetics of metal sorp-
tion and desorption are often slow. These oversights lead to improper evaluation of con-
taminant behavior in the environment, resulting in regulations that may be improper,
and models and remediation strategies that may be unsuccessful. This chapter will inves-
tigate the effects of residence time (aging) and slow kinetics on sorption and desorption
reaction mechanisms of metals with soils and soil materials (e.g., clay minerals, metal
oxides, and organic matter). Such information is important, and can be used in combina-
tion with transport models to predict the fate of trace metals through the vadose zone,
and can provide information on metal bioavailability and speciation.
Trace elements exist in the soil as either aqueous species, as structural elements in
solids, or sorbed onto the surfaces of soil materials. While many of these trace elements
are present naturally in the environment, their indigenous levels are usually nonthreat-
ening. The buildup of these elements to dangerous levels is a result of commercial use
and disposal practices. The following are a few examples of common sources of contami-
nation: disposal of batteries that contain Pb, Cd, and Hg; exhaust from automobiles that
2 Fate and Transport of Heavy Metals in the Vadose Zone
burn gasoline with Pb additives; application of pesticides that contain Pb and As, e.g.,
Pb
3
(As0
4
)2; the use of Pb in paint; trace elements which are used in manufacturing that
end up in waste disposal and the environment from either discarding the product or as a
by-product of the manufacturing process; desiccation of agricultural runoff water in
ponds which results in Se and As concentrating to dangerous levels; disposal of sewage
which contains several trace elements, in particular heavy metals; and mine drainage
which is often acidic and can increase the mobility of metals.
Scientific studies have clearly shown that exposure to metal contaminants at higher
than natural levels is toxic. As a result, many past uses and disposal practices of metals
are now illegal, and trace element contamination of the environment is now regulated
more closely. However, due to the relatively low solubility of many trace metals, and
often strong sorption to soils, environmental contamination persists, and the threat from
contaminants remains a problem that merits continued scientific investigation.
While toxicity from trace elements, and their presence in the environment at danger-
ous levels are well-established facts, the questions remain: how does one remediate con-
taminated soils effectively, and how can significant risks be accurately evaluated? Finding
effective answers to these questions hinges on a clear understanding of the behavior and
interactions of trace elements with soils. In particular, an understanding of slow desorp-
tion and release kinetics from environmental settings which have been contaminated for
long periods is critical. For example, Smith and Comans (1996) conducted sorption and
desorption experiments on Cs contaminated sediments. They found that failure to in-
clude slow reactions in their model gave much lower estimates of the remobilization
potential of the Cs. They concluded from model fits that sorption half-lives were be-
tween 50 and 125 days, and desorption half-lives were on the order of 10 years.
Many studies rely on an equilibrium approach to predict the retention of contami-
nants on natural materials and subsequent migration through the vadose zone. Researchers
often focus on determining parameters such as distribution coefficients, and the maxi-
mum amount of sorption possible. These studies are often based on the contaminant-
solid interactions over a short period (24 hours or less) because it is assumed that the
reaction has reached completion (Griffin et aI., 1986). However, field soils are seldom, if
ever, at equilibrium, often laboratory studies are also far from equilibrium, and slow
sorption may change the distribution between solid and solution over a period of time
(Smith and Comans, 1996; Sparks, 1998). This is primarily due to slow metal sorption
and desorption kinetics. The failure to account for the slow kinetics results in either
underpredictions of the amount of contaminants retained by soils and minerals, or
overpredictions of contaminant availability in the environment. A better approach is to
base mobility estimates, remediation strategies, and risk assessments on the true avail-
ability of the contaminant, which is often controlled by a rate-limited sorption reaction.
Most soils are heterogeneous media that contain a host of different minerals, solids,
and organic materials. Thus, the interaction of trace elements with soils is a heteroge-
neous process. Several possible sorption mechanisms have been proposed (Figure 1.1):
diffusion into micropores and solids followed by subsequent sorption onto interior sur-
faces; sorption to sites of variable reactivity, including sites which involve different bonding
mechanisms, i.e., inner-sphere vs. outer-sphere and monodentate vs. bidentate; and sur-
face precipitation (Fuller et aI., 1993; Loehr and Webster, 1996; Scheidegger and Sparks,
1996). Due to the heterogeneity of soil, these processes can occur simultaneously. A
Sorption Kinetics of Trace Elements in Soils and Soil Materials 3
Figure 1.1. Schematic of soil particles illustrating the
different types of sorption that are possible. See text
for definitions.
measured sorption or desorption rate often reflects a combination of all of the sorption
mechanisms. However, it is possible that one mechanism may dominate at a particular
time in the sorption reaction and the measured rate is primarily an expression of that
reaction rate. For example, outer-sphere complexation can precede inner-sphere com-
plexation, which can precede surface precipitation. The significance of this continuum
in sorption is that while many sorption and desorption reactions may appear to have
reached equilibrium, in fact the reaction can be continuous, and the slow process will not
be measured if the experimentalist studies a short reaction time. In such cases, important
secondary processes which are slower than the primary process may be completely over-
looked. Thus, predictions on the fate of the contaminant may be inaccurate. This can
cause increased threats of toxic exposure, improper evaluation of risks, and/or misap-
propriation of valuable cleanup and public safety funds.
To protect human health and the environment from overexposure there must exist
effective cleanup strategies, accurate risk assessment technologies, and models that
correctly predict the fate of trace elements. For these tasks to be accomplished, time
dependent reactions of trace elements with soils must be taken into consideration.
Thus, the goals of this chapter are to discuss the kinetics of trace element interactions
with soil and soil components, including the importance of slow reactions and possible
sorption mechanisms.
EVIDENCE FOR SLOW SORPTION AND DESORPTION REACTIONS
There are two separate phenomena associated with slow kinetic sorption processes:
(1) a continuous slow removal of the sorptive from solution (sorption), and (2) a slow
release of the sorbate from the sorbent (desorption). The second of these phenomena,
desorption or release, may be influenced by the length of time in which the contaminant
is in contact with the sorbent; i.e., there may be a decrease in the ability of the sorbate to
be removed from the surface with increasing incubation or residence time. As mentioned
above, several hypotheses for the cause of these two phenomena have been proposed
(they are discussed in detail in later sections).
An early report on the effect of incubation time on desorption reactions of metals from
soils was given by McKenzie (1967). It was observed that manganese nodules present in
Australian soils accumulated a large amount of Co. To account for this selective accumu-
lation, a continuous sorption reaction was hypothesized. To test this, McKenzie (1967)
4 Fate and Transport of Heavy Metals in the Vadose Zone
determined both sorption and desorption kinetics of Co on manganese nodules isolated
from soils. He found that removal of Co from solution slowed considerably after two
days, but the extent of desorbability showed a continuous decrease with increasing ag-
ing periods. Thus, Co that was sorbed would become increasingly resistant to desorption
from the nodule with time, resulting in an accumulation over time.
Sorption processes commonly come to a state of quasi-equilibrium rapidly, and many
researchers terminate their sorption experiments at relatively short times. However, it
has been shown that sorption is a continuous process, and that the sorption mechanism
can change over time, with little additional uptake. For example, Nyffeler et aI. (1984)
found that the distribution coefficients for Be, Mn, Zn, Co, and Fe sorption on particu-
late matter from surface sediments and sediment traps increased over the entire time of
observation, 108 days (Figure 1.2), suggesting that sorption is a slow process. Similarly
Bruemmer et al. (1988) found that Ni, Zn, and Cd uptake by soils was continuous for
times up to 42 days; e.g., Ni removal from solution at pH = 6 was 12% in two hours and
70% in 42 days. Bibak et al. (1995) studied the retention of Co by various goethite poly-
morphs and impure goethite. They found that Co sorption behavior varied between the
different polymorphs and minerals, but in all samples the Co uptake increased with con-
tact time (sorption kinetics measured from two hours to 504 hours). McBride (1982)
found that sorption of Cu on noncrystalline aluminum oxide increased over periods of
weeks, and proposed that different bonding mechanisms were responsible for the slow
sorption process.
McLaren et al. (1983) studied the desorption of Cu from humic acid, ferro-manga-
nese concretions, and montmorillonite. In the desorption procedure the sorptive solution
was replaced by the electrolyte solution (no metal), the suspension was allowed to incu-
bate for four hours, and then, new electrolyte solution was added. The repeated washing
of the soil removed little of the Cu, demonstrating that Cu sorption was strong. Young et
al. (1987) compared Cu sorption and desorption reactions on river sediments with Cr
and Zn. They observed that sorption of Zn was complete in four hours, Cr sorption was
far from complete after 48 hours, and Cu sorption kinetics were intermediate. In addi-
tion, Young et al. (1987) concluded that desorption was not irreversible as McLaren et
al. (1983) found, but that the observed irreversibility was a result of the slow kinetics
involved. This slow desorption phenomenon was also observed for phosphate by Lookman
et al. (1995). They found that slow phosphate desorption from soils continued for up to
1,600 hours, and showed no signs of reaching a plateau. In fact, using a rate constant
derived from a first-order fit of the slow reaction, they predicted that 500 days would be
required for desorption of 90% of the phosphate.
Several researchers have noted that not only are trace elements strongly sorbed and
exhibit slow desorption kinetics, but that the rate of desorption decreases with increas-
ing residence times. Padmanabham (1983) conducted desorption experiments ofCu from
goethite and concluded that Cu was sorbed in two different ways: a fraction was associ-
ated with low bonding energy and the rest was associated with high bonding energy. It
was observed that a gradual interchange with increasing incubation time occurs be-
tween the readily desorbed fraction (low energy) and the less readily desorbed fraction
(high energy). Similar results were found by Kuo and Mikkelsen (1980), Schultz et al.
(1987), and Backes et al. (1995), who showed that the desorption rate of several transi-
tion metals (Zn, Co, and Cd) from soils and soil components decreased with increasing
--- --------
10000000
1000000
~
100000
0
'ai
10000
a:
c
0
.';=
1000
::s
.c
c
"Iii
is
100
10
o
Sorption Kinetics of Trace Elements in Soils and Soil Materials 5
- - - ~ - " ' - - - - - -
_____ ---------------------------.:e
20 40 60 80 100
Incubation Time (Days)
o Fe
:eMn
oBe
aCo
.Zn
120
Figure 1.2. Effect of incubation time on the distribution coefficient (1<cJ) for the sorption of selected
trace elements (radioactive tracers) on sediments. From Nyffeler et al. (1984), with permission.
aging time. Their findings further support the hypothesis that a slow process occurs
between trace metals and soils that affects the availability of the metal. Smith and Comans
(1996) observed an increase in the slowly desorbed fraction of sorbed Cs from sediments
with increasing incubation time. Modeling of their data using a two-compartment model
(Figure 1.3) suggests that there exists an exchangeable sorbed fraction and a "fixed"
fraction. Slow transfer between the two fractions was responsible for slow kinetics of
sorption and desorption.
Another method for determining the reversibility of sorption, or the effects of aging, is
to measure the exchangeability of the sorbate using isotopic exchange. This approach
was used by McLaren et al. (1986) who studied the sorption and desorption behavior of
Co from soil components. For humic acid, a large proportion of the Co was isotopically
exchangeable, even for sorption incubation times as long as 50 days. However, for a soil
oxide (ferro-manganese concretions) and montmorillonite, the fraction of nonisotopically
exchangeable Co increased continuously as sorption time increased. Comans (1987)
determined that the isotopic exchangeability of Cd on illite was 100%, but required an
equilibration time of seven to eight weeks. These results suggest that with increasing
incubation time the association between the sorbate and the sorbent changes to that of a
more stable complex, i.e., less readily desorbed.
Hysteresis, or nonsingularity, is a phenomenon in which the sorption and desorption
isotherms do not coincide because of a shift in the equilibrium point. Pseudo-hysteresis
is often observed in systems that have a slower desorption reaction than the sorption
reaction (or vice versa), and is a result of not carrying the experiment out for a long
enough period, i.e., the system has not reached equilibrium (McBride, 1994, p. 91). For
example, Comans (1987) found that sorption and desorption isotherms of Co on illite are
(i Fate and Transport of Heavy Metals in the Vadose Zone
_kt
I Aqueous H Exchangeable I
.J Fixed I
I<r
Equilibrium Process Rate-Limited Process
Figure 1.3. Schematic illustrating a modeling approach used to describe the slow transition of (s
from a mobile (exchangeable phase) to a fixed phase. From Smith and (omans (1996), with
permission.
singular (nonhysteretic) only if the desorption reaction is allowed to come to equilib-
rium, which took 54 days.
Many researchers have found that the magnitude of hysteresis increases with longer
sorption incubation periods. Ainsworth et al. (1994) found that despite increasing the
desorption times from 16 hours to nine weeks, hysteresis persisted for Co and Cd sorbed
on hydrous ferric oxide (HFO). They also found that Cd and Co displayed increasing
hysteretic behavior upon aging from two weeks to 16 weeks (Figure lA, Cd data not
shown), while Pb sorption/desorption behavior was reversible (Figure 1.5).
Oftentimes it is observed that the amount of sorbate able to be desorbed decreases
with increasing incubation time. For example, the results of McKenzie (1980) suggest
that Pb sorption on Fe-oxides is a slow process, and can affect the total amount of Pb
able to be desorbed. McKenzie (1980) incubated Pb on hematite and goethite for peri-
ods from one day to 28 weeks, and then extracted the samples with 2.5% acetic acid until
no additional Pb was removed. A comparison of the Pb removal from the different samples
showed that increasing the incubation time from one day to 28 weeks increased Pb re-
tention by 50% on hematite, and 100% on goethite. Similar experiments were conducted
by Bibak et al. (1995) for Co sorbed on iron oxides; they used a strong acid extractant to
measure the amount of Co released from iron oxides incubated for different periods.
Their results indicated that the percentage of Co released decreased with increasing
aging times. The results of McKenzie (1980) and Bibak et al. (1995) suggest that many
adsorption reactions that appear to be at equilibrium are undergoing slow transforma-
tions that decrease the amount of sorbate able to be desorbed.
Hysteresis has also been observed in cation exchange reactions, where the exchange
of one sorbed cation with another is not completely reversible, i.e., the forward and
reverse exchange reactions do not result in the same isotherms. The hysteretic behavior
of cation exchange is abundantly reported in the literature; an excellent critical review of
this literature was published by Verburg and Baveye (1994). From a survey of the litera-
ture they were able to categorize several elements into three categories (Table 1.1). The
elements in each category were found to show hysteretic exchange between groups, but
not within groups. Verburg and Baveye (1994) proposed that exchange reactions are
most likely a multistage kinetic process in which the later rate-limiting processes are a
result of physical transformation in the system; e.g., surface heterogeneity, swelling hys-
teresis, and formation of quasi-crystals, rather than simply a slow kinetic exchange pro-
cess where there exists a unique thermodynamic relationship for forward and reverse
reactions. While this may be true in some circumstances, an apparent (pseudo) hyster-
esis also can result from slow sorption and desorption reactions, i.e., lack of equilibrium.
100
"0
(])
.0
....
0
III
0
()
-
c
40 (])

(])
Il..
0
2
Sorption Kinetics of Trace Elements in Soils and Soil Materials 7


1:1
c
a
odl
e
e e


...

h
4 6
pH
Adsorption pH edge
aging time = 0 wk
Desorption
aging time = 2 wk
o aging time = 9 wk
C aging time = 16 wk
8 10 12
Figure 1.4. Effect of sorption incubation time on the desorption of Co from Fe oxide. From
Ainsworth et al. (1994), with permission.
1
"0
(])
.0
....
o
III 60
.0
a..
-
c::
(])

(])
a..
Adsorption pH edge
aging time = 0 wk
Desorption
o aging time = 2 wk
aging time = 12 wk
C aging time = 21 wk
pH
12
Figure 1.5. Effect of sorption incubation time on the desorption of Pb from Fe oxide. From
Ainsworth et al. (1994), with permission.
Regardless of the different reasons for hysteresis, we agree with the proposal of Verburg
and Baveye (1994) that to be of practical value, kinetic models need to be complemented
by detailed information on the mechanism (s) responsible for the slow kinetic reaction (s).
The above discussion has presented many examples of slow kinetics for both sorption
and desorption processes. Several of the researchers have speculated on the mechanism(s)
that control slow reactions. However, these hypotheses are based primarily on macro-
scopic data, while sorption and desorption processes are microscopic phenomena. At
8 Fate and Transport of Heavy Metals in the Vadose Zone
Table 1.1. Classification of Cations into Three Groups
in Such a Way that Hysteresis Has Been Documented
in the Literature for Binary Reactions Involving Cat-
ions from Different Groupsa
Group 1 Group 2 Group 3
Na+ K+ Ca
2
+
Li+ Rb+ Ba
2
+
Cs+ sr2+
NH/
Mg2+
Mn
2
+
Cu
2
+
Ni
2
+
a From Verburg and Baveye (1994), with permission.
best, macroscopic investigations suggest a particular mechanism may be occurring; they
provide little evidence that other mechanisms are not involved (Sposito, 1986; Chisholm-
Brause et aI., 1990). Despite this problem, reasonable predictions of sorption mecha-
nisms based on macroscopic observations are often made. However, the uncertainty
becomes clear when one observes the discrepancies in predictions of mechanisms be-
tween published reports on similar systems. For example, Ainsworth et al. (1994) pre-
dicted that the similarities of the ionic radii between Co and Fe suggested a coprecipitation
mechanism as responsible for aging. This hypothesis was supported by their observation
that as ionic radius decreased, i.e., Pb>Cd>Co, hysteresis increased. However, Bibak et
al. (1995) predicted that the mechanism responsible for the slow reaction of Co on vari-
ous iron oxides was diffusion. This prediction was based on a good fit of the data to a
diffusion model. An important point to note about comparing these two systems is that in
the experiments of Ainsworth et al. (1994) the initial Fe-oxide was amorphous and un-
derwent recrystallization, while the Fe-oxides used in the experiments of Bibak et al.
(1995) were crystalline and did not undergo a solid phase transformation. Such differ-
ences can have important consequences on sorption mechanisms. Despite this discrep-
ancy, one can conclude from these studies that in order to better predict the mechanisms
responsible for the slow kinetic processes, microscopic as well as macroscopic data are
necessary.
DIFFUSION-CONTROLLED KINETIC REACTIONS
Diffusion is an activated process driven by the necessity of a system to be at its lowest
possible energy, i.e., uniformly distributed throughout space. Since soils are porous ma-
terials containing both macropores (>2 nm) and micropores 2 nm) (Pignatello and
Xing, 1996), diffusion is a mechanism that can control the rate of sorption of trace ele-
ments on soils. These pores can be interparticle (between aggregates) or intraparticle
(within an individual particle). Intraparticle pores can form during weathering, upon
solid formation, or may be partially collapsed interlayer space between mineral sheets;
i.e., vermiculite and montmorillonite. The rate of diffusion through a pore is dependent
on pore size, particle size, tortuosity, chemical interactions, chemical flux through the
soil, and whether the pore is continuous or discontinuous. Besides pore diffusion, solid-
Sorption Kinetics of Trace Elements in Soils and Soil Materials 9
phase diffusion is also a transport-limited process. Solid phase diffusion is dependent on
the characteristics and interactions of the diffusant and the solid (Pignatello and Xing,
1996). Since there exists a range of diffusion rates in the soil, it follows that with increas-
ing exposure time the fraction of contaminants in the more remote areas of soil particles
(accessible via slow diffusion) will increase. This slow sorption phenomenon is often the
explanation researchers use to account for the slow continuous sorption and desorption
observed between metals and soil (Sparks, 1989; Burgos et aI., 1996).
Bruemmer et aI. (1988) measured sorption and desorption of Cd, Zn, and Ni with
goethite, a porous iron oxide known to have defects within the structure in which metals
can be incorporated to satisfy charge imbalances. They found that the kinetics were
described well with a solution to Fick's second law (a linear relation with the square root
of time), and proposed that the uptake of the metal followed a three-step mechanism: "(i)
adsorption of metals on external surfaces, (ii) solid-state diffusion of metals from exter-
nal to internal sites,O and (iii) metal binding and fIxation at positions inside the goethite
particle," suggesting that the second mechanism is responsible for the slow reaction
(Bruemmer et aI., 1988). Similar observations on sorption of divalent metal ions were
made by Coughlin and Stone (1995). They suggested that the slow sorption and desorp-
tion could be a result of slow diffusion that occurred because their synthetic goethite
may have had an unusually high level of pores and cavities. Axe and Anderson (1997)
also found that sorption of Cd and Sr could be characterized by a model which included
two steps: a rapid reversible sorption step followed by a slow, rate-limiting process in-
volving the diffusion of the cations through small pores existing along the surface.
While the above examples have hypothesized that diffusion is the rate-limiting step
based on good model fIts to data and some speculation, macroscopic sorption experi-
ments are not defInitive proof of a mechanism (Sposito, 1989, p. 150). To give additional
support to diffusion as a mechanism for sorption onto porous media, Papelis (1995)
measured surface coverages of Cd and selenite on porous aluminum oxides using X-ray
photoelectron spectroscopy (XPS). Papelis (1995) calculated the expected thickness of
sorbed Cd and selenite from the total metal loss from solution using both external and
internal surface areas. A good agreement was found between the calculated and the
measured (using XPS) surface coverage thickness when the total surface area (i.e., in-
ternal and external surface area) was used. When the surface layer thickness was calcu-
lated without considering internal surface area, then the calculated thickness exceeds
the thickness observed using XPS. Therefore, the most likely sorption mechanisms were
sorption to external sites, diffusion of Cd into the internal structure, and subsequent
sorption. While Papelis (1995) didn't measure the kinetics of the reaction, it seems prob-
able that the sorption to the interior sites is slower than the exterior sites, and thus a slow
kinetic sorption step would exist.
Fuller et aI. (1993) combined kinetic sorption and desorption experiments with spec-
troscopic observations (Waychunas et aI., 1993) to conclude that the rate-limiting pro-
cess in arsenate sorption by ferrihydrite is diffusion into the solid structure. Using X-ray
Classical solid-state diffusion is a very slow process in crystalline structures, and usually only
significant at very high temperatures (McBride, 1994, p. 28). In this case, solid state diffusion
should be interpreted as diffusion processes through faults and micropores.
10 Fate and Transport of Heavy Metals in the Vadose Zone
0.12
0.1
CI>
1!i
'0.
'0 0.08
e
Il.
o
0.06
...
as
(5
::!!
CI> 0.04
u.

0.02
60
__
.--------_ ....... _ .. _ ... _ .... _ .......... _ ............................ _ ............. _ ....................... - ... .
40 "C
CI>
e
o
III
30

:.!!
20 0

./
/ 10
/
0+----------.---------.----------.----------+0
o 100 150
Time (Hours)
Figure 1.6. Pore-space diffusion fit of As(V) adsorption density as a function of time for total (dark
line), diffusion-limited (dotted line) and exterior surface components of adsorption (thin solid line).
The solid triangles represent the adsorption data. Exterior sites are modeled based on equilibrium.
From Fuller et al. (1993), with permission.
absorption fine structure (XAFS) spectroscopy, Waychunas et al. (1993) found that
arsenate is sorbed predominantly as inner-sphere bidentate complexes, regardless of
whether the arsenate was adsorbed post-mineralization of the ferrihydrite, or present
during precipitation. Thus, at the pH of their study (8.00), arsenate surface precipitates
were not formed. Slow sorption and desorption were explained as slow diffusion of the
arsenate to or from interior surface complexation sites that exist within disordered ag-
gregates of crystallites. The arsenate sorption and desorption kinetics (Figure 1.6) were
explained well using a model which included two types of sorption sites: those easily
accessible were described assuming equilibrium (thin solid line), while the sites which
had limited accessibility (dotted line) were well represented by an equation which is
based on Fick's second law of diffusion.
Kinetics and Mechanisms of Adsorption Processes
Adsorption is a phenomenon in which matter accumulates at the interface between a
solid phase and a solution phase; it is largely considered to be two-dimensional (Sposito,
1989, p. 132). Adsorption reactions are governed by the laws of thermodynamics: en-
ergy is conserved, and the entropy of a system increases to a maximum. These two con-
cepts can be combined to create the Gibbs free energy (G) function. For a reaction to
occur, the products must have a lower free energy than the reactants < 0). This can
occur by either a decrease in enthalpy, an increase in entropy, or both. It is important to
note that a change in enthalpy can dominate the free energy function creating a negative
even when the entropy is decreased in the reaction, and vice versa. Therefore, an
Sorption Kinetics of Trace Elements in Soils and Soil Materials 11
adsorption process leads to an association between an ion and a surface, driven by the
desire of the system to achieve an overall lower free energy.
While thermodynamics can be used to determine if a reaction is favorable, it does not
indicate the rate of the reaction, nor the pathways involved in arriving at the state with
the lowest free energy. This information can be gained by measuring reaction kinetics.
In real systems, such as soils and sediments where there exist several different types of
sorption sites, reaction mechanisms and kinetics can be heterogeneous. In these systems
kinetics plays an important role in the fate of trace elements since such systems are not at
equilibrium, but are continuously undergoing chemical changes as they seek to produce
the most stable species (Steinfield et al., 1989, p. 1). The change may be slow, resulting
in the sorbate becoming less available with time (aging) (Koskinen and Harper, 1990),
and can result in a change from one type of sorbed complex to another. This process is
similar to the concept of the Ostwald-step rule: the first product in a precipitation reac-
tion is that which has the highest solubility, followed by a slow continuous transforma-
tion to a more stable species (Stumm and Morgan, 1996, p. 807). An analogous process
in adsorption would result in a multitude of adsorbed complexes, some of which may be
in a metastable equilibrium state, undergoing continuous transformation to the most
stable species.
Evidence for this slow, continuous change to a more stable species is commonly ob-
served for solid materials. Upon initial precipitation the solid is in an active form that has
a disordered lattice (amorphous), and exists in a metastable equilibrium with the solu-
tion (Stumm and Morgan, 1996, p. 356). With time the solid slowly converts to the more
stable inactive form. The inactive form is more crystalline-like, and has a lower solubility.
This slow kinetic phenomenon may continue for geological time spans. An example is
aragonite (a polymorph of calcite), which is found in rocks < 300 million years old. Ara-
gonite is not thermodynamically stable, but forms under surficial temperatures and pres-
sures, and slowly reverts to the more stable calcite (Blackburn and Dennen, 1994, p.
102). Waychunas et al. (1993), using XAFS data fitting, found that aging and continued
polymerization of ferrihydrite resulted in a transformation of the number of linkages and
interatomic distances to those suggesting a progression to the more ordered polymorph
goethite. The slow transformation of a solid to a state with a lower free energy is often
observed as an aging mechanism for precipitates, but transformations between sorption
mechanisms is more difficult to distinguish, and little direct evidence exists for such
processes. However, it seems reasonable to suggest that the energetics of sorption and
desorption reaction processes are analogous to those of precipitation; i.e., kinetically
limited by a transformation to the most stable sorption configuration (lowest ~ G ) .
Adsorption reactions occur via three different mechanisms: inner-sphere complexes,
outer-sphere complexes, and diffuse ion (Figure 1.7, diffuse ion not indicated) (Sposito,
1989, p. 132). Outer-sphere bonds consist of a solvated ion that forms a complex with a
charged functional group; the primary bonding force is electrostatic. An inner-sphere
complex is partially dehydrated; the ion forms a direct ionic or covalent bond with the
surface functional groups. A diffuse ion exists in the water layers near the surface, and is
held by electrostatic attraction from permanent charges that exist in the solid structure.
A major difference between the outer-sphere complex and the diffuse ion complex is in
the strength of the electrostatic force, which is directly correlated to the proximity of the
ion to the surface (McBride, 1994, p. 73). The type of sorption and bonding mechanism
12 Fate and Transport of Heavy Metals in the Vadose Zone
Metal Oxygen a
j H+
aerD
'H 0
"d
Doa
Other Examples
Outer-Sphere
Surface
Complexes
Monodentate
Inner-Sphere
Surface
Complexes
Bidentate
Figure 1.7. Schematic showing the different types of adsorption complexes that can occur on solid
surfaces. See text for definitions. From Hayes (1987), with permission.
depends on several factors: (1) ionic radius, (2) electronegativity, (3) valence charge, (4)
surface type, and (5) ionic strength of the sorptive solution. There are two major types of
surface sites: variable charged sites, e.g., silanol and alumino!; and permanent charge
sites that result from isomorphic substitution.
To model surface complexation and understand the controlling mechanisms, scien-
tists often assign a hypothetical bonding mechanism between an ion and a given surface.
However, ions can bond to surfaces via several different mechanisms, and can undergo
a continuous transition between adsorption mechanisms (Stumm and Morgan, 1995, p.
586). Waychunas et al. (1993) found that arsenate adsorbed onto ferrihydrate by both
monodentate (30%) and bidentate bonding mechanisms. Bargar et al. (1996) used X-ray
Sorption Kinetics of Trace Elements in Soils and Soil Materials 13
absorption spectroscopy (XAS) to distinguish between outer- and inner-sphere sorbed
Pb on CX-AI
2
0
3
. They found that on the planar 0001 surface Pb-O-AI distances were
consistent with an outer-sphere bond, while on the 1102 plane Pb was sorbed as an
inner-sphere complex. Benjamin and Leckie (1981) conducted sorption experiments at
several different loading levels and equilibrium pHs for Cd, Cu, Zn and Pb on amor-
phous iron oxyhydroxide. Their data suggested that there exist several types of bonding
sites with variable bonding strengths, and that measured equilibrium constants are aver-
age values from these different types of sites. McBride (1982) found similar results on
pure noncrystalline aluminum oxide using electron spin resonance (ESR) spectroscopy
to study the change in Cu sorption mechanisms with time. He found that sorption in-
volved sites of varying reactivity. The first reaction step was the rapid sorption of a low
level of Cu; the second reaction occurred over several weeks and resulted in the uptake
of a greater amount of Cu and ESR spectra distinct from the first reaction step. Such
heterogeneity is enhanced in natural systems that contain materials with a variety of
organic and inorganic surface sites.
Adsorption reactions are often considered to form the most stable bond immediately,
but commonly there are intermediates which can be metastable for long times. In fact,
adsorption may consist of a series of chemical and physical reactions that may limit the
overall reaction rate; i.e., ion and surface dehydration, breaking of a strong bond, bond
formation, and surface diffusion (Stumm and Morgan, 1996, p. 761; McBride, 1994, p.
135). Hayes and Leckie (1986) and GrossI et al. (1994) used pressure-jump relaxation
to measure the kinetics of Pb sorption on aluminum oxide and Cu(II) sorption on goet-
hite, respectively. They found that the best fit to the data was obtained by fitting a ki-
netic model that included a transformation from outer-sphere to inner-sphere
complexation. Their results also suggested that sorption behavior was biphasic, which
they explained by suggesting that the slower reaction was a result of sites with lower
affinities. This concept is similar to the high and low affinity site model proposed by
Dzombak and Morel (1990, p. 92). While the kinetics of these reactions are quite rapid
(reactions considered on a millisecond time scale), the demonstration of a multiple step
adsorption mechanism rationalizes the hypothesis that in some systems one step may be
slow enough to be responsible for the slow adsorption and desorption reactions often
observed in soils (Sposito, 1989, p. 150).
The kinetics of Pb sorption on y-AI203 are shown in Figure 1.8. These data show a
fast initial reaction followed by a slow sorption reaction continuing for several hours.
Such biphasic behavior is likely a result of sorption to sites of variable reactivity and/or
diffusion limited sorption. Slow surface precipitation reactions can be ruled out because
analysis of the radial structure function obtained using XAFS (Figure 1.9) does not
exhibit any major features (e.g., second peaks indicative of second shell neighbors) be-
yond the primary Pb-O structural peak at -1.9 A (uncorrected for phase shifts) with
long incubation times.
Biphasic sorption reactions have also been observed in soils. An example is the result
of Lehman and Harter (1984) who measured the kinetics of chelate-promoted Cu re-
lease from a soil to assess the strength of the bond formed. Their sorption/desorption
data were biphasic, which they attributed to high and low energy bonding sites. They
also found that with increased residence time, 30 minutes to 24 hours, there was a tran-
sition of the Cu from low energy sites to high energy sites (as evaluated by release kinet-
14 Fate and Transport of Heavy Metals in the Vadose Zone
60




--.

0
-
c: 50
0
. .;::;
::J
(5
40
C/)
E
30 0
....
--
"0
Q)
20
>
0
E
Q)
10
c:
.n
a..
0
0 2 4 6 8 188 190 192
Time (hours)
Figure 1.S. Kinetics of Pb removal from solution by y-AI
2
0
3
Ionic strength = 0.1, pH = 6.50, initial
Pb concentration = 0.002 M.
1.0
Q)
"C 0.5
:::I
:: 0
c:
g>

E
....
.E
en
c:

r-
o 2
,-__ __ __________
8 Days

48 Hours
24 Hours
3 4 5
R (A)
6 7 8 9
Figure 1.9. Radial distribution function (uncorrected for phase shifts) for Pb sorbed on y-A1
2
0
3
incubated for 24 hours to 70 days. Incubation conditions are the same as in Figure 1.8.
ics). Incubations for up to four days showed a continued uptake of Cu and a decrease in
the fraction released within the first three minutes, which was referred to as the low
energy adsorbed fraction. The results of Smith and Comans (1996), already mentioned,
also showed that Cs sorption onto sediments is biphasic. They modeled exchange reac-
tions assuming exchangeable and fixed fractions. The fixed fraction was assigned to Cs
that was incorporated in the mineral lattice, i.e., predominantly specific exchange sites
on illitic clay. The Cs adsorption mechanisms proposed by Smith and Comans (1996)
Sorption Kinetics of Trace Elements in Soils and Soil Materials 15
'were based on kinetic experiments, i.e., macroscopic observations. Kim et al. (1996)
used nuclear magnetic resonance (NMR) spectroscopy to make microscopic observa-
tions of Cs sorption mechanisms on kaolinite, boehmite, silica gel, and illite. Their ex-
periments coincide with those of Smith and Comans (1996), suggesting that Cs formed
two distinct types of complexes on the surfaces of the minerals: inner-sphere and outer-
sphere.
The energy and stability of adsorbed species varies depending on the type of surface
complex formed. It is generally accepted that surface complexes with more than one
bond are more stable than complexes with a single bond (Stumm and Morgan, 1996, p.
276; McBride, 1994, p. 134), and likewise for inner-sphere vs. outer-sphere sorption
(McBride, in Bolt, 1991, p. 168). One explanation for the increased stability of a
multidentate bond over a monodentate bond may be the increased entropy gained from
a more stable configuration (steric effect) (Steinfield et al., 1989; McBride, 1994, p. 80) .
. w analogous phenomenon is the Chelate Effect; for example, the for the ethylenedi-
amine complex, a chelate ring with bidentate bonding to a cation, is lower than of the
diamine complex, which forms monodentate complexes with cations (Stumm and Mor-
gan, 1996, p. 279, from Schwarzenbach, 1961). The lower for the ethylenediamine
complex means it is more stable. Since the enthalpies for the complexation of cations by
the two chelates are similar, the lower is a result of an increased entropy for the
bidentate ring complex; as mentioned above, this phenomenon is often referred to as a
steric effect or configurational entropy (Stein field et al., 1989, p. 250; McBride, 1994, p.
80). Since the reactive sites on minerals (silanol and aluminol sites) and organic matter
(carboxyls and phenolic-OH) are often considered to be analogous to ligand functional
groups, the steric effect is likely to be an important consideration when determining
mechanisms of trace element adsorption reactions in soil. Thus, it is reasonable to con-
clude that if the coordination environment is appropriate, multidentatebonding will be
favored (thermodynamically) over monodentate bonding. However, the formation of
multiple bonds may have intermediate products that have a higher activation energy
than a complex with only a single bond. As discussed below, an increase in the activation
energy may limit the kinetics of complex formation.
The formation of a surface complex, or conversion of an adsorbate from one bond
type to another, may be thermodynamically favored but inhibited by an activation en-
ergy, which is the extra energy, beyond the difference in the free energy between the
products and reactants required to complete the reactions (Figure 1.10). The
activation energy results from the energy required to form intermediate products not
accounted for in the reaction stoichiometry (Noggle, 1989, p. 532). A large activation
energy will result in slower adsorption and desorption kinetics compared to sorption
processes which have a lower activation energy.
Since the strength of adsorption varies depending on the surface and adsorptive being
considered, the adsorbate availability (via desorption) and kinetics are variable (Pignatello
and Xing, 1996). For many adsorbed ions it is found that the rate of adsorption is faster
than desorption (McBride, 1994, p. 134; Swift and McLaren, in Bolt, 1991, p. 285). A
possible reason for the slower rate of desorption is an increase in the activation energy
required to break the adsorption bonds. The activation energy for desorption can be
quantified as follows: = adsorption + adsorption' where desorption = activa-
tion energy for desorption, = activation energy for adsorption and
16 Fate and Transport of Heavy Metals in the Vadose Zone

Species
Activated
Complex*
---
--------------- ---------------- ------------------------------
Sorbed
Complex
........ E---------Desorption
Sorption --------...
Reaction Coordinates
Figure 1.10. Schematic diagram of G vs. reaction coordinate for sorption and desorption pro-
cesses. Adapted from Sparks and Jardine (1981), with permission.
= energy of adsorption, see Figure 1.10 (McBride, in Bolt, 1991, p. 168).
This equation indicates that desorption of chemisorbed ions yields a larger activation
energy than adsorption reactions, causing desorption to be a slower process. This may
be the cause of the pseudo-hysteresis that is commonly observed in sorption and desorp-
tion experiments; i.e., the forward and reverse isotherms do not overlie when given the
same reaction time.
The experiments of McLaren et al. (1986) were discussed briefly in an earlier section;
however, another look at their results is merited at this point to evaluate possible mecha-
nisms. They found that Co sorbed by a soil oxide demonstrated a continuous decrease in
isotopic exchangeability as sorption times increased (only 20% was exchangeable when
sorption was carried out for 50 days) (Figure 1.11). For humic acid, the isotopic ex-
changeability of sorbed Co decreased only slightly with increased sorption incubation
time (Figure 1.12) (the amount of Co that was isotopically exchangeable remained as
high as 80% for 50 days of sorption incubation time). It is difficult to prescribe a particu-
lar mechanism as the cause for the aging observed in McLaren's studies; however, it is
possible that a more stable complex is being formed on the oxide with increasing sorp-
tion incubation time, increasing the energy required for isotopic exchange. Eliminating
diffusion as a slow exchange mechanism seems reasonable in this case since the humic
acid fraction, a porous material, lacked a slow exchange portion. However, more de-
tailed studies and measurements of the porosity of the two materials is needed for diffu-
sion to be completely ruled out. Surface precipitation is difficult to eliminate; the authors
Sorption Kinetics of Trace Elements in Soils and Soil Materials 17
- - ---- ---- - --- -------
900
800


700
";"0) 600
~
:; 500
(I)
" 400
III
-g
0 300
(,)
200
100
a
a 10 20 30 40 50
Time (days)
Figure 1.11. Isotopic exchangeability of Co sorbed by soil oxide: total Co sorbed (+), and isotopic
exchangeable (.). The space in between the two lines indicates the nonisotopic exchangeable
fraction. From McLaren et al. (1986), with permission.
20
18
-16
~
";"0) 14
0)
.a. 12
"C
(I)
10
~
0
If)
8
a1
0
6
u
4
2
0
0 10 20

<r--____ ~ ~ _______
30
Time (days)
40 50 60
Figure 1.12. Isotopic exchangeability of Co sorbed by soil organic matter: total Co sorbed (.), and
isotopic exchangeable (+). The space in between the two lines indicates the nonisotopic exchange-
able fraction. From McLaren et al. (1986), with permission.
discounted it as the predominant sorption mechanism since surface coverages were low.
For more conclusive evidence, microscopic measurements are necessary.
18 Fate and Transport of Heavy Metals in the Vadose Zone
In this section we have discussed adsorption and desorption kinetics and sorbate sta-
bility. The kinetics of sorption and stability of a surface complex is a factor of both
entropy (steric factors) and enthalpy (bond energetics). However, the formation of the
most stable adsorbed species may be limited by intermediate complexes. Thus, if a sor-
bate slowly converts from one sorption or bonding species to a more stable complex that
has a lower free energy, the result is important toward controlling the rate of uptake and
affecting the availability of trace metals.
Kinetics and Mechanisms of Surface Precipitation
In contrast to adsorption, surface precipitation is a 3-dimensional growth phenom-
enon that occurs on surfaces. Classical solution chemistry defines aqueous systems in
three states: undersaturated, saturated, and supersaturated, with respect to the solubil-
ity of inorganic precipitates. A system saturated or supersaturated has a negative ~ G ,
indicating that the precipitation of a solid product is favored. Precipitation that occurs in
a saturated system proceeds more slowly than a supersaturated system (Stumm and
Morgan, 1996, p. 802). Surface precipitation during trace metal sorption has been ob-
served in systems undersaturated with respect to the pure hydroxide, and below mono-
layer surface coverage (Fendorf et aI., 1992; Fendorf and Sparks, 1994; O'Day et al.,
1994a,b; Scheidegger et aI., 1996). This means that the availability and transport of a
cation or anion may be controlled by precipitation and dissolution mechanisms, as op-
posed to adsorption phenomena. Veith and Sposito (1977) showed that traditional sorp-
tion data are described equally well by both surface precipitate models and adsorption
isotherms. In addition, it has been noted that solubility lines of many soil solutions (loga-
rithm of metal activity plotted as a function of pH) reveal undersaturation with respect
to common precipitates; however, they often have slopes paralleling those of pure pre-
cipitates (McBride, in Bolt et aI., 1991, p. 171, from Lindsay, 1979). Such examples
display the ambiguity of macroscopic models in describing microscopic processes; i.e.,
surface precipitation and adsorption models seem to describe sorption data equally well.
Since precipitation and dissolution reactions exhibit slower kinetics than adsorption and
desorption (Farley et aI., 1985) they may be the mechanism responsible for aging and
the slow kinetics of sorption and desorption often observed in experimental systems
(Fendorf et aI., 1992).
In this section we categorize surface precipitation into three different types that are
commonly discussed in the literature. These include formation or sorption of metal poly-
mers (dimers, trimers, etc.) on the surface (Chisholm-Brause et aI., 1990); a solid solu-
tion or coprecipitate that involves coions dissolved from the sorbent; and a homogeneous
precipitate formed on the surface composed of ions from the bulk solution, or their hy-
drolysis products (Farley et aI., 1985). The continuum between surface precipitation
and chemisorption is controlled by several factors, including (1) the ratio of the number
of sites vs. the number of metal ions in solution, (2) the strength of the metal-oxide bond,
and (3) the degree to which the bulk solution is undersaturated with respect to the metal
hydroxide precipitate (McBride, in Bolt, 1991, p. 163). The different types of surface
precipitation are explained in more detail below.
Polymeric metal complexes that form at the surface, and/or the sorption of aqueous
polymers, may be a mechanism that typifies surface precipitate-like complexes (Fendorf
Sorption Kinetics of Trace Elements in Soils and Soil Materials 19
et al., 1992). Chisholm-Brause et al. (1990) interpreted the presence of Pb in the second
coordination shell of a sorbed Pb (determined using XAS) as small clusters or poly-
nuclear structures that are analogous with hydroxy metal complexes formed in water
solution. The formation of complete surface precipitates was ruled out because the num-
ber of Pb atoms in the second shell was small (0.3 to l.5). Bargar et al. (1997) observed
similar polymer formation at high loading levels for Pb on AI-oxide surfaces. Fendorf
and Sparks (1994) found that Cr polymerization, and eventually Cr-hydroxide surface
clusters, began at surface coverages as low as 20%. It was proposed that when the struc-
tures of the sorbate and sorbent are dissimilar, epitaxial growth is energetically unfavor-
able and thus nucleation growth is away from the surface, i.e., surface clusters.
The formation of a homogeneous solid on a surface can occur when a solution be-
comes saturated and the surface acts as a nucleation site, or from a chemisorption-pre-
cipitation continuum, i.e., when adsorption reaches monolayer coverage sorption continues
on the newly created sites resulting in a precipitate on the surface (multilayer surface
coverage) (McBride, in Bolt et al., 1991, p. 171; Farley et al., 1985). This phenomenon is
analogous to the assumptions used in the classical Brunauer-Emmett-Teller (BET) iso-
therm model of gas sorption onto surfaces (Borg and Dienes, 1992, p. 400). The distinc-
tion between a surface precipitate and a sorbed metal complex can be subtle, and somewhat
confusing, especially since polymer sorption can lead to, or preface, surface precipita-
tion. Adding to the difficulty of distinguishing surface precipitation from sorbed metal
complexes is the fact that methods for distinguishing between the two phenomena are at
present in their early development, and few studies exist on this subject matter.
The solid solution concept of surface precipitation was presented in detail by Farley et
al. (1985); it is described as a process similar to homogeneous coprecipitation. The com-
position of the surface precipitate varies, "continuously between that of the original solid
and a pure precipitate of the sorbing metal" (Farley et al., 1985). The solid solution
concept differs from the multilayer precipitation concept in that it includes both desorp-
tion and/or dissolution of structural ions from the sorbent and the inclusion of ions from
solution. The result of coprecipitation of the solution ions with ions dissolved from the
surface is a solid with isomorphic substitution, or a stable mixture of two solids (Stumm
and Morgan, 1996, p. 814). An important factor controlling which ions will form a solid
solution is the ionic radius. For example, Ainsworth et al. (1994) found that the extent of
reversibility with aging for Co, Cd, and Pb was inversely proportional to the ionic radius
of the ions, where ionic radii increase in the order Co < Cd < Pb. Since the ionic radius of
Co is the most similar to Fe, they concluded that the hysteresis was a result of the forma-
tion of a solid solution. Solid solution formation is probably limited by the rate of mineral
dissolution, rather than a lack of thermodynamic favorability (McBride, 1994, p. 163;
Scheidegger et al., 1998). O'Day et al. (1996) observed a small amount of Si backscat-
tering from the XAFS spectra of Co sorption on quartz (a-Si0
2
). They explained this by
proposing that Co was coordinated in Si tetrahedra, which occurred by either diffusion
to defect sites, or a small amount of quartz dissolution and reprecipitation of a mixed
Co/Si phase (solid solution).
Surface precipitation and dissolution are slower processes than adsorption and de-
sorption. Farley et al. (1985) noted that the rate of Cd uptake by amorphous iron hy-
droxide was lower when the initial solution concentration exceeded that required for
monolayer coverage. One possible reason for the slower precipitation reactions is that a
20 Fate an.d Transport of Heavy Metals in the Vadose Zone
precipitated ion must form several bonds, which requires more activation energy than
adsorption complexes which have fewer bonds. Likewise, surface precipitates may be
more stable than adsorbed species because of the formation of high energy bonds and
increased coordination. Another factor which makes surface precipitates more stable is
that only the surface of the precipitate is accessible to the solution for dissolution to
occur (for a 3-dimensional object only the exposed surfaces are surrounded by solution).
The formation of a surface precipitate involves several reactions, including (1) ad-
sorption of the ion on the surface, (2) surface nucleation, and (3) crystal growth (Stumm
and Morgan, 1996, p. 812). Each of these steps contains several independent reaction
sequences, and the rate of precipitate formation is determined by the slowest reaction
step. While the formation of surface precipitates is important for predicting the fate of
trace elements in the environment, dissolution reactions are also important processes
that may be the controlling mechanisms for trace element mobilization when a soil has
been contaminated for long periods. For the dissolution of surface precipitates the reac-
tion sequence is similar to the steps of dissolution of a pure solid: (1) transport of reac-
tants from the bulk solution to the surface, (2) adsorption of solutes, (3) interlattice
transfer of reacting species, (4) chemical reactions, (5) detachment of reactants from the
surface, and (6) mass transport into the bulk solution (Stumm and W ollast, 1990). These
steps can be summarized as transport and surface reaction mechanisms. The mechanism
controlling the rate of dissolution is dependent on several factors; i.e., solution composi-
tion, pH, mixing, etc.
The kinetics of surface precipitate formation and dissolution has not been extensively
studied. In a recent study by Scheidegger and Sparks (1996) the rate of release of Ni
from a pyrophyllite surface known to have Ni precipitates showed both a fast and slow
reaction. The fast reaction was attributed to desorption of specifically sorbed Ni. The
slow reaction was attributed to the slow dissolution of polynuclear Ni complexes, which
were found to dissolve more slowly than pure Ni(OHh In another study, Scheidegger
et al. (1998) monitored the kinetics of surface precipitate formation on pyrophyllite,
montmorillonite, and gibbsite using XAFS. Surface precipitate formation was initially
fast on pyrophyllite and gibbsite (within minutes), but did not occur until 48 hours on
montmorillonite. Figure 1.13 shows that Ni uptake by pyrophyllite is initially rapid, with
approximately 25% of the Ni being sorbed within the first 30 minutes. Then the reaction
slowed considerably, but was continuous for times as long as 72 hours (97% of the Ni is
removed from solution). Analysis of the radial structure function (Figure 1.14) derived
from XAFS spectroscopic characterization of the samples after different sorption peri-
ods shows an increase in a second shell at -2.75 A (uncorrected for phase shifts). This
suggests that the slow development of polynuclear Ni complexes is responsible for the
slow sorption reaction. These complexes have been identified as mixed Ni-Al (takovite-
like) hydroxide phases (Scheidegger et al., 1998).
O'Day et al. (1996) used XAS and kinetic experiments to hypothesize the mecha-
nisms of surface precipitation on two different minerals. Their hypotheses were strength-
ened by comparing and contrasting the spectroscopic and kinetics results for different
mineral surfaces. XAFS results for Co on rutile (Ti0
2
) showed an increase in the num-
ber of backscattering Co atoms for aging times of one day to 11 days, suggesting an
increase in the size of multinuclear complexes formed on the surface. However, similar
results were not seen for Co aging on quartz (a-Si0
2
), which had Co(OH)2 surface
100
90
;e 80
e..
c:
o 70

5
'0
en 60
E
e 50
-
'tJ 40
~
E 30
CD
a:
Z 20
10

Sorption Kinetics of Trace Elements in Soils and Soil Materials 21




O + - - - - - . - - - - . - - - - - . - - - - . - - - - - . - - - - . - - - - - . - - - - - r - - ~
o 24 48 72 96 120 144 168 192 216
Time (hours)
Figure 1.13. Kinetics of Ni removal from solution by pyrophyllite. From Scheidegger et al. (1997b),
with permission.
2.0
1.5
1.0
Q)
0.5
"'0
Months
:::J
0.0
~
c:
0)
rn
::E
E
...
0
24 Hours
-
en
c:
rn
...
I-
o
Figure 1.14. Radial distribution function (uncorrected for phase shifts) for Ni sorption on pyrophyl-
lite incubated for 24 hours to 3 months. From Scheidegger et al. (1997b), with permission.
22 Fate and Transport of Heavy Metals in the Vadose Zone
precipitates present. They hypothesized that the reason for the observed slow change in
the multinuclear surface precipitate on the rutile and not on the quartz was a result of the
similar radii between Co and Ti, while the radius of Si is larger than Co. As a result of
this difference in atomic radii, the Co hydroxide-like precipitate formed on quartz was
attached only to corners of select Si tetrahedra on the surface. However, Co sorption on
the rutile was consistent with the formation of a precipitate that had similar lattice di-
mensions as the surface, effectively extending the lattice structure of the bulk solid; i.e.,
an epitaxial growth. The resulting Co surface precipitate was structurally strained due
to charge imbalances and distortion of the Co0
6
octahedra. Thus, they proposed that the
change with increasing equilibration times was due to the formation of an anatase-like
structure (conclusion made based on similar octahedra bond distances between the ana-
tase and the surface precipitate). The anatase structure, a Ti0
2
polymorph, has favor-
able lattice dimensions for Co because of a more open structure which better
accommodates the slightly larger Co ion.
There are several thermodynamic reasons for the formation of surface precipitates in
unsaturated systems. For example, the solid surface may lower the energy of nucleation
by providing sterically similar sites (McBride, in Bolt, 1991, p. 171), the activity of the
surface precipitate is less than unity (Sposito, 1986), and the solubility of the surface
precipitate is lowered because the dielectric constant of the solution near the surface is
less than that of the bulk solution (O'Day et al., 1994a). It has not been established
which one of these mechanisms predominates; however, it is possible that the three phe-
nomena simultaneously influence precipitation reactions on surfaces. To ensure that pre-
cipitation is truly a surface-induced phenomenon, experimental systems should be run at
conditions undersaturated with respect to known phases. However, the solubility prod-
ucts of many possible phases are unknown, making it difficult to determine if such phases
will precipitate in a given system. This is particularly true for mixed cation hydroxides
and coprecipitation reactions on surfaces (d'Espinose de la Caillerie et al., 1995; Towle
et al., 1997; Scheidegger et al., 1998) The theories on the enhancement of surface pre-
cipitation by the three mechanisms mentioned above are discussed in more detail below.
As discussed above, Farley et al. (1985) presented a solid solution model to explain
the continuum between precipitation and chemisorption onto solid surfaces. The model
suggests that sorption is a process that includes solid dissolution, and then reprecipitation
onto the surface. Thus, the formation of a surface complex involves the coprecipitation
of both the ions dissolved from the sorbent, and the ions present in the bulk solution.
Therefore, assuming the solid phase is a pure crystal and has unit activity (relative to the
pure macro crystal) is an inappropriate assumption, and invalidates solubility determina-
tions based on the law of mass action, and ion activity products that do not account for
surface activity. In addition, the resulting surface complex may not be compositionally
homogeneous and completely free from inclusions, causing the activity of the solid phase
to be even lower (Sposito, 1986). Sposito (1986) illustrated this idea by considering the
dissolution of CdC0
3
:
(1)
Ion Activity Product (lAP) = [Cd][C0
3
] = K.o [CdC0
3
(s)] (2)
Sorption Kinetics of Trace Elements in Soils and Soil Materials 23
where the brackets indicate ion activity. The activity of a pure solid is often considered
to be one; however, if Cd forms a mixed precipitate (coprecipitate) with another ion in
solution, such as Ca, the result is not [CdC0
3
(s)] = 1, but Cd
x
C<lyC0
3
(s) with activity
d. This means that the lAP (including Ca) at the surface of the precipitate is greater
than the lAP of the bulk solution. Thus, Cd
x
C<lyC0
3
(s) will precipitate at the surface
before CdC0
3
. The actual lAP of a solid solution depends on the concentration of the
foreign ion (Ca in the above case) in the solid mixture phase:
(3)
where gj is the activity coefficient of solid i, ~ is the mole fraction of solid i and K; so is the
solubility product of the pure mineral i (Van Riemsdijk and Van der Zee, in Bolt, 1991,
p. 251). From this equation one sees that a coprecipitate with one constituent present in
minor amounts has a decreased solubility product with respect to the pure mineral.
The enhancement of precipitate formation on the surface may also be due to the re-
duction of the energy barrier necessary for nucleation processes to occur in an aqueous
solution. This is a factor of the lattice dimensions of the solid, and those of the precipitate
to be formed, i.e., a steric interaction (McBride, in Bolt et al., 1991, p. 171). The result of
surface nucleation sites is that the extent of supersaturation required for precipitation is
decreased. However, there may be other important factors contributing to surface nucle-
ation interactions. For example, Fendorf et al. (1992) observed AI surface precipitates
on Mn0
2
using high resolution transmission electron micrography (HRTEM), but not
on Ti0
2
under the same conditions (undersaturated with respect to the most likely AI
hydroxide precipitates, and equivalent surface coverages). If promotion of surface pre-
cipitation below saturation was a result of the presence of nucleation sites, then one
would expect to see precipitates on both surfaces. Thus, another factor inhibited surface
precipitate formation on the Ti0
2
, i.e., steric hindrances between the two surfaces.
The dielectric constant of the solution at the interface of a solid is much less than it
would be in the bulk solution (Hiemenz, 1986, p. 725). This is a result of the surface
charge aligning the dipoles of the water layers nearest the surface. This phenomenon is
called dielectric saturation, and results in a dielectric constant an order of magnitude or
less in the first few angstroms of the surface (McBride, 1994, p. 296). The activity of
individual ions in solution is inversely proportional to the dielectric constant of water.
Consequently, near the surface the lowered dielectric constant of the water causes an
increased ion activity, and the lAP near the surface will exceed that of the bulk solution.
O'Day et al. (1994a) used a more direct approach to this concept that calculates the
change in the free energy near the surface as a function of the dielectric constant using
the Born charging equation for a spherical ion. However, calculations for sorption of Co
on kaolinite revealed that the average surface dielectric constant is not low enough to
account for surface precipitation of a pure hydroxide phase as the predominant mecha-
nism of sorption. Thus, if surface precipitation was the mechanism of sorption, then
either the value of the dielectric constant near the surface was incorrect (possible, since
dielectric constants near surfaces are difficult to determine), or precipitation was en-
hanced by one of the aforementioned mechanisms (O'Day et al., 1994a).
Regardless of the mechanism, surface precipitation is an important process affecting
trace metal reactions with natural materials. Since surface precipitation kinetics can be
24 Fate and Transport of Heavy Metals in the Vadose Zone
slow, the extent of precipitation and subsequent dissolution of surface precipitates are
affected by residence time; thus, they are important slow kinetic mechanisms which can
control the fate of trace elements in the environment.
SUMMARY
In this chapter we have presented evidence that slow kinetics are important when
estimating the extent and reversibility of trace metal sorption on soils and soil materials.
We have also discussed three possible mechanisms for such slow kinetics: diffusion,
rate-limited adsorption processes, and precipitation reactions on surfaces. While evi-
dence exists that suggests all three mechanisms occur, the slow sorption mechanism
occurring in a particular system is highly dependent on the prevailing environmental
conditions; e.g., solution pH, sorbent characteristics, ionic strength, trace metal physico-
chemical characteristics, dissolution rate of the solid, and the microporosity of the solid.
In addition to a review of relevant studies in the literature, two examples from our
own research were given that suggested different mechanisms as the rate controlling
processes responsible for the slow sorption of metals on model soil components. In one
case Pb sorption on y-AI203 resulted in little change in the local chemical environment
(determined using XAFS) with increased incubation time. In another example, Ni sorp-
tion on the clay mineral pyrophyllite resulted in increased polynuclear surface complex-
ation with increasing reaction time. While it is difficult to make direct comparisons of
the two metals since the surfaces present are different, we think it is justifiable since
additional experiments (data not shown) for Pb on pyrophyllite (AIcacio, 1997) and Ni
on gibbsite (a form of AI hydroxide) (Scheidegger et al., 1997a) showed similar sorption
behavior to the systems being compared above. We propose that one reason for the
different apparent bonding mechanisms is the difference in the ionic radius of the two
metals (1.20 A for Pb and 0.69 A for Ni). The sorbents studied in these two cases have AI
present as a structural component. Since AI can dissolve and has a similar radius as Ni
they can form a coprecipitate, while the Pb ion is too large to form such a coprecipitate.
Ainsworth et al. (1994) observed that the extent of sorption reversibility was positively
correlated to the ionic radius of the sorbing metals. Coughlin and Stone (1995) also
suggested that coprecipitation of metal ions with Fe is directly dependent on ionic ra-
dius. While these studies do not provide direct evidence on the formation of a coprecipitate,
they agree well with our data; Ni (the smaller ion with an ionic radius similar to AI)
forms a mixed precipitate, while Pb does not. This information can be used to improve
predictions on the fate of these metals in the environment, and will allow for better
simulations in the laboratory.
When predicting the transport of trace elements in the vadose zone, researchers must
know the kinetics of sorption and desorption behavior. If slow kinetics are controlling
these mechanisms then reaction-transport models should include such chemical pro-
cesses. This will result in more accurate predictions and improved management of exist-
ing and potential risks. In addition, if the mechanism responsible for slow sorption or
desorption is known, researchers can design remediation strategies more efficiently. This
may include mobilizing or immobilizing contaminants based on the pH of the soil solu-
tion, treating the soil with chelating ligands, or creating treatments for specific exposure
times based on the kinetics of the reactions.
Sorption Kinetics of Trace Elements in Soils and Soil Materials 25
To obtain complete sorption and desorption kinetic behavior, researchers should con-
duct experiments in the lab for extended time periods. When possible, mechanisms of
slow kinetics should be determined to better predict the fate of trace elements in the
environment. To ascertain the mechanisms, both macroscopic and microscopic experi-
ments should be used. In this chapter we have presented several methods that have been
used for the determination of mechanisms. With the rapid advancement of technology
the future should bring an even better understanding of soil sorption mechanisms and
kinetic processes. It is critical that researchers combine their efforts with those in related
fields so that the most contemporary and valid models can be developed and employed to
predict the fate of trace elements in the environment.
REFERENCES
Ainsworth, C.C., J.L. Pilon, P.L. Gassman, and W.G. Van der Sluys. Cobalt, cadmium, and lead
sorption to hydrous iron oxide: Residence time effect. Soil Sci. Soc. Am. J. 58, pp. 1615-1623,
1994.
Alcacio, T.E. An XAFS Survey of Pb Complexes at the Gibbsite Solid/Liquid Interface. Masters
Thesis, University of Delaware, 1997.
Alexander, M. How toxic are chemicals in soils? Environ. Sci. Tec/moL. 29(11), pp. 2713-2717,
1995.
Axe, L. and P.R Anderson. Experimental and theoretical diffusivities of Cd and Sr in hydrous
ferric oxide. J. Collow Interface Sci. 185(2), pp. 436-448, 1997.
Backes, C.A., RG. McLaren, A.W. Rate, and RS. Swift. Kinetics of cadmium and cobalt
desorption from iron and manganese oxides. Soil Sci. Soc. Am. J. 59, pp. 778-785, 1995.
Bargar, J.R, G.E. Brown, Jr., and G.A. Parks. Surface complexation of Pb(II) at oxide-water
interfaces: I. XAFS and bond-valence determination of mono- and polynuclear Pb(II) sorp-
tion products on Al-oxides. Geochimica et CO.JmochimicaActa, 61(13), pp. 2617-2637, 1997.
Bargar, J.R, S.N. Towle, G.E. Brown, Jr., and G.A. Parks. Outer-sphere Pb(II) adsorbed at
specific surface sites on single crystal a-alumina. Geochimica et CO.Jmochimica Acta 60(18), pp.
3541-3547, 1996.
Benjamin, M.M. and J.O. Leckie. Multiple-site adsorption of Cd, Cu, Zn, and Pb on amorphous
iron oxyhydroxide. J. CoLww Interface Sci. 79(1), pp. 209-221, 1981.
Bibak, A., J. Gerth, and O.K Borggaard. Retention of cobalt by pure and foreign-element
associated goethites. Clay.J Clay Minerau 43(2), pp. 141-149, 1995.
Blackburn, W.H. and W.H. Dennen.PrincipluofMinerawgy. Wm. C. Brown Publishers, Dubuque,
lA, 1994.
Borg, RJ. and G.J. Dienes. The Pby.JicaL Chemutry of SoLw.J. Academic Press, Inc., London, 1992.
Bruemmer, G.W., J. Gerth, and KG. Tiller. Reaction kinetics of the adsorption and desorption
of nickel, zinc and cadmium by goethite. I. Adsorption and diffusion of metals. J. SoiL Sci. 39,
pp. 37-52, 1988.
Burgos, W.D., J.T. Novak, and D.P. Berry. Reversible sorption and irreversible binding of
naphthalene and a-Naphthol to soil: Elucidation of processes. Environ. Sci. TechnoL. 30(4), pp.
1205-1211, 1996.
Chisholm-Brause, C.J., KF. Hayes, A.L. Roe, G.E. Brown, Jr., G.A. Parks, and J.O. Leckie.
Spectroscopic investigation of Pb(II) complexes at the y-Al
2
0iwater interface. Geochimica et
CO.JmochimicaActa 54, pp. 1897-1909, 1990.
Comans, RN.J. Adsorption, desorption and isotopic exchange of cadmium on illite: Evidence for
complete reversibility. Water RM. 21 (12), pp. 1573-1576, 1987.
26 Fate and Transport of Heavy Metals in the Vadose Zone
Coughlin, B.R and A.T. Stone. Nonreversible adsorption of divalent metal ions (Mn", Co", Ni",
Cu", and Pb") onto goethite: Effects of acidification, Fe" addition, and picolinic acid addition.
Environ. Sci. Technol. 29(9), pp. 2445-2455, 1995.
d'Espinose de la Caillerie, J.B., M. Kermarec. and O. Clause. Impregnation of y-alumina with
Ni(II) and Co(II) ions at neutral pH: Hydrotalcite-type coprecipitate formation and charac-
terization. J. Am. Chem. Soc. 117, pp. 11471-11481. 1995.
Dzombak, D.A. and F.M.M. Morel. Sur/ace Complexation MOdeling, HYdrOLM Ferric Oxide. John
Wiley & Sons, New York, 1990.
Farley KJ., D.A. Dzombak, and F.M.M. Morel. A surface precipitation model for the sorption
of cations on metal oxides. J. Colloid Inter/ace Sci. 106(1), pp. 226-242, 1985.
Fendorf, S.E., M. Fendorf. D.L. Sparks, and R Gronsky. Inhibitory mechanisms of Cr(I1l)
oxidation by 8-Mn0
2
J. Colloid Inter/ace Sci. 153(1), pp. 37-54, 1992.
Fendorf. S.E., G.M. LambIe, M.G. Stapleton, M.J. Kelley, and D.L. Sparks. Mechanisms of
chromium(III) sorption on silica. 1. Cr(III) surface structure derived by extended X-ray
absorption fine structure spectroscopy. Environ. Sci. Technol. 28(2), pp. 284-289. 1994.
Fendorf, S.E. and D.L. Sparks. Mechanisms of Chromium(III) sorption on silica. 2. Effect of
reaction conditions. Environ. Sci. Technol. 28(2), pp. 290-297, 1994.
Fendorf, S.E., D.L. Sparks, and M. Fendorf. Surface precipitation reactions on oxide surfaces.
J. Colloid Inter/ace Sci. 148(1), pp. 295-298, 1992.
Fuller, C.C., J.A. Davis, and G.A. Waychunas. Surface chemistry offerrihydrite: Part 2. Kinetics
of arsenate adsorption and coprecipitation. Geochimica et COdmochimica Acta 57, pp. 2271-2282,
1993.
Griffin. RA., W.A. Sack, W.R Roy, C.C. Ainsworth, and LG. Krapac. Batch Type 24-h
Distribution Ratio for Contaminant Adsorption by Soil Materials. In HazardOLM anO IndLMtrial
Solid Wadte Tedting and DiJpodal, Vol. 6, D. Lorenzen, RA. Conway, L.P. Jackson, A. Hamza,
C.L. Perket. and W.J. Lacy, Eds., American Society for Testing and Materials. Philadelphia,
1986. pp. 390-408.
GrossL P.R. D.L. Sparks, and C.C. Ainsworth. Rapid kinetics of Cu(II) adsorption/desorption
on goethite. Environ. Sci. Techno!. 28(8), pp. 1422-1429. 1994.
Hayes, KF. and J.O. Leckie. Mechanism of Lead Ion Adsorption at the Goethite-Water
Interface. In Geochemical PrOCedded at Mineral Sur/aced, J.A. Davis and K.F. Hayes. Eds . Ameri-
can Chemical Society, Washington, DC, 1986, pp. 115-141.
Hayes. KF. Equilibrium, Spectroscopic, and Kinetic Studies of Ion Adsorption at the Oxide/
Aqueous Interface. Ph.D. Dissertation, Stanford University, 1987.
Hiemenz, P.C. Principled of Colloid and Sur/ace ChemiJtry. Marcel Dekker, Inc., New York. 1986.
Kim, Y., RJ. Kirkpatrick, and RT. Cygan. J33Cs NMR study of cesium on the surfaces of
kaolinite and illite. Geochimica et C(},jmochimicaActa 60(21), pp. 4059-4074. 1996.
Koskinen. W.C. and S.S. Harper. The Retention Process: Mechanisms. In Pedticided in the Soil
Environment: PrOCedded, Impactd, and MOdeling, H.H. Cheng, Ed., Soil Science Society of America,
Madison. WI, 1990, pp. 51-77.
Kuo, S. and D.S. Mikkelsen. Kinetics of zinc desorption from soils. Plant Soil 56, pp. 355-364,
1980.
Lehmann, RG. and RD. Harter. Assessment of copper-soil bond strength by desorption kinet-
ics. Soil Sci. Soc. Am. J. 48. pp. 769-772, 1984.
Loehr, R.C. and M.T. Webster. Behavior of fresh vs. aged chemicals in soils. J. Soil Contam. 5(4),
pp. 361-393, 1996.
Lookman, R, D. Freese. R. Merckx. K. Vlassak, and W.H. Van Riemsdijk. Long-term kinetics
of phosphate release from soil. Environ. Sci. Technol. 29(6). pp. 1569-1575, 1995.
McBride. M.B. Cu
2
+ characteristics of aluminum hydroxide and oxyhydroxides. Clayd and Clay
Minerau 30(1), pp. 21-28. 1982.
Sorption Kinetics of Trace Elements in Soils and Soil Materials 27
McBride, M.M. Processes of Heavy and Transition Metal Sorption by Soil Minerals. In Interac-
tWIM at the Soil Col!.JlJ-Soil Solution Interface, Vol. 190, G.H. Bolt, M.F.D. Boodt, M.H.B. Hayes,
and M.B. McBride, Eels., Kluwer Academic Publishers, 1991, pp. 149-176.
McBride, M.M. Environmental Chemiltry of SoiU. Oxford University Press, New York, 1994.
McKenzie, RM. The sorption of cobalt by manganese minerals in soils. AliA. J. Soil Ru. 5, pp.
235-246, 1967.
McKenzie, RM. The adsorption oflead and other heavy metals on oxides of manganese and iron.
AlMt. J. Soil Ru. 18, pp. 61-71, 1980.
McLaren, RG., D.M. Lawson, and RS. Swift. Sorption and desorption of cobalt by soils and
soil components. J. Soil Sci. 37, pp. 413-426, 1986.
McLaren, RG., J.G. Williams, and R.S. Swift. Some observations on the desorption and
distribution behavior of copper with soil components. J. Soil Sci. 34, pp. 325-331, 1983.
Noggle, J.H. PhYdical Chemiltry. Harper Collins, New York, 1989.
Nyffeler, U.P., H.Y. Li, and P.H. Santschi. A kinetic approach to describe trace-element distri-
bution between particles and solution in natural aquatic systems. Geochimica et COdmochimica
Acta 48, pp. 1513-1522, 1984.
O'Day, P., G.E. Brown, Jr., and G.A. Parks. X-Ray absorption spectroscopy of cobalt (II)
multinuclear surface complexes and surface precipitates on kaolinite. J. CofWlJ Interface Sci. 165,
pp. 269-289, 1994a.
O'Day, P., G.A. Parks, and G.E. Brown, Jr. Molecular structure and binding sites of cobalt (II)
surface complexes on kaolinite from X-ray absorption spectroscopy. CIaYd and Clay Minerau
42(3), pp. 337-355, 1994b.
O'Day, P.A., C.J. Chisholm-Brause, S.N. Towle, G.A. Parks, and G.E. Brown. X-ray absorption
spectroscopy of Co(II) sorption complexes on quartz (u-Si0
2
) and rutile (Ti0
2
). Geochimica
et COdmochimica Acta 60(14), p. 2515, 1996.
Padmanabham, M. Adsorption-desorption behavior of copper (II) at the goethite-solution inter-
face. AlMt. J. Soil Ru. 21, pp. 209-320, 1983.
Papelis, C. X-Ray Photoelectron spectroscopic studies of cadmium and selenite adsorption on
aluminum oxides. Environ. Sci. Technol. 29, pp. 1526-1533, 1995.
Pignatello, J.J. and B. Xing. Mechanisms of slow sorption of organic chemicals to natural
particles. Environ. Sci. Technol. 30(1), pp. 1-11, 1996.
Scheidegger, A.M., G.M. Lambie, and D.L. Sparks. Investigation ofNi sorption on pyrophyllite:
An XAFS study. Environ. Sci. TechnoL 30(2), pp. 548-554, 1996.
Scheidegger, A.M. and D.L. Sparks. Kinetics of the formation and the dissolution of nickel
surface precipitates on pyrophyllite. Chem. GeoL 132(1-4), p. 157, 1996.
Scheidegger, A.M., G.M. Lambie, and D.L. Sparks. Spectroscopic evidence for the formation of
mixed-cation hydroxide phases upon metal sorption on clays and aluminum oxides. J. Col!.JlJ
Interface Sci. 186, pp. 118-128, 1997a.
Scheidegger, A.M., G.M. Lambie, and D.L. Sparks. The kinetics of nickel sorption on pyrophyl-
lite as monitored by X-ray absorption fine structure (XAFS) spectroscopy. Journal de PhYdique
IV, 7(C2-773), 1997b.
Scheidegger, A.M., D.G. Strawn, G.M. Lambie, and D.L. Sparks. The kinetics of mixed Ni-Al
hydroxide formation on clays and aluminum oxides: A time-resolved XAFS study. Geochimica
COdmochimica Acta, in press, 1998.
Schultz, M.F., M.M. Benjamin, and J.F. Ferguson. Adsorption and desorption of metals on
ferrihydrite: Reversibility of the reaction and sorption properties of the regenerated solid.
Environ. Sci. TechnoL 21(9), pp. 863-869, 1987.
Smith, J. T. and RN.J. Comans. Modeling the diffusive transport and remobilization of CS
137
in
sediments: The effects of sorption kinetics and reversibility. Geochimica et COdmochimica Acta
60(6), pp. 995-1004, 1996.
28 Fate and Transport of Heavy Metals in the Vadose Zone
Sparks, D.L. and P.M. Jardine. Thermodynamics of potassium exchange in soil using a kinetic
approach. Soil Sci. Soc. Am. 1. 45, pp. 1094-lO99, 1981.
Sparks, D.L. Kinetic.J of SoiL Chemical Proce.JJe.J. Academic Press, San Diego, 1989.
Sparks, D.L. EnvironmentaL SoiL ChemiJtry. Academic Press, San Diego, 1995.
Sparks, D.L. Kinetics of Soil Chemical Phenomena: Future Directions. In Future ProJpectJ for SoiL
ChemiJtry. Soil Science Society of America Special Publication, P.M. Huang et al., Eds., Soil
Science Society of America, Madison, WI, in press, 1998.
Sposito, G. Distinguishing Adsorption from Surface Precipitation. In GeochemicaL Proce.JJe.J at
MineraL Surfaced, J.A. Davis and K.F. Hayes, Eds., American Chemical Society, Washington,
DC, 1986, pp. 217-228.
Sposito, G. The ChemiJtry of Soil!. Oxford University Press, New York, 1989.
Steinfeld, J.I., J.S. Francisco, and W.L. Hase. ChemicaL Kinetic.J and DymanicJ. Prentice Hall,
Englewood Cliffs, NJ, 1989.
Stumm, W. and J.J. Morgan. Aquatic ChemiJtry, ChemicaL EquiLihria and Rate.J in NaturaL WaterJ.
John Wiley & Sons, New York, 1996.
Stumm, W. and R. W ollast. Coordination chemistry of weathering: Kinetics of the surface-
controlled dissolution of oxide minerals. Rev. GeophYJ. 28(1), pp. 53-69, 1990.
Swift, R.S. and R.G. McLaren. Micronutrient Adsorption by Soils and Soil Colloids. In Interac-
tionJ atthe SoiL CoL"'w-SoiL SoLution Inter./ace, Vol. 190, G.H. Bolt, M.F.D. Boodt, M.H.B. Hayes,
and M.B. McBride, Eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, 1991,
pp. 257-292.
Towle, S.N., J.R. Bargar, G.E. Brown, Jr., and G.A. Parks. Surface precipitation of cobalt on
Al
2
0
3
.1. CoL"'w Interface Sci. 187, pp. 62-82, 1997.
Van Riemsdijk, W.H. and S.E.A.T.M. Van der Zee. Comparison of Models for Adsorption, Solid
Solution and Surface Precipitation. In InteractionJ at the SoiL CoL"'w-SoiL SoLution Interface, Vol.
190, C.H. Bolt, M.F.D. Boodt, M.H.B. Hayes, and M.B. McBride, Eds., Kluwer Academic
Publishers, The Netherlands, 1991, pp. 241-256.
Veith, J.A. and G. Sposito. On the use of the Langmuir equation in the interpretation of
"adsorption" phenomena. Soil Sci. Soc. Am. 1. 41, pp. 697-702, 1977.
Verburg, K. and P. Baveye. Hysteresis in the binary exchange of cations on 2: 1 clay minerals: A
critical review. ClayJ and CLay Mineral! 42(2), pp. 207-220, 1994.
Waychunas, G.A., B.A. Rea, C.C. Fuller, and J.A. Davis. Surface chemistry of ferrihydrite: Part
1. EXAFS studies of the geometry of coprecipitated and adsorbed arsenate. Geochimica et
COJmochimica Acta 57, pp. 2251-2269, 1993.
Young, T.C., J.V. DePinto, and T.W. Kipp. Adsorption and desorption of Zn, Cu, and Cr by
sediments from the Raisin River Valley. J. Great LaIee.J &d. 13(3), pp. 353-366, 1987.
CHAPTER 2
Adsorption Isotherms of Nickel
Acid Forest Soils
Franz Zehetner and Walter W. Wenzel
INTRODUCTION

In
Recent studies reveal that the deposition of Pb, As, Co, Cr, and Cd in Europe has
decreased since the early eighties while the level of Ni deposition has not changed sig-
nificantly (Schulte and Gehrmann, 1996; Schulte et al., 1996). Ni concentrations in soil
solutions may be controlled by either adsorption/desorption or by precipitation/dissolu-
tion processes. According to Ma and Lindsay (1995), adsorption/desorption may be im-
portant in uncontaminated soils oflow pH whereas precipitation/dissolution may operate
in soils of high pH and/or high Ni levels. Ni adsorption is influenced by a number of soil
factors, including CEC, pH, texture, CaC0
3
content, organic matter, sesquioxides, and
chelating agents (Adriano, 1986). The most important adsorbents in soils are organic
matter, layer silicates, hydrous oxides of Si, Al, Fe, and Mn, and carbonates. The range
of different (external and internal) surfaces available for Ni adsorption in soils results in
a continuum of adsorption sites with various affinities for the species of Ni. The most
strongly adsorbed species are the cationic species Ni
2
+ and NiOH+ (Uren, 1992).
In this contribution adsorption isotherms of Ni are presented for individual acid for-
est soils. Individual mass-based isotherms are only of limited value for modeling Ni mo-
bility in soils since determination for each individual soil is required. Therefore, the
potential of a general surface-based adsorption isotherm was tested using specific sur-
face area (SSA) instead of soil mass as the reference quantity. Since typically, cation
exchange capacity (CEC) rather than SSA is available from soil databases, a charge-
based approach was developed by further substituting SSA for CEC.
ADSORPTION
Definition
Adsorption, as defined by Everett (1972), is the process producing net accumulation
of a substance at the common boundary of two contiguous phases. In soil/soil solution
.,0
30 Fate and Transport of Heavy Metals in the Vadose Zone
systems the term adsorption refers to the accumulation of molecules and ions at the
interface between the soil solid phases and the soil solution forming two-dimensional
arrangements. Adsorption does not include three-dimensional processes such as (sur-
face) precipitation and diffusion into crystal structures (Sposito, 1989; Scheidegger and
Sparks, 1996a).
The adsorbed substance is termed an addorbate, the solid surface on which it accumu-
lates is the addorbent. A molecule or an ion in the soil solution that potentially can be
adsorbed is termed an ad(JOrptive.
The Diffuse Double-layer
Adsorption phenomena can be described by means of molecular adsorption models,
which are based on hypotheses about the interactions between an adsorptive and an
adsorbent resulting in a particular arrangement of an adsorbate on a surface.
The diffLMe double-layer model is the oldest molecular adsorption model. Gouy (1910)
and Chapman (1913) derived an equation describing the ionic distribution in the diffuse
layer formed on a uniformly charged plane surface. The classical approach to under-
standing the distribution of cations in the solution close to a negatively charged surface
is the application of diffuse double-layer theory, which provides a mathematical descrip-
tion of the decrease in electrical potential with increasing distance from a charged plane
surface. The typical distribution curve for cations away from a negatively charged sur-
face (Figure 2.1) can be considered as a balance between electrical attraction and a
diffusion away from the surface due to the concentration gradient that is established as a
result of the attraction.
In the diffuse double-layer theory, charge is the exclusively important property whereas
ion size, as well as the effects of hydration are not accounted for. Thus, the main use of
the diffuse double-layer approach is the description of effects at some distance from the
surface in a region beyond that in which surface complexes are formed. Stern (1924)
modified the model proposed by Gouy and Chapman accounting for the special nature
of the first layers (within about 0.5 nm from the surface) of counterions against the
charged surface compared to the truly diffuse ions beyond. This generally improves the
application of diffuse double-layer theory but still gives little explanation to the real
situation close to the surface. In summary, a diffuse layer of ions may extend from a
charged surface, but only after the ions in solution have formed any specific interaction
with the surface.
Adsorption Mechanisms
According to Sposito (1989), adsorption on soil particle surfaces can involve three
mechanisms, i.e., in the order of decreasing interaction strength, inner-dphere dur/ace com-
plexation, outer-dphere dur/ace complexation, and diffLMe ion addoclation (Figure 2.2). In inner-
sphere surface complexes, the molecule or ion is directly bound to a surface functional
group, involving ionic, as well as covalent bonding. Inner-sphere surface complexation
can be considered as the molecular basis for the term dpecific addorption, since covalent
bonding is significantly influenced by the particular electron configurations of the atoms
involved. In the less stable outer-sphere surface complexes at least one water molecule is
interposed between the surface functional group and the bound molecule or ion, e.g., by
(/)
c:
.Q
-
o
c:
o
..
e!
-- c:
~
c:
o
()
Adsorption Isotherms of Nickel in Acid Forest Soils 31
~ " " - - ~ " - ~ "
I Stern layer
I Diffuse layer
-. . - - - - - - - - - - - - - - ~ - - - - - - - - - - + . -------------
External solution
-
--
,....
/' Anions
Distance from surface
Figure 2.1. Distribution of cations and anions away from a negatively charged surface (after
Russell, 1 988).
the solvation shell of an ion. If a solvated ion does not form a complex with a charged
surface functional group, but balances charge in a delocalized sense, it is adsorbed in the
diffuse ion swarm. These ions are fully dissociated from surface functional groups and
are free to move about. The outer-sphere surface complexation and the diffuse ion asso-
ciation involve electrostatic bonding and can be considered as the molecular basis for the
term nOfMpecijic ad<fOrptwn due to the weak influence of the electron configuration on in-
teractions of solvated species, for which ion valence and surface charge should be critical
to determining adsorption affinity (Sposito, 1989). Outer-sphere surface complexes and
especially the diffuse ion swarm are involved in wn exchange reactions whereas inner-
sphere surface complexes are formed by chemi:Jorptwn, i.e., through chemical bonding.
Chemisorption as a mechanism retaining transition and heavy metals in nonexchange-
able forms in soils is described in McBride (1989). The type of surface complex affects
the rate and reversibility of adsorption reactions. Outer-sphere complexation is usually
rapid and reversible whereas inner-sphere complexation is slower and may be "irrevers-
ible" (Sparks, 1995). The affinity of a soil adsorbent for an adsorptive species will in-
crease with the tendency to form inner-sphere surface complexes.
There is a continuum between surface complexation (adsorption) and surface pre-
cipitation (Figure 2.2). At low surface coverage the former tends to dominate, and as
surface loading increases, the latter becomes the dominant mechanism (Scheidegger and
Sparks, 1996a). Recent studies (Fendorf et al., 1994; O'Day et al., 1994a, 1994b;
Scheidegger and Sparks, 1996b; Scheidegger et al., 1996a, 1996b) have revealed surface
precipitates of Cr(Ill), Co, and Ni hydroxides to form at metal surface loadings far
below a theoretical monolayer coverage and at pH values that would not, according to
thermodynamic solubility products, cause the metals to form hydroxide precipitates.
32 Fate and Transport of Heavy Metals in the Vadose Zone
s i
Stern layer
a d
Diffuse layer
w
zn2+ Diffuse
_ ion swarm
: : ~
Outer-sphere
surface complexes
Inner-sphere
surface complexes
o Metal
o Oxygen
cIo Water molecule
Surface precipitate
Figure 2.2. Schematic representation of possible reactions at a mineral surface (after Sposito, 1984;
Russell, 1988; Scheidegger and Sparks, 1996a). The s plane is that of surface hydroxyl groups, the
i and 0 planes are associated with inner-sphere and outer-sphere surface complexes, respectively,
and the d plane indicates the beginning of the diffuse layer.
Adsorption Isotherms of Nickel in Acid Forest Soils 33
Q Q
L-curve isotherm
S-curve isotherm
c c
C-curve isotherm
Q H-curve isotherm Q
c c
Figure 2.3. The four main classes of adsorption isotherms (after Sposito, 1984, 1989).
ADSORPTION ISOTHERMS
If the adsorbed amount (Q) of a substance and the corresponding concentration in
solution (C) are known at fIxed temperature (ilotherm) and applied pressure, the pairs of
Q and C can be plotted against one another with Q as the dependent variable, obtaining
an adsorption isotherm.
Classification
Adsorption isotherms for solutes in dilute solution can be classifIed according to ini-
tial slope (Giles et aL, 1960, 1974a, 1974b). The four main classes, which are shown in
Figure 2.3, are briefly described in Sposito (1984, 1989).
The S-curve isotherm is characterized by an initially small slope that increases with
adsorptive concentration. This indicates that at low concentration the affInity of the soil
solid phases for the adsorbate is less than that of the soil solution for the adsorptive.
Metal adsorption in soil may follow an S-shaped isotherm if dissolved organic com-
pounds form strong, nonadsorbing complexes with the metaL When the metal concen-
tration exceeds the complexing capacity of these ligands, adsorption on the solid soil
particles increases and the isotherm takes on its characteristic S-shape. The S-curve
isotherm occurs also as a result of interactions among adsorbed organic molecules. These
interactions (e.g., surface polymerization) cause the adsorbate to become stabilized on
the solid surface and lead to enhanced affInity of the surface for the adsorbate with
increasing amounts adsorbed.
34 Fate and Transport of Heavy Metals in the Vadose Zone
The L-curve isotherm, which typically is concave to the concentration axis, shows an
initial slope that does not increase with adsorptive concentration. This is the result of
high affinity of the soil solid phases for the adsorbate at low surface coverage combined
with a decreasing amount of adsorbing surface with progressing adsorption.
The H-curve isotherm represents an extreme version of the L-curve isotherm. Its
characteristic large initial slope indicates very high affinity of the soil solid phases for the
adsorbate, which is caused either by very specific interactions between the solid phases
and the adsorbate (inner-sphere surface complexation) or by significant van der Waals
interactions in the adsorption process.
The C-curve isotherm shows an initial slope that remains independent of adsorptive
concentration until the maximum possible adsorption. This type of isotherm is produced
either by a constant partitioning of a substance between the interfacial region and the
soil solution or by a proportionate increase in the amount of adsorbing surface as the
amount adsorbed increases.
The adsorption of ions by soil particles usually follows an L-curve isotherm. The math-
ematical description of this isotherm almost invariably involves either the Langmuir equa-
tion or the van Bemmelen-Freundlich equation. The development of these two equations
to describe solute adsorption from aqueous solutions was reviewed from a historical
point of view by Forrester and Giles (1972).
The Langmuir Equation
By means of kinetic theory, Langmuir (1918) developed an equation describing gas-
eous adsorption on plane solid surfaces (glass, mica, platinum). Langmuir assumed that
the surface of a solid possesses a finite number of adsorption sites. If a gas molecule
strikes an unoccupied site, it is adsorbed whereas if it strikes an occupied site, it is re-
flected back into the gas phase. This model leads to the concept of an upper limit of
adsorption which occurs when the surface of the solid is covered with a closely packed
mono-molecular layer of gas molecules. According to Langmuir, adsorption can be re-
garded as a chemical reaction, given by
A + S ~ AS (1)
and described as a dynamic equilibrium. A denotes some gaseous substance that is
adsorbed on a plane solid surface (S). At equilibrium, the rate of adsorption equals the
rate of desorption, i.e.,
(2)
where kads and kdes are the rate coefficients in the forward and reverse direction, respec-
tively. The rate of adsorption is proportional to the pressure of the gas (p) -i.e., the
amount of gas molecules per unit of volume, and to the fraction of the uncovered surface
(1 - 8). The desorption rate is proportional to the fraction of the surface covered (8),
that is obtained as
Adsorption Isotherms of Nickel in Acid Forest Soils 35
(3)
By introducing a constant quantity (adsorption coefficient K) for the ratio of the rate
coefficients (kad/kdeJ, Equation 3 can be rearranged to
8= Kp
I+Kp
(4)
Since the adsorption coefficient (K) is defined as the ratio of the forward to the reverse
rate coefficient, it can be regarded as equilibrium constant of the adsorption reaction
(Eq. 1) and directly related to the standard free energy change of the reaction ( ~ G O ) by
(5)
where R is the gas constant [8.3143 J K-
I
mol-I] and T is absolute temperature. The
fraction of the surface covered (8) can be expressed by the ratio of the adsorbed gas
volume (y) to the value of y that is approached asymptotically as p becomes arbitrarily
large; i.e., the maximum adsorption capacity (Yma,). Thus,
8=--L.
Ymax
(6)
Equations 4 and 6 can be combined, obtaining the nonlinear form of the traditional
Langmuir equation, given by
that can be linearized to
YmaxKp
y=
I+Kp
y
-=y K-Ky
max
p
(7)
(8)
by multiplying both sides by (lip + K) and solving for yip. At vel}' small values of p, K p
is negligibly small and Equation 7 can be rearranged to
y=YmaxKp
(9)
36 Fate and Transport of Heavy Metals in the Vadose Zone
The adsorbed gas volume (y) increases linearly with increasing pressure (p). At high
pressure, K p is much greater than 1 and Equation 7 can be simplified to
Y =Ymax
(10)
Under these conditions the maximum adsorption capacity is obtained.
For the description of ion adsorption reactions in soil/soil solution systems, the adsorbed
gas volume (y) is replaced by the adsorbed amount of the studied ion per unit soil (Q)
and the equilibrium pressure (p) by the equilibrium soil solution concentration of the ion
(C). Parameter b is introduced as a measure of the adsorption maximum instead of Ymax'
and the Langmuir equation is obtained as
and in its linear form as
Q= bKC
I+KC
Q
-=bK-KQ
C
(11)
(12)
Quotient Q/C is known as the distribution coefficient (K
d
), and parameter K determines
the magnitude of the initial slope of the isotherm and is therefore a measure of adsorp-
tion energy. The advantages of Equation 12 over other linear forms of the Langmuir
equation are discussed in Veith and Sposito (1977). A graph of Q/C against Q should be
a straight line with the slope -K and an x-intercept equal to b if the Langmuir equation is
applicable to the studied system. The typical shape of the nonlinear Langmuir isotherm
is shown in Figure 2.4.
Harter and Smith (1981) summarized the assumptions on which the theoretical model
of Langmuir is based:
The plane surfaces have a fixed number of only one kind of elementary space.
Each space is able to hold only one adsorbed molecule.
The surface is covered with a monolayer only.
The adsorption reaction is reversible.
Adsorbed molecules are not free to move laterally on the surface.
Adsorption energy is the same for all sites and is not dependent on surface
coverage.
There is no interaction between adsorbate molecules.
The probability of a molecule condensing on an unoccupied site or dissociating
from an occupied site is not affected by coverage of adjoining sites.
Many of these assumptions are not valid in soil/soil solution systems, where different
types of adsorption sites exist, and the surfaces can be covered by more than one layer.
Irreversible reactions are involved, adsorbed molecules can move laterally on the sur-
face showing interactions between each other, and the adsorption energy, as well as the
Adsorption Isotherms of Nickel in Acid Forest Soils 37
b ................................................................................................ .
Q
c
Agure 2.4. The Langmuir isotherm (b = adsorption maximum).
probability of adsorption or desorption are dependent upon surface coverage. The appli-
cation of the Langmuir equation for describing adsorption of solutes in dilute solution
requires further assumptions (Harter and Smith, 1981) which are seldom attained in
soil/soil solution systems:
There is no specific adsorption of the solvent by the surface.
There is no interaction between solvent and solute (Henry's law obeyed).
Adsorption sites are empty "holes" having an activity coefficient of 1, as well as
o entropy and adsorption energy.
Nevertheless, the Langmuir equation has been widely used in the literature, as reviewed
by Travis and Etnier (1981). Adsorption data of ions in soils often show the shape of
Langmuir isotherms and can be accurately described with the Langmuir equation. This
may be explained according to Brunauer et al. (1967), who pointed out that the deriva-
tion of the Langmuir equation does not require all of the usual assumptions if it is devel-
oped, as originally by Langmuir (1918), on the basis of kinetic arguments instead of
statistical thermodynamics, which has imposed further restrictions on the original as-
sumptions. Moreover, Brunauer et al. (1967) noted that for adsorption on energetically
heterogeneous surfaces, the free energy of adsorption would decrease with increasing
surface coverage whereas the lateral interaction energies between adsorbed molecules
would raise the free energy of adsorption with increasing surface coverage. In certain
cases the two opposing effects would compensate for each other, making the free energy
of adsorption approximately constant for a particular adsorption isotherm.
Veith and Sposito (1977) found the Langmuir equation to apply, on the basis of statis-
tical analysis, even when the concentration of the studied ions was much larger than the
threshold value required to initiate precipitation. Thus, the Langmuir equation cannot
be used statistically to determine whether adsorption or precipitation is occurring and
the Langmuir parameters cannot be interpreted in terms of surface reactions unless there
is independent evidence that only adsorption is involved. Furthermore, the calculation
of the maximum adsorption capacity can involve errors of 50% and more (Harter, 1984)
38 Fate and Transport of Heavy Metals in the Vadose Zone
if the isotherm does not have the correct Langmuir shape and only low concentration
data are used for the calculation.
Although, according to Equation 12, a linear relationship is expected between QlC
and Q, a curvilinear relationship is frequently reported, which has been attributed to
nonuniformity of soil adsorption sites with respect to adsorption energy (Holford et al.,
1974; Holford, 1978) or to the effect of desorbed ions in the equilibrium solution on the
competition for adsorption sites (Griffin and Au, 1977; Harter and Baker, 1977, 1978).
Holford et al. (1974) used the equation proposed by Langmuir (1918) for gaseous
adsorption on surfaces with more than one type of elementary space of different adsorp-
tion energies, given by
(13)
and described o-phosphate adsorption by 41 soils of widely varying physical and miner-
alogical properties almost perfectly with the equation
Q= b1K1C + b2K2C
I+K
1
C I+K
2
C
(14)
where subscripts indicate different discrete energy adsorption sites. The traditional
Langmuir equation (Eq. 11) fitted none of the adsorption data as well as the "two-sur-
face" Langmuir equation (Eq. 14). However, according to Posner and Bowden (1980)
and Sposito (1982), the adjustable parameters in Equation 14 (b
l
, K
I
, b
2
, and K
2
) cannot
be interpreted in terms of surface reactions unless there is independent evidence for two
types of surface sites involved in the adsorption reaction.
For the description of Zn adsorption by soils, Harter and Baker (1977) used the equa-
tion proposed by Boyd et al. (1947), who modified the Langmuir equation to describe
simultaneous competitive adsorption of two equally charged cations, obtaining
(15)
where subscripts 1 and 2 refer, respectively, to the adsorbate ion and the ion originally
on the adsorbent surface. The calculated adsorption maxima (b) were similar to those
obtained from the traditional Langmuir equation (Eq. 11). Thus, the use of the tradi-
tional Langmuir equation seems satisfactory for the calculation of adsorption maximum
values that are subsequently used for correlation with chemical and physical soil proper-
ties. However, according to Harter and Baker (1977), adsorption energy (K) is signifi-
cantly affected by desorbed ions and should therefore be calculated by Equation 15.
Following suggestions from Holford (1978), Harter and Baker (1978) combined Equa-
tions 15 and 14, accounting for heterogeneity of adsorption energy over the entire sur-
face. The resulting equation is given by
Adsorption Isotherms of Nickel in Acid Forest Soils 39
(16)
where subscripts 1 and 2 refer to the adsorbate ion and the ion originally on the adsor-
bent surface, respectively, and the single and double primes indicate different discrete
energy adsorption sites.
The van Bemmelen-Freundlich Equation
In the works of van Bemmelen (1888) and Freundlich (1909), an empirical equation
was developed which applies to the adsorption of trace amounts of molecules or ions on
solid surfaces. For gaseous adsorption, the van Bemmelen-Freundlich equation has the
form
A
lin
y= p
(17)
where A and lin are positive-valued empirical constants that decrease with increasing
temperature, with n being greater than unity in most cases of practical interest.
For soil/soil solution systems, Equation 17 can be rearranged to
Q=AC
13
(18)
with 13 constrained to lie between 0 and 1. By logarithmizing Equation 18, the linear
form of the van Bemmelen-Freundlich equation is obtained as
log Q = log A + 13 log C
(19)
Thus, a plot of log Q against log C should be a straight line, with the slope 13 and a
y-intercept equal to log A if the van Bemmelen-Freundlich equation is applicable to the
studied system. The typical shape of the nonlinear van Bemmelen-Freundlich isotherm
is shown in Figure 2.5.
Although the van Bemmelen-Freundlich equation is empirical, it was tried to estab-
lish a theoretical derivation. Actually the equation may be obtained from Langmuir's
theory of monolayer adsorption by assuming that the free energy of adsorption decreases
logarithmically with increasing surface coverage (Halsey and Taylor, 1947), which may
be caused by particle interactions or by surface heterogeneity. The van Bemmelen-
Freundlich equation does not provide direct information on adsorption capacity and
adsorption energy, however, according to Slejko (1985), the constants A and 13 may be
regarded as relative indicators of capacity and affinity, respectively.
The adsorption of ions on soil particle surfaces can often be described accurately by
the van Bemmelen-Freundlich equation, which has been extensively used in the litera-
ture, as reviewed by Travis and Etnier (1981). However, the same authors noted that
the flexibility of the two constants allows for easy curve fitting, but does not guarantee
accuracy if the data are extrapolated beyond the experimental points. Furthermore,
Hemwall (1957) observed the van Bemmelen-Freundlich isotherm to apply also to pre-
40 Fate and Transport of Heavy Metals in the Vadose Zone
----.... - - - . - - . ~
Q
c
Figure 2.5. The van Bemmelen-Freundlich isotherm.
cipitation reactions. Thus, it cannot be used to differentiate between adsorption and
precipitation.
In the range of low adsorption, the van Bemmelen-Freundlich isotherm is similar to
the Langmuir isotherm, and with proper choice of constants both isotherms can almost
be brought to coincidence. Schulte (1988) reported that both equations accurately de-
scribed the adsorption of heavy metals in soils. However, neither of the two equations
provides information on the adsorption mechanisms involved or even whether adsorp-
tion or precipitation has occurred. Thus, both may be regarded as curve-fitting models
without particular molecular significance, but with predictive capability under limited
conditions (Sposito, 1989).
CASE STUDY
Five forest sites in the Giinser Mountain region, eastern Austria, were selected for the
study. The mean annual precipitation in the region is 1000 mm and the mean annual
temperature is 8C. The parent material of the soils is phyllite. Soil samples (n = 20)
were collected according to genetic horizons, air-dried, and passed through a 2-mm sieve.
The screened soils were mixed to obtain homogeneous samples and stored in plastic bags
until analysis. The studied soils represent acid forest soils [pH(CaCI
2
) between 2.9 and
4.7] showing high variation in organic carbon content, specific surface area, and cation
exchange capacity. Soil texture, however, displays low variation. The characteristic soil
properties are given in Zehetner (1997).
Particle size analysis of the fraction < 2 mm was carried out by a combined sieve and
pipette method (Blum et al., 1996). The specific surface area (SSA) was determined by
means of a single point in the water adsorption isotherm according to Puri and Murari
(1964), who found for various soils and clays that a mono-molecular water layer was
completed in equilibrium with a relative water vapor pressure (p/po) of 0.S273. Three
grams of air-dried soil were equilibrated with p/po = 0.S2 maintained by a saturated
Mg(N0
3
h solution at 24.SoC. The samples were weighed, dried to constant weight at
lOsoC, and weighed again. The time required for equilibration (seven days at p/po = 0.S2
and 48 hours at lOsoC, respectively) was determined in preliminary experiments. The
specific surface area [m
2
g-l] was calculated by the equation
Adsorption Isotherms of Nickel in Acid Forest Soils 41
(20)
where X [g g-I] is the amount of water adsorbed per mass soil in equilibrium with p/po =
0.52, N is Avogadro's number [6.02205 X 10
23
mol-I], A is the area occupied per mol-
ecule of water [1.08 X 10-
19
m
2
], and M(H
2
0) is the molecular weight of water [18.015 g
mol-I]. For the maintenance of a constant relative water vapor pressure, a saturated salt
solution is preferable to conc. H
2
S0
4
since the concentration of the latter changes due to
evaporation and condensation during the tests. The experiments were conducted with-
out vacuumizing (Ponizovskiy et al., 1993) and by applying a desorption rather than an
adsorption method (Farrar, 1963) in order to avoid irreversible changes in soil struc-
ture.
Soil pH was measured in a 1:2.5 soil:O.Ol M CaC1
2
extract after two hours of equili-
bration using a combination pH electrode (Blum et al., 1996). The cation exchange ca-
pacity (CEC) at soil pH was calculated as the sum of Al
3
+, Ca
2
+, Fe
3
+, H+, K+, Mg2+,
Mn
2
+, and Na+ extracted by 0.1 M BaCl
2
at 1:20 (Blum et al., 1996), and corrected for
H+ due to Al hydrolysis (Meiwes et al., 1984). The sum of mono- and divalent cations in
the BaCl
2
extract was calculated and termed CEC(2+)' Soil organic carbon (OC) corre-
sponds to the total carbon content measured with a CNS analyzer since the studied soils
are carbonate-free. Oxide-bound Al, Fe, and Mn were extracted with the bicarbonate-
citrate-dithionite method according to Mehra and Jackson (1960).
For each soil sample the quantitative soil:solution ratio necessary to obtain a soil paste
at the Atterberg "upper plastic limit" was determined with the liquid limit apparatus
described in Hillel (1980). It varied between 1:0.4 and 1:0.7 for the studied soils, corre-
sponding to their specific surface area. The determined soil:solution ratio of each soil
was used for a saturated water extract (equilibrium soil solution) and for the adsorption
experiments. Aqua bidest. and Ni solutions, respectively, were added to 100 g of the air-
dried and screened soil. The soil slurry was repeatedly stirred during the 48-hour-equili-
bration period at 20C, which was determined in preliminary tests. The solution was
separated from the soil by centrifugation and membrane filtration (cellulose acetate;
0.45 pm).
In the equilibrium soil solution, major cations were measured by ICP-AES, Ni by
GF-AAS, and anions by ion chromatography. pH and electrolytic conductivity (EC)
were determined, and dissolved organic carbon (DOC) was estimated by correlation
between DOC and absorbance at 254 nm measured with a UV-VIS spectrophotometer
(Brandstetter et al., 1996). The properties of the equilibrium soil solutions are given in
Zehetner (1997).
For each soil a Ni adsorption isotherm with 13 points was obtained by applying NiC1
2
solutions in increasing concentrations (between 2 X 10-
7
and 10-
2
mol L-
I
). Depending
on the quantitative soil:solution ratio and on the concentration of Ni in the added solu-
tion, between 0.08 and 7000 }lmol Ni per kg soil were added. The soluble Ni was mea-
sured with Flame- and GF -AAS. The total amount ofNi retained by the soil was calculated
by the difference between the initial and the final concentration in each solution.
The horizons of a spodosal (soils no. 8-12) were used for adsorption experiments at
soil:solution ratio of 1:5. NiCl
2
solutions of 10-
2
, 10-4, and 2 X 10-6 mol L-
I
, and a blank
42 Fate and Transport of Heavy Metals in the Vadose Zone
solution (aqua bidest.) were added to 8 g of the air-dried and screened soil. The obtained
soil suspensions were shaken for 48 hours at 20C, centrifuged and passed through mem-
brane filters (cellulose acetate; 0.45 pm).
For the calculation of adsorption isotherms, the initial amounts of potentially mobile
Ni on the soil solid surfaces have to be taken into account. They were estimated by the
fraction extractable with 0.05 M Na2EDTA at 1:10 according to Blum et al. (1996).
Since EDTA does not attack the silicate structures (Brummer et aI., 1986; Konig et aI.,
1986), it may be considered as an extractant for potentially mobile heavy metals (Schulte,
1988, 1994a, 1994b; Schulte and Beese, 1994a, 1994b).
The horizons of a spodosal (soils no. 8-12) were used for fractionating adsorbed Ni.
Subsequently to the adsorption experiments (saturation extracts with NiCl
2
solutions of
10-
2
, 10-4, and 2 X 10--6 mol L-
i
, and aqua bidest. as blank solution) an aliquot of each soil
was washed chloride-free, air-dried and sequentially extracted by a seven-step sequen-
tial extraction procedure (Table 2.1) developed by Zeien and Brummer (1989). The
residual fraction (fraction 7) was obtained by microwave digestion with H
2
0
2
IHN0
3
instead of the procedure suggested by Zeien and Brummer (1989) using HCIO/HN0
3
or HF/HCI0
4

Adsorption versus Precipitation
Sadiq and Zaidi (1981) suggested the precipitation of nickel ferrite (NiFe204) to be
the major retention process of Ni in 27 studied soils (pH range from 5.5 to 8.6), and
Sadiq and Enfield (1984a, 1984b) concluded from theoretical and experimental studies
on the solid phase formation ofNi that NiFe204 was the most stable solution controlling
solid species of Ni in soils. Other solid phases of Ni including carbonate, phosphate,
sulfates, halides, oxides and hydroxides, sulfides, as well as nickel silicate (Ni
2
Si0
4
) and
nickel aluminate (NiAl
2
0
4
) , were metastable with respect to nickel ferrite under the
conditions most commonly encountered in soils. However, according to Ma and Lindsay
(1995), NiFe204 has not yet been observed in soils and the conditions necessary for its
formation have not yet become clear. Thus, the validity of NiFe204 as a mineral control-
ling Ni
2
+ activities in soil solutions needs further examination.
In the presented study the soil solutions were undersaturated with respect to NiFe204
even at the highest applied initial Ni concentration (10 mmol L-
i
), assuming Fe
3
+ activity
in equilibrium with soil-Fe according to Lindsay (1979). Thus, adsorption including ion
exchange is likely to be the process that controlled solution concentrations of Ni under
the studied conditions.
langmuir and van Bemmelen-Freundlich Isotherms
The traditional, as well as the "two-surface" Langmuir equations (Eqs. 11 and 14,
respectively) were fitted to the experimental data of individual soils by means of an
iterative nonlinear x2-technique. The results of the fitting procedure are presented in
Zehetner (1997). Nonlinear fitting of Equation 11 was preferred over linear regression
according to Equation 12 since data variability is reduced in a graph of Q/C against Q.
The quality of fit was improved, especially in the low concentration range, by applying
the "two-surface" instead of the traditional Langmuir equation. This is evident from
obtained X2 values, which are lower by an order of magnitude for the "two-surface"
Table 2.1. Sequential Extraction Procedure Used for the Fractionation of Adsorbed Ni
a
Extraction Soil:Solution
Fraction Mobility and Binding Form Extractant Conditions Ratio Wash Step
1 Mobile fraction: Water-soluble 1 M NH
4
N0
3
24 hr shaking at 20
0
( 1 :25
and exchangeable (Le., (unbuffered)
nonspecifically adsorbed)
metals as well as easily
soluble metal-organic
complexes
2 Easily mobilizable fraction: 1 M NH
4
0Ac 24 hr shaking at 20
0
( 1 :25 1 M NH
4
N0
3
(unbuffered) at
More specifically adsorbed (pH 6.0) 1 :12.5; 10 min shaking at
metals, close to the surface 20
0
(
;x:.
occluded and carbonate
Q.
VI
bound metals as well as
0
....
"t:I
less stable metal-organic
....
o
complexes ::
3 Metals bound to Mn-oxides 0.1 M NH
2
OH-H(1 30 min shaking at 20
0
( 1 :25 2 times 1 M NH
4
0Ac (pH 6.0)
iii
0
+ 1 M NH
4
0Ac at 1 :12.5; 10 min shaking at
....
::r
(pH 6.0) 20
0
(
fb
....
4 Organically bound metals 0.025 M NH4EDTA 90 min shaking at 20
0
( 1 :25 1 M NH
4
0Ac (pH 4.6) at
3
VI
(pH 4.6) 1 :12.5; 10 min shaking at
0
-+.
20
0
(
z
5 Metals bound to amorphous 0.2 M NH
4
-oxalate 4 hr shaking at 20
0
( 1 :25 0.2 M NH
4
-oxalate (pH 3.25)
;::;.
;s:;
Fe-oxides (pH 3.25) in the dark at 1 :12.5; 10 min shaking at
!:E.
:i"
20
0
( in the dark
;x:.
6 Metals bound to crystalline 0.1 M ascorbic acid 30 min shaking in a water 1 :25 0.2 M NH
4
-oxalate (pH 3.25)
,.,
0:
Fe-oxides + 0.2 M NH
4
-oxalate basin at 96 3( in the at 1 :12.5; 10 min shaking at
."
(pH 3.25) light 20
0
( in the dark
0
..,
fb
7 Residual fraction 65% HN0
3
+ Microwave digestion 1 :20
VI
....
30% H
2
0
2
Vi
9.
a Adapted from Zeien and Briimmer, 1989.
iii
J:;;.
w
44 Fate and Transport of Heavy Metals in the Vadose Zone
10
....-.
......
,
en
0.1
.::s:.
0
E
E 0.01
.........
0
0.001
AEh (soil no. 8)
- - - - - Langmuir
--"two-surface" Langmuir
- - - van Bemmelen - Freundlich
0.001 0.01 0.1
C [mmol L."1]
10
Figure 2.6. Adsorption isotherms fitted to the experimental data of soil no. 8.
10
Bw (soil no. 11)
....-.
......
,
en
0.1
.::s:.
0
E
E
0.01
.........
0
0.001
- - - - - Langmuir
--"two-surface" Langmuir
- - - van Bemmelen - Freundlich
0.001 0.01 0.1 10
C [mmol L-
1
]
Figure 2.7. Adsorption isotherms fitted to the experimental data of soil no. 11.
Langmuir equation compared to the traditional form (Zehetner, 1997), and can be seen
in Figures 2.6 and 2.7 for a topsoil and a subsoil horizon (soils no. 8 and 11). In the "two-
surface" Langmuir equation, subscript 1 indicates surfaces with low adsorption maxi-
mum and high adsorption energy, and subscript 2 indicates surfaces with high adsorption
maximum and low adsorption energy (Zehetner, 1997).
The linear form of the van Bemmelen-Freundlich equation (Eq. 19) was fitted to the
experimental data of individual soils by linear regression, obtaining correlation coeffi-
cients (R) ~ 0.988, and significance at the 0.001 level (Zehetner, 1997). For soils no. 8
Adsorption Isotherms of Nickel in Acid Forest Soils 45
--------------- --_._-- ---- -- ------.. - - ---- -------- ----- ---.. -
50
0
0
AEh (soil no. 8) 0
40
0
0
Bw (soil no. 11)
0
0

,........,
0 ..-
30
,
Ol
0
.::t:- o

"0 20
0

10
0





0
0.0001 0.001 0.01 0.1 1 10
Initial Ni concentration [mmol L-
1
]
Figure 2.8. Effect of applied initial Ni concentration on distribution coefficient values for soils
no. 8 and 11.
and 11, the resulting van Bemmelen-Freundlich isotherms are shown in Figures 2.6 and
2.7, respectively.
Correlation coefficients of calculated adsorption isotherm parameters and the soil prop-
erties SSA, CEC, and CEC(2+) are presented in Zehetner (1997). Except for b
l
, the R
values were lowest for correlation with CEC, intermediate with SSA, and highest with
CEC(2+). Thus, ion exchange against mono- and divalent cations was possibly a major
mechanism of Ni adsorption under the studied conditions whereas sites occupied by
trivalent cations like Al
3
+ and Fe
3
+ were apparently less involved. Partial hydrolysis of
trivalent cations may result in particularly strong adsorption due to decreased hydration
and increased charge intensity. The adsorption maximum of the high energy sites (b
l
)
displayed the tightest correlation with SSA, followed by CEC and CEC(2+). This may be
due to specific adsorption on organic surfaces, of which SSA is a better measure than
CEC and CEC(2+)' as indicated by correlation with OC (R = 0.957***, 0.841 ***, and
0.781 ***, respectively).
The adsorption energy parameters (K, Kl' and K
2
), which may be considered as pH-
dependent (Schulte, 1988; and Beese, 1994a, 1994b), did not correlate with
pH(CaCI
2
) nor with pH(H
2
0). This may be due to the relatively narrow pH range of
the studied soils. EDTA-extractable Ni correlated with KI (R = but not with
K2 and K. This indicates that initially adsorbed Ni is predominantly bound to high en-
ergy sites, which may be caused by specific adsorption on organic surfaces.
As shown in Figure 2.8 for soils no. 8 and 11, distribution coefficients were
higher for topsoils than for subsoils and decreased with increasing applied initial Ni
concentration. This is consistent with the results of Basta and Tabatabai (1992), and
may be attributed to affinity of Ni for highly selective sites (specific adsorption) at low
concentration and to adsorption on less selective sites (nonspecific adsorption) at larger
amounts of Ni added.
In topsoil horizons (OC > 15 g kg-I) organic carbon content (OC) was the soil con-
stituent mainly affecting adsorption capacity parameters whereas in subsoil horizons
46 Fate and Transport of Heavy Metals in the Vadose Zone
10

AEh (soil no. 8)
0
0
......
.,....

0
I
Q)
0
.::t:.
0
0.1
E
0
E
0
........
. ~ o
a
0.01
0 saturation extract

1 : 5 extract
0.001
0.0001 0.001 0.01 0.1 10
C [mmol L-
1
]
Figure 2.9. Quantity-intensity relationships of Ni obtained in saturation extracts and at 1 :5 for soil
no. 8.
(oe < 15 g kg-I) dithionite-extractable Mn (Mn
d
) was the major influencing constitu-
ent. Correlation coefficients are presented in Zehetner (1997). In the topsoils, oe dis-
played tighter correlation with b
l
than with b
2
whereas in the subsoils, Mnd showed the
opposite trend. Thus, organic matter may be strongly involved in (specific) adsorption
of Ni on high energy sites. Neither in topsoils nor in subsoils did clay content show
significant correlations with adsorption capacity parameters. This may be due to the
relatively low variation in the clay content of the studied soils and the old age of the soil
material, indicated by low feldspar and smectite contents and high kaolinite contents,
which was determined for selected soils by X-ray diffraction.
Effect of Soil:Solution Ratio on Quantity-Intensity Relationships
The adsorption experiments were conducted in saturation extracts at soil:solution
ratios between 1 :0.4 and 1 :0.7, which are closer to field conditions than tighter soil:solution
ratios most commonly applied in adsorption experiments in the literature. For the hori-
zons of a spodosal (soils no. 8-12), the adsorption isotherms obtained in saturation ex-
tracts were compared with quantity-intensity relationships at a soil:solution ratio of 1:5
(Figures 2.9 to 2.13).
In the low concentration range (Ni solution concentrations in the order of 10-4 to 10-.3
mmol L -I), the quantity-intensity relationships were hardly affected by the soil:solution
ratio. At applied initial Ni concentrations of 0.1 and 10 mmol L -I, however, the amounts
of Ni adsorbed per mass soil were significantly higher in the 1:5 extracts than in the
saturation extracts, and despite a slighter decrease of Ni solution concentrations due to
adsorption in the 1:5 extracts, the obtained datapoints were significantly above the ad-
sorption isotherms obtained in saturation extracts. Thus, in the low concentration range
quantity-intensity relationships may be satisfactorily described by using tighter
soil:solution ratios whereas at higher concentrations these would cause false assessment
of adsorption. At a given solution concentration the amount adsorbed per mass soil would
......
....
'0>
::t:.
o
E
E
.........
a
Adsorption Isotherms of Nickel in Acid Forest Soils 47
10
Bhs (soil no. 9)
o
0.1
0.01
~ o
o
o

o
o
o
o saturation extract
1: 5 extract
0.001 +-...,.......,...,........,.".---..-........... rTTTTr-....-r-rTTTnr-.....-......... n-nr--,-.......,..........."
0.0001 0.001 0.01 0.1 10
C [mmol L-
1
]
Figure 2.1 O. Quantity-intensity relationships of Ni obtained in saturation extracts and at 1 :S for soil
no. 9.
10
Bs (soil no. 10)

0
0
......
....
,
0>
0
0
::t:.
0
0.1
E
0
E
0 .........
a
0.01
~ o
0 saturation extract

1 : 5 extract
0.001
0.0001 0.001 0.01 0.1 10
C [mmol L-
1
]
Figure 2.11. Quantity-intensity relationships of Ni obtained in saturation extracts and at l:S for
soil no. 10.
be overestimated, and at a given adsorbate concentration the corresponding equilibrium
solution concentration would be underestimated. Thus, soil:solution ratios of 1:5 and
tighter, used in the majority of (Ni) adsorption studies in the literature, may result in
quantity-intensity relationships that strongly overestimate adsorption, especially at higher
concentrations, and are therefore of limited value for modeling metal mobility in natural
systems.
Specific adsorption may be responsible for the observed low influence of the
soil:solution ratio on quantity-intensity relationships in the low concentration range. In-
volving covalent and ionic bonding, specific adsorption can be regarded as a chemical
48 Fate and Transport of Heavy Metals in the Vadose Zone
.......
......

10
0.1
E
........
a
0.Q1
0.0001
Bw (soil no. 11)

o
0.001
o
o
o
0
o
o
o saturation extract
1: 5 extract
0.01 0.1 10
C [mmol L-
1
]
Figure 2.12. Quantity-intensity relationships of Ni obtained in saturation extracts and at 1:5 for
soil no. 11.
10
Cw (soil no. 12)

..........
......
'0)

0.1
E
........
a 0.01
o
o
o
0
o
o
o
o saturation extract
1: 5 extract
o. <i 1
0.0001 0.001 0.01 0.1 10
C [mmol L-
1
]
Figure 2.13. Quantity-intensity relationships of Ni obtained in saturation extracts and at 1:5 for
soil no. 12.
equilibrium reaction, establishing a certain solution concentration at a given adsorbate
concentration regardless of the soil:solution ratio. However, at higher concentrations,
where nonspecific coulombic interactions tend to dominate, adsorption was enhanced
due to larger amounts of adsorptive per mass soil at 1:5 compared to saturation extracts.
Applied initial Ni concentration of 10 mmol L-
i
provided 4 to 7 mmol Ni per kg soil in
saturation extracts depending on the respective soil:solution ratio, but as much as 50
mmol Ni per kg soil at 1 :5.
Differences in ionic strength and pH may contribute to enhanced Ni adsorption at 1:5
compared to saturation extracts. In 1:5 extracts ionic strength was lower than in satura-
tion extracts, as indicated by electrolytic conductivity (Zehetner, 1997). This may cause
Adsorption Isotherms of Nickel in Acid Forest Soils 49
100 -
AEh (soil no. 8) r- - r- r-
~ r*,
c
(]) -
()
L- 50-
(])
C.
.........
c
a
. ~
c.
L-
a
C/)
'0
ro
~
:.;::::;
-50 -
0.0005 0.001 0.002 0.005 0.01 0.02 0.05 0.1 0.5
Initial Ni concentration [mmol L-
1
]
-
5 10
ro
(])
0:::
D saturation extract
* 1: 5 extract
-100- '-
Figure 2.14. Effect of applied initial Ni concentration and soil:solution ratio on relative adsorption
for soil no. 8.
lower ion competition and greater width of diffuse layers. Ni adsorption did not affect
solution pH at applied initial Ni concentrations < 1 mmol L-I, however, it resulted in a
significant decrease of pH at 10 mmol L-i, which was more strongly pronounced in
saturation extracts than in l:S extracts (Zehetner, 1997).
Relative adsorption (Arel) [percent] was calculated as
(21)
where Co and CI are the Ni concentrations in the applied solution before and after equili-
bration, respectively. For a topsoil and a subsoil horizon (soils no. 8 and 11) the influ-
ence of applied initial Ni concentration and soil:solution ratio on relative adsorption is
shown in Figures 2.14 and 2.1S.
In the low concentration range, desorption played a significant role leading to a nega-
tive value of 1\.el at the lowest applied initial Ni concentration. Relative adsorption then
increased with applied initial Ni concentration until a maximum was reached at around
O.OOS to 0.01 mmol L-I, and decreased again above 0.1 mmol L-
1
in the presented subsoil
horizon and above 1 mmol L -I in the presented topsoil horizon. This difference may be
explained by adsorption capacities and affinities higher in topsoils than in subsoils
(Zehetner, 1997). Although the amount of Ni adsorbed per mass soil was significantly
higher in l:S extracts than in saturation extracts, relative adsorption was significantly
lower at the highest applied initial Ni concentration in the topsoil horizon and at all
studied concentrations in the subsoil horizon. The similar 1\.el at applied initial Ni con-
centrations of 0.002 and 0.1 mmol L -I in the topsoil may again result from higher adsorp-
tion capacities and affinities compared to the subsoil (Zehetner, 1997).
50 Fate and Transport of Heavy Metals in the Vadose Zone
........
+-'
C
~
I-
Q)
c..
..........
c
o
:;::::;
c..
I-
o
en
"'0
ro
Q)
>
:;::::;
ro
Q)
c::
100 -
Bw (soil no. 11)
-
-
-
50 -
-
o 0.0002 n
-50 -
-
-100 -
0.0005 0.001 0.002 0.005 0.01 0.02 0.05 0.1 0.5 1
5 10
Initial Ni concentration [mmol L-
1
]
o saturation extract
* 1: 5 extract
Figure 2.15. Effect of applied initial Ni concentration and soll:solution ratio on relative adsorption
for soil no. 11.
Fractionation of Adsorbed Nickel
The h izons of a spodosal (soils no. 8-12) were used to study the influence of applied
imtla i concentration on the fractionation of adsorbed Ni (Figures 2.16 to 2.20).
e blank treatment displays the distribution of initially present Ni. In all the studied
rizons, the dominant initially present Ni fractions were, in the order of decreasing Ni
concentration, residual Ni, Ni bound to crystalline Fe-oxides, and Ni bound to amor-
phous Fe-oxides (fractions 7, 6, and 5). In the Bs, Bw, and Cw horizons (soils no. 10, 11,
and 12) these three fractions almost exclusively contributed to the total Ni contents,
which were higher than in the horizons above. In the Bhs horizon (soil no. 9), organi-
cally bound Ni (fraction 4) also significantly contributed to the total Ni content, and
notable Ni contents could be identified, in the order of decreasing Ni concentration, in
the mobile, the Mn-oxide bound, and the easily mobilizable fractions (fractions 1, 3, and
2). In the AEh horizon (soil no. 8) besides fractions 7, 6, and 5, only organically bound
and mobile Ni were identified in substantial concentrations.
According to Tu (1996), previous studies on the distribution of Ni in soils identified
50-80% and 20-30% of native Ni in the residual and in the Fe- and Mn-oxide bound
fractions, respectively, and usually less than 2% in the soluble plus exchangeable frac-
tions. Generally, this was observed in the presented study in which, however, the ratio of
residual to oxide-bound Ni was shifted toward the oxide-bound fractions.
Applied initial Ni concentrations of 0.002 and 0.1 mmol L-
1
were too small to signifi-
cantly affect total Ni contents and to identify clear trends in the fractionation of ad-
sorbed Ni, ho'Wever, especially in the uppermost horizons AEh and Bhs, the mobile Ni
fraction clearly increased with applied initial Ni concentration. At 10 mmol L -I, total Ni
contents were significantly raised in all the studied horizons, and the major part of ad-
sorbed Ni was in the mobile fraction that became the dominant fraction in the AEh and
6
~ 5
~
~
o 4
E
.s
Z 3
"0
Q)
t5 2
~
x
UJ
Adsorption Isotherms of Nickel in Acid Forest Soils 51
AEh (soil no. 8)
o 0.002 0.1 10
Initial Ni concentration [mmol L-
1
]
Fraction:
= 7
~ 6
_ 5
EEEE!I 4
= 3
= 2
= 1
Figure 2.16. Distribution of Ni among the seven sequentially extracted fractions at different
applied initial Ni concentrations for soil no. 8.
6
~ 5
~
(5 4
E
.s
z 3
"0
Q)
t5 2
~
x
UJ
Bhs (soil no. 9)
o 0.002 0.1 10
Initial Ni concentration [mmol L-
1
]
Fraction'
= 7
= 6 _ 5
EITEll 4
= 3
= 2
= 1
Figure 2.17. Distribution of Ni among the seven sequentially extracted fractions at different
applied initial Ni concentrations for soil no. 9.
Bhs horizons. The easily mobilizable fraction was also increased in all the studied hori-
zons whereas Mn-oxide bound and organically bound Ni were substantially increased
only in the AEh and Bhs horizons and to a smaller extent in the Bs horizon. An increase
in the fraction bound to amorphous Fe-oxides was identified in the AEh horizon, as well
as in the Bw and in the Cw horizons. Fractions 6 and 7 were apparently not affected by
adsorption at applied initial Ni concentrations of lO mmol L-
1
and below.
The fractionation of adsorbed Ni was conducted immediately after the adsorption
experiments, which may explain the strong dominance of mobile Ni in all the studied
horizons. Redistribution of adsorbed Ni toward more stable forms may occur with time.
52 Fate and Transport of Heavy Metals in the Vadose Zone
6
~ 5
'0)
.::<:
(5 4
E
.s
z 3
"C
(])
tl 2
~
><
w
o
Bs (soil no. 10)
o 0.002 0.1 10
Initial Ni concentration [mmol r1]
Fraction'
= 7
= 6
- 5
= 4
= 3
= 2
= 1
Figure 2.18. Distribution of Ni among the seven sequentially extracted fractions at different
applied initial Ni concentrations for soil no. 1 O.
6
5
~
(5 4
E
~
Z 3
"C
~ 2
~
x
UJ
o
Bw (soil no. 11)
o 0.002 0.1 10
Initial Ni concentration [mmol L-
1
]
Fraction:
= 7
= 6
- 5
= 4
= 3
= 2
= 1
Figure 2.19. Distribution of Ni among the seven sequentially extracted fractions at different
applied initial Ni concentrations for soil no. 11.
Adsorption Density and General Adsorption Density Isotherms
For different acid forest soils, Schulte (1988, 1994a, 1994b), as well as Schulte and
Beese (1994a, 1994b) were able to approximately describe quantity-intensity relation-
ships of various heavy metals by a single adsorption isotherm for each, by plotting the
equilibrium solution concentration (C) against the adsorption density based on the spe-
cific surface area (AD
ssA
) [ions m-
2
], defined as
AD _ (SEDTA + S) N
SSA - SSA 10
6
(22)
6
~ 5
~
o 4
E
E.
z 3
"0
Q)
13 2
~
x
w
Adsorption Isotherms of Nickel in Acid Forest Soils 53
Cw (soil no. 12)
o 0.002 0.1 10
Initial Ni concentration [mmol r1]
Fraction
= 7
= 6
- 5
= 4
= 3
= 2
= 1
Figure 2.20. Distribution of Ni among the seven sequentially extracted fractions at different
applied initial Ni concentrations for soil no. 12.
where SEDTA [mmol kg-I] is the EDT A-extractable amount of Ni initially adsorbed, S
[mmol kg-I] is the amount adsorbed during the test, N [6.02205 X 10
23
ions mol-I] is
Avogadro's number, and SSA [m
2
g-I] is the specific surface area.
Adsorption processes, especially nonspecific adsorption through ion exchange, are
influenced by the intrinsic surface charge, which CEC is a measure of. By dividing ADsSA
through the intrinsic surface charge density (CEC/SSA) [mole m-
2
], the adsorption den-
sity based on exchange sites (AD
cEC
) [ions mole-I] was obtained as
AD _ (SEDTA + S) N
CEC - CEC
(23)
where CEC [mmol
e
kg-I] is the cation exchange capacity. Since it was shown that ad-
sorption isotherm parameters were closely related to the exchange sites occupied by
mono- and divalent cations (Zehetner, 1997), CEC in Equation 23 was further substi-
tuted for CEC(2+)' obtaining AD
cE
C(2+).
Surface based and charge based adsorption densities were calculated for each point of
the isotherm of each soil, and the linear form of the van Bemmelen-Freundlich equation
(Eq. 19) was fitted by linear regression to the pairs of log ADsSA (log AD
cEC
and log
AD
cE
C(2+)' respectively) and log C, obtaining general adsorption density isotherms by
combining all the studied soils (Figures 2.21 to 2.23).
Similar fits were obtained by using SSA and CEC as reference quantities, however,
the quality of fit was improved when only the proportion of CEC occupied by mono- and
divalent cations (CEC(2+ was used as the reference. By means of the proposed general
adsorption density isotherms, quantity-intensity relationships of native Ni can be esti-
mated and the behavior of deposited Ni can be assessed in acid soils of different compo-
sition if the initially adsorbed amount (Q) or the corresponding solution concentration
(C) and either SSA, CEC, or CEC(2+) are available. Use of CEC as the reference quan-
tity, which is usually available in soil databases, yields similar accuracy as SSA, how-
54 Fate and Transport of Heavy Metals in the Vadose Zone
........
N
,
E
II)
c:
g

II)
II)
Cl

10
16
10
15
10
14
IIIPO
log ADSSA = 16.1948 + 0.5887 log C
R = 0.930***, n = 260

0.0001 0.001 0.01 0.1
C [mmol r1]
10
Figure 2.21. General adsorption density isotherm of Ni with SSA as the reference quantity.
10
23
..; 10
22
()
-0
E
II)
10
21
;=..
()
w
()
Cl 10
20
log AD
cEC
= 22.3098 + 0.5937 log C
R = 0.928***, n = 260
1 0
19
.......... r-rn-nr--r-"""""TTT1"T1
0.0001 0.001 0.01 0.1
10
C [mmol L-
1
]
Figure 2.22. General adsorption density isotherm of Ni with CEC as the reference quantity.
ever, if CEC(2+) is available, the accuracy can be strongly improved. As indicated by the
correlation of adsorption isotherm parameters with SSA, CEC, and CEC(2+) (Zehetner,
1997), the better generalizibility of adsorption density isotherms with CEC(2+) as the
reference shows that ion exchange against mono- and divalent cations was probably the
principal mechanism of Ni adsorption under the studied conditions.
SUMMARY
Nickel adsorption was studied in acid forest soils. The traditional and the "two-sur-
face" Langmuir equations as well as the van Bemmelen-Freundlich equation were fitted
to the experimental data. At low concentration, specific adsorption on organic surfaces
10
23
-
..--
,
u
0
E
10
22
IIJ
c::
g
' 10
21
o
W
()
o
10
20
0.0001
Adsorption Isotherms of Nickel in Acid Forest Soils 55
0.001
log AD
CEC
{2+) = 22.8322 + 0.5948 log C
R = 0.969***, n = 260
0.01 0.1
C [mmoll-
1
]
10
Figure 2.23. General adsorption density isotherm of Ni with CEC(2+l as the reference quantity.
may occur to a certain degree, however, exchange against mono- and divalent cations is
considered as the primary mechanism of Ni adsorption in the studied soils. Organic mat-
ter and Mn-oxides may be the most effective adsorbents in topsoils and subsoils, respec-
tively. By means of sequential extraction, adsorbed Ni was predominantly found in the
mobile fraction, involving water-soluble and exchangeable Ni, as well as easily soluble
metal-organic complexes. In order to obtain close-to-field conditions, adsorption experi-
ments were conducted in saturation extracts. Comparison to adsorption at 1:5 showed
that, especially at higher concentrations, adsorption would be strongly overestimated if
tighter soil:solution ratios were applied. Using the van Bemmelen-Freundlich equation,
general adsorption density isotherms were developed for the studied soils. Similar fits
were obtained by using specific surface area (SSA) and cation exchange capacity (CEC)
as reference quantities, however, the quality of fit was improved when only the propor-
tion of CEC occupied by mono- and divalent cations (CEC(2+) was used as the reference.
REFERENCES
Adriano, D.C. Trace Element.J in the TerrutriaL Environment. Springer-Verlag, New York, 1986.
Basta, N.T. and M.A. Tabatabai. Effect of cropping systems on adsorption of metals by soils: II.
Effect of pH. Soil Sci., 153, pp. 195-204, 1992.
Blum, W.E.H., H. Spiegel, and W.W. Wenzel. Boden.wAand.Jinventur. Konzeptwn, Durchfuhrung und
Bewertung. Empfehlungenmr VereinheitLichung der Vorgang.JweiJe in (j.Jterreich. 2. uberarbeitete Auflage,
Bundesministerium fur Land- und Forstwirtschaft, Bundesministerium fur Wissenschaft,
Verkehr und Kunst, Wien, 1996.
Boyd, G.E., J. Schubert, and A.W. Adamson. The exchange adsorption of ions from aqueous
solutions by organic zeolites. 1. Ion exchange equilibria. J. Am. Chern. Soc., 69, pp. 2818-2829,
1947.
Brandstetter, A., R.S. Sletten, A. Mentler, and W.W. Wenzel. Estimating dissolved organic
carbon in natural waters by UVabsorbance (254 nm). Zeit.Jchrijt fur Pjlanzenerniihrung und
BOden!cunde, 159, pp. 605-607, 1996.
56 Fate and Transport of Heavy Metals in the Vadose Zone
Brummer, G.W., J. Gerth, and U. Herms. Heavy metal species, mobility and availability in soils.
Zeitdchriftfur Pflanunerniihrung und Bodenkunde, 149, pp. 382-398, 1986.
Brunauer, S., L.E. Copeland, and D.L. Kantro. The Langmuir and BET Theories. In The SoLid-
Gad Interface, Volume 1, pp. 77-103, E.A. Flood, Ed., Marcel Dekker, New York, 1967.
Chapman, D.L. A Contribution to the Theory of Electrocapillarity. The London, Edinburgh, and
DubLin PhilfJdophicaL Magazine and JournaL of Science, 6
th
series, 25, pp. 475-481, 1913.
Everett, D.H. ManuaL of Sym6014 and Terminology for PbydicochemicaL QpantitieJ and UnitJ. Appendix II:
Dejinilwnd, Terminology and Sym6014 in CoLLoid and Swface ChemiJtry. Butterworths, London, 1972.
Farrar, D.M. The use of vapour-pressure and moisture-content measurements to deduce the
internal and external surface area of soil particles. J. SoiL Sci., 14, pp. 303-321, 1963.
Fendorf, S.E., G.M. Lambie, M.G. Stapleton, M.J. Kelley, and D.L. Sparks. Mechanisms of
chromium(III) sorption on silica: 1. Cr(lll) surface structure derived by extended X-ray
absorption fine structure spectroscopy. Environ. Sci. TechnoL., 28, pp. 284-289, 1994.
Forrester, S.D. and C.H. Giles. From manure heaps to monolayers. One hundred years of solute-
solvent adsorption isotherm studies. Chem. Ind., pp. 318-325, 1972.
Freundlich, H. KapiLlarchemie. Akademische Verlagsgesellschaft, Leipzig, 1909.
Giles, C.H., T.H. MacEwan, S.N. Nakhwa, and D. Smith. Studies on adsorption. Part XI: A
system of classification of solution adsorption isotherms, and its use in diagnosis of adsorption
mechanisms and in measurement of specific surface areas of solids. J. Chem. Soc., London, pp.
3973-3993, 1960.
Giles, C.H., D. Smith, and A. Huitson. A general treatment and classification of the solute
adsorption isotherm. I: Theoretical. J. CoLloid Interface Sci., 47, pp. 755-765, 1974a.
Giles, C.H., A.P. D'Silva, and LA. Easton. A general treatment and classification of the solute
adsorption isotherm. Part II: Experimental interpretation. J. CoLloid Interface Sci., 47, pp. 766-
778, 1974b.
Gouy, M. Sur la constitution de la charge electrique a la surface d'un electrolyte. JournaL de
PhYdi<JUC Theori<Juc et AppLi<Juee, 4" serie, 9, pp. 457-468, 1910.
Griffin, RA. and A.K. Au. Lead adsorption by montmorillonite using a competitive Langmuir
equation. Soil Sci. Soc. Am. J., 41, pp. 880-882, 1977.
Halsey, G. and H.S. Taylor. The adsorption of hydrogen on tungsten powders. J. Chem. PhYd., 15,
pp. 624-630, 1947.
Harter, RD. Curve-fit errors in Langmuir adsorption maxima. SoiL Sci. Soc. Am. J., 48, pp. 749-
752, 1984.
Harter, R.D. and D.E. Baker. Applications and misapplications of the Langmuir equation to soil
adsorption phenomena. SoiL Sci. Soc. Am. J., 41, pp. 1077-1080, 1977.
Harter, RD. and D.E. Baker. Further reflections on the use of the Langmuir equation in soils
research. SoiL Sci. Soc. Am. J., 42, pp. 987-988, 1978.
Harter, RD. and G. Smith. Langmuir Equation and Alternate Methods of Studying "Adsorp-
tion" Reactions in Soils. In ChemiJtry in the Soil Environment, ASA Special Publication Number
40, pp. 167-182, R.H. Dowdy et aI., Eds., American Society of Agronomy, Soil Science
Society of America, Madison, WI, 1981.
Hemwall, J.B. The fixation of phosphorus by soils. Adv. Agron., 9, pp. 95-112, 1957.
HilleL 0. Fundamenta14 of SoiL Pbyd0. Academic Press, London, 1980.
Holford, I.C.R. Soil adsorption phenomena and the Langmuir equation. SoiL Sci. Soc. Am. J., 42,
pp. 986-987, 1978.
Holford, I.C.R., RW.M. Wedderburn, and G.E.G. Mattingly. A Langmuir two-surface equation
as a model for phosphate adsorption by soils. J. SoiL Sci., 25, pp. 242-255, 1974.
Konig, N., P. Baccini, and B. Ulrich. Der EinfluE der naturlichen organischen Substanzen auf
die Metallverteilung zwischen Boden und Bodenlosung in einem sauren Waldboden. Zeitdchrift
fur Pflanunerniihrung und Bodenkunde, 149, pp. 68-82, 1986.
Adsorption Isotherms of Nickel in Acid Forest Soils 57
Langmuir, I. The adsorption of gases on plane surfaces of glass, mica, and platinum. J. Am. Chem.
Soc., 40, pp. 1361-1403, 1918.
Lindsay, W.L. ChemicaL equilibria in Soiu. John Wiley & Sons, New York, 1979.
Ma, Q. Y. and W.L. Lindsay. Estimation of Cd
2
+ and Ni
2
+ activities in soils by chelation. Geoderma,
68, pp. 123-133, 1995.
McBride, M.B. Reactions Controlling Heavy Metal Solubility in Soils. In AdvanCed in Soil Science,
Volume 10, B.A. Stewart, Ed., Springer-Verlag, New York, 1989, pp. 1-56.
Mehra, a.p. and M.L. Jackson. Iron oxide removal from soils and clays by a dithionite:citrate
system buffered with sodium bicarbonate. ClayJ and Clay Minerau, 7, pp. 317-327, 1960.
Meiwes, K-J., N. Konig, P.K Khanna, J. Prenzel, and B. Ulrich. Chemische Untersuchungsver-
fahren fur Mineralboden, Auflagehumus und Wurzeln zur Charakterisierung und Bewertung
der Versauerung in Waldboden. In Berichte deJ ForJchungJzentrumJ Wa[Jo"lcOJYJtemeIWauJJterben,
7, Gottingen, 1984, pp. 1-67.
a'Day, P.A., G.E. Brown, Jr., and G.A. Parks. X-ray absorption spectroscopy of cobalt(II)
multinuclear surface complexes and surface precipitates on kaolinite. J. CoLlolJ Inter/ace Sci,
165, pp. 269-289, 1994a.
a'Day, P.A., G.A. Parks, and G.E. Brown, Jr. Molecular structure and binding sites of cobalt(II)
surface complexes on kaolinite from X-ray absorption spectroscopy. ClayJ and Clay Minerau,
42, pp. 337-355, 1994b.
Ponizovskiy, A.A., L.P. Korsunskaya, T.A. Polubesov, a.A. Salimgareyeva, Ye.S. Aleksane, and
Ya.A. Pachepskiy. Methods for determining the specific surface area of soil from water-vapor
adsorption. EuraJian SoiL Sci, 25, pp. 12-29, 1993.
Posner, A.M. and J.W. Bowden. Adsorption isotherms: Should they be split? J. SoiL Sci., 31, pp.
1-10, 1980.
Puri, B.R. and K Murari. Studies in surface area measurements of soils: 2. Surface area from a
single point on the water isotherm. SoiL Sci., 97, pp. 341-343, 1964.
Russell, E.W. RMJeLL'.J SoiL ConditionJ and Plant Growth. 11th ed., A. Wild, Ed., Longman Scientific
& Technical, Essex, 1988.
Sadiq, M. and T.H. Zaidi. The adsorption characteristics of soils and removal of cadmium and
nickel from wastewaters. Water Air Soil PolLut. 16, pp. 293-299, 1981.
Sadiq, M. and C.G. Enfield. Solid phase formation and solution chemistry of nickel in soils: l.
Theoretical. Soil Sci, 138, pp. 262-270, 1984a.
Sadiq, M. and C.G. Enfield. Solid phase formation and solution chemistry of nickel in soils: 2.
Experimental. SoiL Sci, 138, pp. 335-340, 1984b.
Scheidegger, A.M. and D.L. Sparks. A critical assessment of sorption-desorption mechanisms at
the soil mineral/water interface. SoiL Sci, 161, pp. 813-831, 1996a.
Scheidegger, A.M. and D.L. Sparks. Kinetics of the formation and the dissolution of nickel
surface precipitates on pyrophyllite. Chem. GeoL., 132, pp. 157-164, 1996b.
Scheidegger, A.M., G.M. Lambie, and D.L. Sparks. Investigation ofNi sorption on pyrophyllite:
An XAFS study. Environ. Sci. Techno!., 30, pp. 548-554, 1996a.
Scheidegger, A.M., M. Fendorf, and D.L. Sparks. Mechanisms of nickel sorption on pyrophyl-
lite: Macroscopic and microscopic approaches. SoiL Sci. Soc. Am. J., 60, pp. 1763-1772, 1996b.
Schulte, A. AdJorption von SchwermetaLlen in repriMentativen Boden IJraeu und NordwedtdeutJchlandJ in
Abhangiglceit von derJpezifiJchen Ober/liiche. Berichte des Forschungszentrums Wald6kosysteme,
Reihe A, 46, Gottingen, 1988.
Schulte, A. Adsorption density, mobility and limit values of Cd, Zn, Cu and Pb in acid forest soils.
Archil' fur Aclcer- und Pjlanzenbau und Bodenlcunde, 38, pp. 71-81, 1994a.
Schulte, A. The Adsorption of Cd, Zn, Cu and Pb in Acid Forest Soils. In BiogeochemiJtry of Trace
ElementJ, D.C. Adriano et aI., Eds., Science and Technology Letters, Northwood, 1994b, pp.
525-535.
58 Fate and Transport of Heavy Metals in the Vadose Zone
Schulte, A. and F. Beese. Adsorptionsdichte-Isothermen von Schwermetallen und ihre okologische
Bedeutung. Zeitdchrijt fur Pflanzenerniihrung und BOdenkunde, 157, pp. 295-303, 1994a.
Schulte, A. and F. Beese. Isotherms of cadmium sorption density. J. Environ. QuaL., 23, pp. 712-
718, 1994b.
Schulte, A. and J. Gehrmann. Entwicklung der Niederschlags-Deposition von Schwermetallen
in Westdeutschland. 2. Arsen, Chrom, Kobalt und Nickel. Zeitdchrijt fur Pjlanzenerniihrung und
Bodenkunde, 159, pp. 385-389, 1996.
Schulte, A., A. Balazs, J. Block, and J. Gehrmann. Entwicklung der Niederschlags-Deposition
von Schwermetallen in Westdeutschland. 1. Blei und Cadmium. Zeitdchrijt fur Pjlanzenerniihrung
und Bodenkunde, 159, pp. 377-383, 1996.
Slejko, F.L. Addorption Technology. Marcel Dekker, New York, 1985.
Sparks, D.L. EnvironmentaL Soil ChemiJlry. Academic Press, San Diego, 1995.
Sposito, G. On the use of the Langmuir equation in the interpretation of" adsorption" phenom-
ena: II. The "two-surface" Langmuir equation. Soil Sci. Soc. Am. J., 46, pp. 1147-1152, 1982.
Sposito, G. The Surface Chemi.Jtry of SoiU. Oxford University Press, New York, 1984.
Sposito, G. The Chemi.Jtry of Soiu. Oxford University Press, New York, 1989.
Stern, O. Zur Theorie der elektrolytischen Doppelschicht. Zeitdchrijt fur Elektrochemie und Angewandte
PhYdikaLi.Jche Chemie, 30, pp. 508-516, 1924.
Travis, C.C. and E.L. Etnier. A survey of sorption relationships for reactive solutes in soil. J.
Environ. QuaL., 10, pp. 8-17, 1981.
Tu, C. Distribution and transformation of native and added Ni fractions in purple soils from
Sichuan Province. Pedodphere, 6, pp. 183-192, 1996.
U ren, N. C. Forms, reactions, and availability of nickel in soils. Adv. Agron., 48, pp. 141-203, 1992.
van Bemmelen, J .M. Die Absorptionsverbindungen und das Absorptionsvermogen der Ackererde.
Die Landwirtdcha/tLichen Verdllchd-Stationen, 35, pp. 69-136, 1888.
Veith, J.A. and G. Sposito. On the use of the Langmuir equation in the interpretation of
"adsorption" phenomena. SoiL Sci. Soc. Am. J., 41, pp. 697-702, 1977.
Zehetner, F. Addorption IdothermJ of NickeL in Acid Fore.Jt Soiu. Diplomarbeit, Institut fur
Bodenforschung, Universitat fur Bodenkultur, Wien, 1997.
Zeien, H. and G.W. Brummer. Chemische Extraktionen zur Bestimmung von Schwermetall-
bindungsformen in Boden. Mitteilungen der Deutdchen BOdenkundLichen Ge.JeLucha/t, 59, pp. 505-
509, 1989.
CHAPTER ;)
Sorption-Desorption Equilibria and
Dynamics of Cadmium During
Transport in Soil
R.S. Kookana, R. Naidu, D.A. Barry, Y.T. Tran, and K. Bajracharya
INTRODUCTION
Cadmium (Cd) is a soil contaminant that is toxic to plants and animals. It is a major
human health hazard due to its potential accumulation in kidneys leading to kidney dys-
function (Alloway, 1990). Both natural and anthropogenic activities are important sources
of Cd to the soil environment. Sources include atmospheric fallout from metallurgical
industries, application of sewage sludge or other waste containing Cd on land, and use of
phosphatic fertilizers in agricultural areas. The result is substantial Cd accumulation in
soil, providing the source for plant uptake, as well as leaching of Cd under certain soil
and environmental conditions (Alloway, 1990; Naidu et al., 1997).
The mean concentrations of Cd in agricultural soil are generally less than 0.4 mg/kg
(Alloway, 1990), with a maximum up to 3 mg/kg. Highly contaminated soils can, how-
ever, contain Cd concentrations up to 10 mg/kg or even more (e.g., Hornburg and
Brummer, 1993). Elevated concentrations of Cd in soil can also originate from geologic
sources. For example, soils developed from black shales can contain Cd concentrations>
20 mg/kg (Alloway, 1990). In soil solutions, however, Cd concentrations rarely exceed
10 /J-g/L under agricultural conditions. McLaughlin et al. (1997) reported that in solu-
tions extracted at moisture contents equivalent to -5 kPa from 50 Australian agricultural
(saline/sodic) soils, the total concentration of Cd ranged from 2.7 to 222.4 nM (0.30-
24.9 /J-g/L). Natural Cd concentrations up to 800 /J-g/L (far exceeding USEPA drinking
water quality standard -10 /J-g/L) have been observed in filtered samples of stream wa-
ters and shallow groundwaters associated with unmined ore deposits in some areas
(Runnells et aI., 1992). Among anthropogenic activities, the disposal of sewage sludge
and/or metal rich industrial wastes on soils can lead to unacceptably high concentrations
of Cd in both soils and associated water resources.
59
60 Fate and Transport of Heavy Metals in the Vadose Zone
The fate of Cd from anthropogenic sources is largely determined by its retention and
mobility in soils (Naidu et al., 1997). The chemodynamics of Cd in the soil environment
are influenced by its interactions with both the solid and aqueous phases of soil. In addi-
tion, the soil solution and solid phase concentrations of Cd can have significant bearing
on the nature of the interactions. Leaching of Cd through soil profiles has implications
for both Cd buildup in the root zone and the potential for contamination of surface and
groundwater associated with soils showing low sorption capacity. Some soils can have
an inherently low sorption capacity for Cd due to either their low surface negative charge
density or their sandy nature and high permeability. Cd mobility in such soils can be
rather high (Boekhold and Van der Zee, 1991; Kookana et aI., 1994). Groundwater
monitoring studies (Lieber et aI., 1964) as well as simulations of long-term leaching
behavior of Cd (Boekhold and Van der Zee, 1991) indicate that Cd leaching in some
instances, e.g., in sandy soils receiving high input (> 50 glha.yr), can potentially exceed
the acceptable levels in groundwater.
The objective of this report is to provide an overview of important processes which
largely determine the fate of Cd in soil environment. The processes covered here include
sorption, desorption, precipitation, complexation - all factors affecting transport or mo-
bility. The influences of soil solid as well as the solution phase compositions, and of the
soil physical conditions, e.g., static and flow conditions, on sorption, desorption, and
mobility of Cd in soils are discussed.
PROCESSES GOVERNING FATE OF CADMIUM IN THE SOil PROFilE
The processes of Cd retention in a soil include adsorption/desorption, ion exchange,
and precipitation/dissolution reactions, their occurrence depending on the nature of the
solid phase, the concentration of Cd present in soil solution, and soil solution composi-
tion. Cd mobility and transport through the soil profile are directly linked to retention
and release processes. Precipitation and dissolution reactions of metals including Cd
may involve discrete solid phases or solid phases adsorbed to soil component surfaces
and usually occur either in the presence of specific anions or at high concentration of
metal ions (Naidu et aI., 1997). While ion exchange causes an exchange between ionic
species in solution phase and those held by the soil surface, adsorption reactions can
involve ionic as well as molecular species of metals (Amacher et al., 1986). Since it is
often not possible to make a clear distinction between various retention processes of
metals in soils, the term sorption is more appropriate than adsorption. These key pro-
cesses (sorption, desorption, precipitation, complexation, and transport) regulate the
mobility and availability of Cd in soil solution (Naidu et al., 1997), and are discussed in
the following sections, particularly in relation to the soil and solution phase composition.
Sorption
Sorption is one of the most important processes that govern the bioavailability and
mobility of Cd in soil. Sorption of metals has generally been related to metal ion hydroly-
sis (Hodgson et aI., 1964; Forbes et al., 1976) and therefore to their pKa values. On the
basis of pKa values (in parentheses), metal sorption is expected to decrease in the follow-
ing sequence: Hg (3.4) Pd (7.7) > eu (7.7) Zn (9.0) > Co (9.7) > Ni (9.9) > Cd
(l0.1). Based on studies carried out on soil clay isolates from a range of soils, Tiller et al.
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 61
(l984a) concluded that both specific and nonspecific sorption mechanisms contribute to
Cd retention in soils. Cd can form inner-sphere complexes with amorphous Fe and Ai
oxyhydroxides (Hayes and Leckie, 1987). Silicate minerals exhibit a range of affinities
for Cd (Garcia-Miragaya and Page, 1978; Tiller et al., 1979) and the nature of sorption
mechanisms also varies with the type of mineral. As Zachara and Smith (1994) stated,
Cd can form at least two types of complexes with layer silicates, including outer-sphere
ion exchange complexes on the basal plane, and coordination complexes with SiOH or
AlOH groups exposed at the crystalline edges. However, they found that on soil smectite
edge complexation reactions were more important. Sorption of Cd on kaolinite is mainly
ascribed to ion exchange and possibly some inner-sphere complexation at pH > 6.5
(Schindler et al., 1987). While Cd sorption through metal ion hydrolysis and specific
adsorption in the presence of different electrolytes has been documented (e.g., Tiller et
al., 1979; Brummer et al., 1988; Tiller 1989), the nature of mechanisms operative in the
presence of specifically adsorbed ligand ions such as sulfate, phosphate, and organic
ligands remains unclear (Naidu et al., 1997).
Sorption of Cd on amorphous metallic oxides has been postulated to be a continuum
between surface reactions and precipitation (Farley et al., 1985). Farley et al. (1985)
extended the well-established surface complexation approach (coordination reactions
between solute and functional groups on the surface) through a "surface precipitation"
model which is based on a series of adsorption followed by precipitation reactions on the
oxide surfaces. The precipitation on the solid phase was described by the formation of a
solid-solution. The model satisfactorily described the sorption behavior of Cd on iron
hydroxide surfaces (Farley et al., 1985; Dzombak and Morel, 1986).
Silicate clay minerals, amorphous oxides of Fe and Mn, and particulate organic C
present in soil all influence the sorption potential of metals in soils. However, the relative
contribution of these soil fractions to Cd sorption varies depending on soil type and its
solution composition. For example, in the presence of Ca concentrations high enough to
suppress ion exchange mechanisms, layer silicates have a relatively lesser role than or-
ganic carbon and Fe and Mn oxides, whereas layer silicates can contribute significantly
to Cd sorption on Na-saturated clay-sized isolates from soils (e.g., Zachara et al., 1992).
Obviously the contribution will vary with the type of layer silicates. Often, layer silicates
act as substrates on which amorphous Fe and Al oxides precipitate. Therefore, fixed-
charge sites on soil clays with a potential to bind Cd can be blocked by oxides (Zachara
et al., 1992). The presence of mineral-bound organic material can enhance Cd sorption
(Haas and Horowitz, 1986) or have no effect (Davis, 1984; Harter and Naidu, 1995).
Not only the nature of the sorbent but also a range of other soil properties and factors
such as soil pH, ionic strength of the soil solution, the presence of inorganic and organic
ligands, competing ions, etc., affect Cd sorption. The influence of some of these factors is
discussed in the following sections.
Factors Affecting Cd Sorption il1 Soils
Cd interactions with the soil solid phase include ion adsorption at surface sites, ion
exchange with clay minerals, binding by organically coated particulate matter or organic
colloidal material, or adsorption of metal ligand complexes (Naidu et al., 1997). All of
these interactions are influenced by soil solution composition and characteristics such as
62 Fate and Transport of Heavy Metals in the Vadose Zone
pH and ionic strength, nature of the metal species, dominant cation, and ligands (inor-
ganic and organic) present in the soil solution.
pH
A large body of literature based on both pure mineral systems and soils has estab-
lished that soil solution pH has the most critical influence on sorption of Cd (e.g., Tiller
et al., 1984b; Haas and Horowitz, 1986; Briimmer et al., 1988; Biirgisser et al., 1991;
Gerth et al., 1993; Naidu et al., 1994a; Naidu et al., 1997). These studies show that over
a narrow pH range, the sorption of Cd increases very rapidly leading to a so-called
adsorption edge (Figure 3.1). Recently, Tran et al. (1998a) performed Cd sorption ex-
periments at a range of pH values in a sandy soil. They found that for every increase of
0.5 unit of pH for the range of 5.5 to 6.5, twice as much sorption of Cd was observed.
Such increases in sorption presumably occur because of the rapid increase in the con-
centration of the metal-hydroxy species, believed to be the active component adsorbed
by soils (Hodgson et al., 1964; Davis and Leckie, 1978; Tiller et al., 1979). However,
speciation calculations show that the concentration of CdOH+ is negligible relative to
Cd
2
+ at the pH of the adsorption edge (Naidu et al., 1994a). Therefore, for the metal-
hydroxy species to be adsorbed, it must have a very high affinity for the soil surface
which will then drive the metal hydrolysis reaction (Eqs. 1 and 2) to the right, maintain-
ing Le Chatelier's principle of equilibrium (Naidu et al., 1994a):
(1)
MOH+ + Soil :::} Soil - MOH (2)
On surfaces exhibiting variable charge, in addition to the metal species, the surface charge
density of the adsorbent is strongly influenced by pH. Increasing solution pH leads to a
rapid increase in net negative surface charge which may explain the enhanced affinity
for metal ions in such systems, e.g., Fe and Al oxides, organic matter (Garcia-Miragaya
and Page, 1978; Naidu et al., 1997).
Ionic Strength
Soil properties such as pH, charge density distribution, thickness of the diffuse double
layer, and the activity of Cd present in solution are all influenced by the composition of
soil solution (Harter and Naidu, 1995). Several studies (e.g., Homann and Zasoski, 1987;
Boekhold et al., 1993; Zachara et al., 1993; Naidu et al., 1994a,b) have demonstrated the
importance of soil solution composition on the nature and extent of Cd sorption by soils
and their constituents.
Increasing ionic strength (1) generally reduces metal sorption due to its influence on
both sorbate and sorbent properties (Naidu et al., 1997). Depending on the nature of the
surface properties of the sorbent, the effect of ionic strength on the sorption of Cd can be
marked. On minerals with permanent charge density, a substantial reduction in Cd sorp-
tion with increasing ionic strength of NaCI0
4
, NaCl, or Na2S04 was noted by several
workers (e.g., Garcia-Mirgaya and Page, 1976, 1977; Zhu and Alva, 1993). However,
when Cd forms inner-sphere complexes with sorbents (e.g., with amorphous Fe and Al
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 63
- - - ~ , - ' " " ' - - - - - - " " ' . , , - - - - ~ - ---- --- --,----""'" "--------
Xeralf
100 100
-
~
~
"C
Q)
.0
'-
0
en
"C
U
-
0
c::
0
t
0
61lmal L'10
Cl.
1.51lmal L'l ... 0 151lmal L'l ...
'-
a.. 20 30 Ilmal L"
1
I:J. 3 Ilmal r
1
I:J.
150llmal L'l 61lmal L'l
0 0
2 4 6 8 2 4 6 8
Oxisol Oxisol
-
0
c::
0
t Maianda
0
0.75 1lmal L'l ... Cl.
0
20
1.5llmal L'l I:J.
'-
a..
31lmol C
1

61lmal L'10
0
2 4 6 4 6 8
Figure 3.1. Effect of pH on sorption of Cd in four soils at different initial concentrations (Naidu et
aI., 1994, with permission).
oxyhydroxide, smectite), sorption is little affected by ionic strength (Hayes and Leckie,
1987). For example, Naidu et al. (1994a) observed that doubling of I did not influence
Cd sorption in a smectite-dominated Vertisol, whereas this increase in I caused a 50-fold
drop in Cd sorption on an Inceptisol (Figure 3.2), On variable charge surfaces, the effect
of ionic strength depends on pH, In an Oxisol, Cd sorption was found to increase with
ionic strength at pH values below the point of zero net charge (PZNC) but the reverse
was observed at pH values above the PZNC (Naidu et al., 1994a). The effect of ionic
strength on metal sorption via its effect on electrostatic potential in the plane of adsorp-
tion is likely to be most marked in variable charge soils (Barrow, 1987).
Complexation with Ligands
Complexation of heavy metal cations with a variety of ligands in soils (both inorganic
and organic) has been long been recognized (Harter and Naidu, 1995), Indeed, several
64 Fate and Transport of Heavy Metals in the Vadose Zone
4.4
.. Xeralf
4.0 4.0

'0
en
b 3.6 3.6
H20 0.01 0.030.150.300.75 1.5 H20 0.01 0.030.150.300.75 1.5
(5
E
.6 Oxisol
-
c::
5.04
E
<t:
.03
.01

H20 0.01 0.030.150.300.75 1.5 H20 0.010.030.150.300.75 1.5
Ionic strength NaN0
3
(mol L-1)
Figure 3.2. Effect of ionic strength on Cd sorption in six soils (Naidu et aI., 1994, with permission).
computer models including GEOCHEM-PC (Parker et al., 1995) and MINTEQA2/
PRODEFA2 (Allison and Brown, 1995), which are recent improvements on previous
versions of these models, are available to calculate various ionic speciations of the metals
in soil solution. Model simulation as well as experimental results show that complex
formation can significantly affect the activity of Cd in soil solution and therefore its
sorption and leaching potential through a soil profile (Garcia-Miragaya and Page, 1976;
Naidu et al., 1994b; Temminghoff et al., 1995: Bolton et al., 1996). Cd+
2
makes stable
complexes with CI-, SO/-; principally CdCI+ and CdS0
4
, respectively. However, model
calculations show that at the same concentration of Cd
2
+, Cl-, and SO/-, the fraction
complexed with CI- is higher than that with sol-. Chlorocomplexes of Cd can signifi-
cantly reduce the sorption of Cd in soils, depending on chloride concentration in soil
solution (Boekhold et al., 1993).
Formation of metal-ligand complexes can have a variable impact on Cd sorption. It
has been observed, for example by Homann and Zasoski (1987), that the presence ofel-
and S04
2
- had either no effect or reduced the Cd sorption on some forest soils. In calcar-
eous soils with pH > 7.5 (Typic Torrifluvents, Ustollic Caciorthids, and Petro calcic
Paleustolls), O'Connor et al. (1984) observed that Cd sorption decreased in the pres-
ence of CaCl
2
but increased in the presence of CaS04' the latter being the result of a
decreased Ca activity and reduced competition with Cd for sorption. Cd-ligand com-
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 65
plexes can have important consequences in terms of Cd bioavailability and food quality.
Studies on the effect of chloride salinity on Cd bioavailability in soil solution and plant
uptake (Bingham et aI., 1986; McLaughlin et aI., 1994) show that Cd availability is
enhanced in the presence of CI ions in soil solution.
Organic ligands through their complexation with Cd can either enhance or reduce Cd
sorption in soil, depending on the functional groups of the organic molecules and the
charge characteristics of the soil (Harter and Naidu, 1995). For example, Elliott and
Denneny (1982) studied sorption of Cd in the presence of acetate, oxalate, nitrilotriacetate,
and EDT A in three soils. They observed that in two of the three soils, oxalate increased
the sorption of Cd at pH values below 5.0, whereas the other ligands caused a reduction
in sorption. Neal and Sposito (1986) observed that, in soils with permanent charge, Cd
sorption was reduced by the presence of organic matter from sewage sludge. Bolton et
al. (1996) studied the Cd complexation properties of a humic acid fraction extracted
from a soil and concluded that it forms relatively strong complexes with soil humic acid
at pH values relevant to natural environments. Low molecular weight organic acids such
as acetic, citric, fumaric, oxalic, and succinic acids have been shown to enhance the
release behavior of Cd present in soil and, therefore, its bioavailability (Krishnamurti et
al., 1997). Complexation of organic and inorganic ligands can result in enhanced trans-
port of Cd through the soil profile, as discussed in a later section.
Presence Of Other Metals and Cations and Competition for Sorption Sites
Competition by other metal ions (Cu, Ni, Pb, Cr, etc.) has been shown to cause a
reduction in Cd sorption by several workers (e.g., Garcia-Mirgaya and Page, 1976;
Homann and Zasoski, 1987; Christensen, 1987). Christensen (1987), while studying Cd
sorption on 12 Danish soils, found that in the presence of mixtures of heavy metals such
as Ni, Co, and Zn, the sorption coefficient for Cd was up to 14 times lower than those
observed for Cd alone. The reduction was ascribed mostly to Zn due to its relatively
higher concentration than other competing metals. Other cations such as Ca, Mg, and Na
in soil solution can also compete with Cd for sorption sites. However, Ca has been re-
ported to be a much stronger competitor than Na for Cd sorption on soils and pure sys-
tems (Christensen, 1984a; Zhu and Alva, 1993; Zachara et aI., 1993; Boekhold et aI.,
1993; Naidu et al., 1994a; Kookana and Naidu, 1998). Boekhold et aI., (1993) reported
that due to competition between Cd and Ca, sorption of Cd in soil was reduced by 80% in
the Ca-electrolytes as compared to the Na-electrolytes. Similarly, Naidu et aI. (1994a)
found that, in Australian soils, even when the ionic strengths were kept constant, Ca
2
+
caused much greater reduction in Cd sorption than Na+. Such reduction in sorption can
have a major influence on Cd mobility and transport behavior in soils, as discussed below.
Effect Of pH and Soil Solution Composition on Cd Transport
As mentioned already, soil pH, through its effect on both surface charge density and
the formation of hydroxy metal species, exerts a strong influence on Cd sorption and
hence transport in soils. The influence of pH may, however, vary with soil type. In soils
with high surface charge densities, limited movement of heavy metals has been observed
at pH values above 6.0. For example, in a Typic Hapludoll from Minnesota (pH = 6.4),
66 Fate and Transport of Heavy Metals in the Vadose Zone
Cd did not leach beyond 0.4 m depth over a period of three years even after the incorpo-
ration of sludge in the top 0.2 m at a loading of 25 kg Cd ha -1 (Dowdy and Yolk, 1983).
In contrast, these authors noted that in a Typic Hapludult at pH 5, despite the high OM
content (10.6%) in the topsoil, 4-7% of the applied Cd leached beyond a depth of 1.2 m.
On the basis of a field study, Streck and Richter (1997a) reported that following 29 yr of
wastewater (Cd concentration ranging from 2-40 /lg/L during 1980-1990) application,
only about 5% of Cd and Zn were found below 0.7 and 0.9 m depths, respectively, in a
soil (current pH = 5.2-5.4 in 1:2.5, 0.01 M CaCI
2
). In flow-through experiments,
Bajracharya et al. (1996) found that sorption was reduced drastically when the pH is
reduced from 6 to 4.3. This resulted in an early breakthrough of Cd and a reduction in
the value of the partition or Freundlich coefficients by an order of magnitude, consistent
with Al-Soufi (1994).
A reduction in Cd sorption due to increase in ionic strength or the presence of com-
peting species is expected to influence its transport through soils. Recently, Kookana
and Naidu (1998) studied the Cd transport behavior in laboratory columns of an Oxisol
and an Alfisol, as influenced by varying ionic strengths and index cations. They ob-
served that when the concentration of NaN0
3
was increased in the background solu-
tions from 0.03 M to 0.30 M, the breakthrough of Cd through the Oxisol soil column
occurred 3-4 times faster (Figure 3.3A). While for breakthrough only approximately 30
pore volumes were needed at 0.30 M NaN0
3
, about 110 pore volumes were needed in
the presence of 0.03 M NaN0
3
At the same ionic strength, Cd eluted much earlier in the
presence of 0.01 M Ca(N0
3
)2 than in the presence of 0.03 M NaN0
3
An increase in
Ca(N0
3
)2 concentration, however, had relatively less impact than NaN0
3
, mainly due
to the significant competition for sorption sites in the presence of even relatively low
(0.01 M) concentration of Ca(N0
3
h. In contrast, even at a very high ionic strength of
NaN0
3
(0.075 M), the Cd breakthrough in the Alfisol took more than 100 pore vol-
umes, whereas only 10 pore volumes were needed in the presence of 0.05 M Ca[N0
3
h
(Figure 3.3B). An increase in ionic strength of Ca[N03h from 0.05 M to 0.25 M en-
hanced the Cd transport by a factor of 2 in the Alfisol. For both soils, Cd movement at
constant ionic strength was an order of magnitude faster in the presence of Ca(N0
3
)2
solution compared to NaN0
3

The presence of inorganic and organic ligands in soil solution can markedly affect the
mobility of Cd through soil. Chloride forms stable complexes with Cd and given the
natural abundance of CI- in soils, it is particularly significant to consider the role of CI- in
Cd mobility. Doner (1978) studied the mobility of Cd, Cu, and Ni through a Typic
Xerorthent with pHsat of 6.6, demonstrating that in the presence of 0.5 M NaCI solution,
Cd moved up to 4 times more rapidly than in the presence of NaCI0
4
- solutions. It is
worth noting that very high Cd concentrations (10 IlglmL) were used in this study. The
role of chlorocomplexes in enhancement of Cd mobility was similarly observed by
Kookana et al. (unpublished data) during miscible displacement studies on Oxisol soil
columns. These experiments were carried out under conditions similar to those in Kookana
et al. (1994) but at a higher effluent pH. They found that the Cd BTC (breakthrough
curve) in the presence of O.OlM CaCl
2
was twice as fast as in O.OlM Ca [N0
3
]2 (Figure
3.4). This was because Cd forms complexes with CI- but not Nol-. This is consistent
with calculations which show that at I = 0.03, in 0.02 M NaCI solutions, only 48% of
dissolved Cd is present as Cd
2
+ (Boekhold et aI., 1993).
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 67
0

0
0
-0
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
A
./
./...,.,.,:
Oxisol
,
;
,
I

1-
-
I
/
0.01 M Ca[N0312
-
I
[J 0.03 M NaN03

.0.3 M NaN03
0.1
....
o
50
1 ,..., ----_._---
B Alfisol
0.9 I
0.8

0.7
0.6

0.5
0.4
0.3
0.2

0.1
0
0
o
o
o
o
o
o
20 40
00.75 M NaN0
3
0.25 M Ca(N0
3
)2
00.05 M Ca(N0
3
)2
60 80
100 150

I
100
0 I
B I
o
o
o
[J
o
[J
o
a
[J
120
I
140 160
Pore Volumes
Figure 3.3. Ionic strength and index cation effect on the mobility of Cd in an (A) Oxisol and (B)
Alfisol (after Kookana and Naidu, 1998, with permission).
Characterization of the Combined Effects of
Ionic Strength, pH, and Cd Concentration
The variation in ionic strength often causes a change in soil solution pH and, there-
fore, the results are often a reflection of combined effects of pH and ionic strength.
However, some workers in recent years have tried to quantify the individual contribu-
tions of pH and ionic strength. To discern the effect of ionic strength, pH, and [Ca], a
simple theoretical relationship was developed by Temminghoff et al. (1995). The rela-
68 Fate and Transport of Heavy Metals in the Vadose Zone
0.9
0.8
0.7
0.6
o 0.5

u 0.4
0.3
0.2
0.1
0.01 M CaCI
2
o 0.01 M Ca[N0312









o
o
o
o
o
o
o
o
o 0
o
o 10 20 30 40 50 60 70 80
Pore Volumes
Figure 3.4. Cd breakthrough curves as influenced by complexation with chloride in an Oxisol [data
for 0.01 M Ca(N0
3
h from Kookana et aI., 1994, with permission). Flow rate 13.5 cm/h (Cn and
12.3 cm/h (N0
3
-); pH 4.8-4.9 (Cn, 4.9-5.0 (N0
3
-). Other experimental conditions were same as
described by Kookana et al. (1994).
tionship describes the pH-dependency of Cd ion binding and also the effect of competi-
tion by Ca. Since [Ca
2
+] [Cd
2
+] in most cases, this simple relation can be expressed as:
(3)
where is the adsorbed quantity of Cd, K' is a modified Freundlich sorption coeffi-
. ned d nCa Cd d C 'fi 'd l' d .
Clent, an are - an a-speCl lC nonl ea lty parameters, an m IS a parameter
based on ned and nCa representing the relative replacement ratio of H+ by Cd
2
+ (for details
see Temminghoff et aI., 1995). In Equation 3, activities instead of concentrations have to
be used. According to Equation 3, the Cd sorption data corresponding to different pH,
[Cd], and [Ca] obtained from batch and transport experiments (Kookana and Naidu,
1998) follows a straight line, as shown in Figure 3.5.
In Equation 3, the exponents for pH and Ca are indicative of the sensitivity of Cd
sorption to these two parameters. These exponents have been reported to vary with soil
type. For example, for a clayey soil (Oxisol) the value of the exponent m was found to be
-1.3 by Naidu et al. (l994a), in contrast to a value of -0.77 for a sandy soil reported by
Boekhold et aI. (1993). The marked differences in m for these soils may signifY the dif-
ferent nature of the soil particle surfaces and electrolytes used during sorption studies.
Similarly, the sensitivity of sorption to [Ca
2
+] has also been found to vary between soils.
Kookana and Naidu (1998) found the nCa values for an Oxisol to be -0.21 and -0.61 for
an AlfisoI. Temminghoff et aI. (1995) reported a value of nCa of -0.34 for a sandy soil,
whereas Chardon (1984) reported an average value of -OA1 0.07 for six different soil
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 69
0.04 -.-------------------------------,
0.03
OJ
.:.:
-(5
E
E
;-0.02
QI
.c
....
o
UI
"C
o
0.01
0
'"
0




'"
'"
'"



0.0005 0.001 0.0015 0.002 0.0025
Figure 3.5. Cd sorption as a function of [WI, [Cd
2
+[ concentration and [Ca
2
+) activities in soil
solution. column data, .A. batch data (Kookana and Naidu, 1998, with permission).
types. The low nCa for the Oxisol may be due to its low inherent affinity for Cd. In this
soil the pH is close to the PZC (point of zero charge), and so the effect of ionic strength
on Cd sorption is much less pronounced.
Precipitation
The role of precipitation and dissolution reactions in determining solubilities of trace
elements in soil solution is generally only important in cases where either the concentra-
tion of metal is very high or when the conditions in soils are such that certain anions or
ligands are present. For example, under reducing conditions sulfide salts of metals such
as Cd, Zn, Hg, Fe, etc., may be important (Tiller, 1996).
Interactions of Cd with surfaces of calcite-the most common carbonate in soils-
have been described by various authors as chemisorbed complexes, surface precipitates,
etc. (see Zachara et al., 1991). However, initial rapid sorption of Cd on calcite surfaces
may be followed by a slower phase of dehydration and coprecipitation (e.g., Davis et al.,
1987; Zachara et al., 1991). There is a possibility that following sorption, Cd may mi-
grate into the solid phase, forming a solid solution, as demonstrated by Stipp et al' (1992).
Stipp et al. (1992), on the basis of studies on Cd uptake by calcite using near-surface
sensitive techniques, provided experimental evidence for the "formation of a thin, surficial,
solid solution precipitate during initial contact with solution," which disappeared on the
order of weeks due to the diffusion of the trace metal into the solid phase. They also
verified by surface analysis, the precipitation of a nearly pure, crystalline otavite (CdC0
3
)
70 Fate and Transport of Heavy Metals in the Vadose Zone
overlayer on calcite. Therefore it is possible that at elevated levels of Cd in the carbon-
ate-rich systems, otavite can playa role in controlling concentrations of Cd in terrestrial
and aquatic environments. However, as studies by Holm et al. (1996) showed, even in
carbonate-rich aerobic soils the Cd concentration in solution was not governed by otavite,
presumably due to the presence of inhibitors of precipitation such as dissolved organic
matter in soil solution.
The precipitation that follows the sorption reaction of Cd on calcite and the probable
incorporation of the sorbate into calcite can result in its limited desorption (Zachara et
aI., 1991). Therefore, in calcareous soils and groundwaters, calcite can act as an impor-
tant sink for Cd and some other metals. The kinetics of calcite recrystallization (Zachara
et aI., 1991) and Cd diffusion into the solid phase (Stipp et aI., 1992) may contribute to a
nonequilibrium sorption behavior of Cd during transport in soil and groundwater.
Kinetics of Cd Sorption
In well-mixed systems involving soils and other sorbents, the sorption of Cd as well as
other ion exchange reactions has generally been found to be fast (Christensen, 1984a;
Hayes and Leckie, 1987; Hachiya et al., 1979). With reaction half-lives on the order of
minutes or less, they are often complete by the time solid and liquid phases can be sepa-
rated (e.g., Zasoski and Burau, 1978; Harter and Lehmann, 1983; Jardine and Sparks,
1984). Some of these studies have been carried out at very high solution to soil ratios,
and involve continuous agitation facilitating the accessibility of sorption sites to sorbent.
The sorption reaction of Cd in soils has been found to be generally complete within
hours. For example, Christensen (1984a), using low concentrations of cadmium, reported
that more than 95% sorption in soils (pH 6.0 to 6.5, background salt 0.001 M CaCI
2
)
studied was complete within 10 minutes and the equilibrium was achieved within 1 hour
with no further increase in sorption for 67 weeks (Figure 3.6). Chardon (1984) studied
sorption of Cd after 23 hand 46 h on 12 soils [pH 3.3 to 7.6; background salt 0.0015 M
Ca(N0
3
hJ and reported that for one soil only a slightly higher sorption (but significant)
was noted after 46 h. Kookana (1997; unpublished data) observed that cadmium sorp-
tion on four soils (Alfisols from South Australia, pH 5.5 to 7.3) was essentially complete
within 3 h of shaking; more than 95% of maximum observed sorption occurred in the
first 30 minutes.
On synthetic minerals and natural soil sorbents, biphasic sorption reactions have been
reported for Cd. On synthetic goethite, for example, Brummer et ai. (1988) noted a
rapid initial sorption of the metals (Cd, Zn, and Ni), followed by a much slower reaction
akin to a diffusion-controlled penetration into goethite. However, among the three met-
als studied, the smallest increase in the magnitude of sorption with time at any given pH
was with Cd. A batch kinetic study on Cd sorption in five soils by Selim (1989) also
showed a fast sorption reaction followed by a slower reaction, more pronounced in some
soils than others. Sorption of trace elements showing a rapid initial sorption followed by
a much slower reaction has been reported by other workers also (Gerth et aI., 1993).
Similarly, Fuller and Davis (1987) noted that Cd sorption on a calcareous aquifer sand
followed two reaction steps, the second slower step continuing for 7 days. In this case,
the sorption might have been followed by a precipitation reaction.
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 71
8 ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
c;;- 66 ...
E
~ 50-
C)
::i.
-;- 40 r-
o
~
-
c:
Q)
c.:>
c:
o
c.:>
"C
U
0
1
1
o 10 20
A
A A
-
1 .. 1 1 ..
30 60 180
t::.. sandy loam
loamy sand
A
1..>.. I..>.
-
I 1 1
1200 1400 1680
Contact time (minutes)
Figure 3.6. Solution concentration of Cd as a function of time in a batch sorption experiment in
two soils at low concentrations (after Christenson, 1984a, with permission).
The rate of sorption reaction can significantly influence the transport behavior of Cd
through soil profiles or aquifer systems. The observation that in calcareous soils or aqui-
fers the sorption may be followed by precipitation is particularly noteworthy, due to its
potential effect on reaction kinetics and thereby on transport of Cd through the systems.
Several studies on Cd transport are available in the literature (e.g., Dowdy et al. 1991;
Dunnivant et aI., 1992; Boekhold and Van der Zee, 1992; Kookana et aI., 1994;
Bajracharya et aI., 1996; Streck and Richter, 1997); however, most have been carried
out in soil columns under laboratory conditions. While a detailed account of Cd trans-
port is beyond the scope of this chapter, studies on sorption of Cd during transport have
been discussed below.
Sorption Behavior of Cd During Transport Through Soil Columns
Batch versus Flow-Through Systems
At a constant pH and for the same adsorbing media, it is often assumed that the amount
of Cd sorbed by soil under batch and flow-through systems should remain the same.
However, a comparison of data from the two systems has shown that sorption coeffi-
cients obtained by the two methods may not always be in good agreement (e.g., Boekhold
and Van der Zee, 1992; Burgisser et aI., 1991; Grolimund et al., 1995; Bajracharya et al.,
1996). Klamberg et al. (1989) found that the maximum sorption capacity for copper with
humic acids was greater in column than in batch experiments. There may be several
reasons for these differences (Grolimund et al., 1995), as described below.
Effect of Solid-Solution Ratio
A major difference between batch and flow-through systems (particularly using packed
soil columns) is the soil:solution ratio. Past research indicates that the solid-to-solution
ratio affects both the rate and extent of sorption (Tan and Teo, 1987; Boesten and Van
der Pas, 1988). However, different solid-to-solution ratios have been used in various Cd
sorption studies published in the literature. Bajracharya et al. (1996) compared Cd sorp-
tion under batch and flow conditions and found that the batch-determined Cd partition
72 Fate and Transport of Heavy Metals in the Vadose Zone
coefficients (linear and Freundlich) were around 60-80% higher than that determined
by any of the column experiments conducted at various flow rates. The solid-to-solution
ratio in their column experiments was about 5 g: 1 mL, in contrast to a ratio of 1 :50 in the
batch experiments. Tan and Teo (1987) observed the influence of solid-to-solution ratio
on equilibrium solid phase concentration and noted that for the same initial concentra-
tion, the equilibrium solid phase concentration was lower at a higher solid-to-solution
ratio. Hence, the higher linear and Freundlich coefficient values observed in batch ex-
periments by Bajracharya et al. (1996) could be due to the solid-to-solution ratio effect.
This was also apparent in their batch experiments with different solid-to-liquid ratios
(Figure 3.7). As shown in the figure, the partition coefficient decreases with increasing
solid-to-liquid ratio. However, for the sand used in these experiments, as the ratio be-
comes smaller, very little effect is observed beyond the ratio of 0.5 g: 1 mL. The partition
coefficient at a 1 g: 1 mL solid-to-liquid ratio (8.65), i.e., similar to that in a flow-through
experiment, was less than the column-determined values (between 10 and 12), suggest-
ing that the dynamic effect of flow could influence Cd sorption. Thus, even though the
solid-to-solution ratio does affect the partition coefficient, it does not explain the differ-
ence in ~ values between batch and column experiments.
Akratanakul et al. (1983) suggested that even when the soil-to-solution ratio is the
same, the sorption in flow-through experiments should be greater than that in corre-
sponding batch experiments. They reasoned that since batch systems are closed, the
amount of ions in the liquid phase decreases with time as more and more ions are adsorbed
onto soil particles. In flow systems, desorbed ions are carried away by the flowing solu-
tion, replacing desorbed ions with a higher concentration of Cd ions. This effectively
exposes the soil particles to a continuous source of ions. They also suggested that the
amount of ions that are exposed to soil particles in a batch system is equal to the concen-
tration of the solution times the volume. In a dynamic system, the product of concentra-
tion, flow rate, and exposure time gives the amount of ions exposed to soil particles. This
implies that there are more Cd ions available for sorption in the flow-through system and
that greater exposure of Cd ions increases sorption of Cd to soil particles. A schematic
representation of the two systems has been provided in Figure 3.8.
In batch and column experiments on copper sorption in soils, Grolimund et al. (1995)
observed a significant dependence on sorbed amount of the solid concentration. How-
ever, once the sorbate was prewashed before sorption was measured, this effect disap-
peared and the results from different techniques were in good agreement. Therefore, it
was concluded that the particle concentration effect was due to incomplete removal of
preadsorbed ions (e.g., AI, K, Mg) or the presence of complexing agents (e.g., dissolved
organic matter). Often, column experiments involve preconditioning of soil with a back-
ground electrolyte solution before sorption is measured, which is usually not done in
batch experiments. The disagreement between batch and column studies, therefore, can
be attributed to the physicochemical differences resulting in a multicomponent effect on
sorption equilibria.
Flow-Affected Sorption
The amount of sorption is expected to be same at different flow velocities as long as
sufficient time of contact with the sorbent is allowed. However, there are conflicting
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 73
20
,-.,

~
....;;l
'-'

.....
s::
Q)
......
C,)
......
1t
10

Q)
0

C,)
s::
0
'p
. ~
p..
0
0.0 0.5 1.0
Solid-to-liquid ratio
Figure 3.7. Effect of sOlid-to-liquid ratio on adsorption coefficients in batch reactors (Bajracharya
et aI., 1996, with permission).
Solution
Sorbed
t
Sorbed
Batch system
Time
Solution
Sorbed
. ~
Sorbed
~

Solution
U
Time
Flow-through system
Figure 3.8. A schematic representation of concentration changes in sorbed and solution phases
with time as an equilibrium between the two phases is approached for the batch and flow-
through systems.
reports in the literature on this phenomenon. For example, Akratanakul et al. (1983)
conducted flow-through experiments at three different flow velocities (1.3, 2.3, and 2.7
cmlh) for Pope Ridge soil developed from volcanic deposits and found the rate of sorp-
tion of Cd increased with flow velocity. Kookana et al. (1994) conducted flow-through
experiments on Cd at two different flow velocities for an Oxisol soil. From their two
breakthrough curves, it is evident that the curve at the higher velocity emerged later,
indicating greater sorption at the higher velocity. However, the higher pH at higher flow
velocity, albeit by only 0.2 unit, would have contributed to the increased sorption. The
linear sorption coefficients evaluated by fitting a two-site model to two other column
experimental breakthrough curves in an Oxisol and a Spodosol were found to be larger
than the corresponding batch-determined ones.
74 Fate and Transport of Heavy Metals in the Vadose Zone
There are other reports in which the pore water velocity has the opposite effect on
solute sorption. For example, Bajracharya (1989) conducted Cd sorption experiments in
river sand at flow velocities of 0.83 cmlh and 9.17 cmlh and found that at the higher
velocity, sand exhibited a lower sorption of Cd. Similar velocity effects on sorption have
also been reported for other less strongly adsorbed solutes (Miller et aI., 1989; Shimojima
and Sharma, 1995). Retardation factors for tritium and bromide have been observed to
depend strongly on pore water velocity, with the retardation factor decreasing with in-
creasing velocity (Schulin et aI., 1987).
There are, again, reports where batch- and flow-determined sorption parameters were
within the same range. For Cd, Theis et al. (1988) reported that the total reactive surface
site density from column studies agreed well with the values obtained from batch stud-
ies. Boekhold and Van der Zee (1992) reported that batch-determined Freundlich sorp-
tion coefficients of Cd for a soil adequately described the observed Cd breakthrough
curves from column experiments. Similarly, Burgisser et al. (1991) in their experiments
on Cd sorption on sand (particle size, 125 to 250 I.lm) found that the batch sorption data
was in good agreement with the isotherm calculated from the BTCs, obtained at two
different flow velocities. They showed that kinetic effects were absent in their experi-
ments, as the BTCs were unaffected by flow velocities.
The factors affecting heavy metal sorption are difficult to isolate from experiments
conducted on natural soils. Characteristics of natural soils are very difficult to ascertain
precisely. Soil heterogeneity makes replicate experiments subject to uncertainty. The
effects of pore water velocity on Cd sorption are difficult to establish even in relatively
homogenous porous media, as shown by Bajracharya et al. (1996). They carried out 13
flow-through Cd sorption experiments on a uniform graded silica sand, mostly at a con-
stant pH of6, but using a wide velocity range (5-214 cmlh). The partition coefficient, Kd
[L
3
/M] , obtained by fitting solutions of the advection-diffusion equation to concentra-
tion breakthrough curves from their experiments, is plotted against pore water velocity
in Figure 3.9. Even in this artificial homogeneous sorbing medium, no conclusive trend
in ~ with pore water velocity could be established.
From the above discussion it is unreasonable to draw any definite conclusion in terms
of similarity in results from batch and the flow-through systems. Not only are the two
systems very different from each other, the sorption of Cd is influenced by several physi-
cal and chemical factors in the two systems, as discussed above. However, it is clear that,
in addition to the differences in factors such as soil solution composition and pH, it is
important to establish that kinetic effects are absent under the experimental conditions
employed in flow-through systems.
Evidence of Sorption Nonequilibrium During Cd Transport Through Soil
Asymmetrical Breakthrough Curves
The linear equilibrium sorption during transport of a solute normally results in sym-
metrical breakthrough curves. On the other hand, a nonequilibrium process is said to
occur when a solute undergoes any time-dependent reaction in addition to the standard
advection and dispersion transport processes. Early and asymmetric breakthrough curves
are characteristic of nonequilibrium processes, and can be caused by various physical
and chemical processes. On the other hand, asymmetrical BTCs can result from nonlin-
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 75
14
-
"""' btl
~
.....l
13
'-"
....
I::
Q)
.(3
1+=
12
4-<
Q)
0

u
I::
0
. ;:1
11 ...
~



~


10

1 10 100 1000
Pore water velocity (cmlh)
Figure 3.9. Variation of linear Cd sorption coefficient with pore water velocity (Bajracharya et aI.,
1996, with permission).
ear equilibrium sorption from the liquid to solid phase. In principle, a nonreactive chemical
would exhibit no asymmetry in the breakthrough curve when passed through the same
homogeneous soil in which a reactive chemical shows a time-dependent sorption reac-
tion. Tailing in the breakthrough data indicating sorption nonequilibrium during Cd
transport has been reported in miscible displacement studies carried out on soils and
clay-humic acid mixtures by several authors (Morisawa and Inoue, 1985; Campbell et
al., 1987; Selim et al., 1992; Boekhold and Van der Zee, 1992; Kookana et al., 1994).
Flow-Interruption as a Test for Sorption Nonequilibrium
Another technique to check for the occurrence of nonequilibrium processes is the
flow-interruption method. The presence of a measurable depression in the rising limb of
the BTC immediately after the flow is resumed signifies a nonequilibrium process method
(Murali and Aylmore, 1980; Brusseau et al., 1989; Kookana et al., 1994). The purpose
of stopping the flow is to enable sufficient time for solute to be adsorbed onto the sorp-
tion sites.
Using the flow-interruption technique, Kookana et al. (1994) noted that nonequilib-
rium conditions existed during their flow experiments involving Cd. Recently, Tran et
al. (1998a) further investigated Cd sorption nonequilibria through a series of flow-inter-
ruption experiments. In these experiments, a steady flow in homogeneous sand was es-
tablished, after which Cd was introduced at constant pH. A flow-interruption period of
24 hr was applied once a complete breakthrough curve had been obtained; i.e., when the
effluent concentrations were close to the influent concentration. Immediately after this
no-flow period, the effluent solution was monitored for both Cd concentration and pH.
As already noted, drop in effluent concentration immediately after the resumption of
flow indicates nonequilibrium behavior. This behavior, however, was not observed. Fig-
ure 3.10 shows a plot reported by Tran et al. (1998a) in which an increase in effluent
concentration was observed immediately after resumption of flow. Clearly this increase
of effluent concentration was the result of desorption of Cd from the solid phase during
76 Fate and Transport of Heavy Metals in the Vadose Zone
u
1.5
0.5

,
,
,
,
,
,

", ...
0.0 .. _"'-----l'-----"_--'-_ ....... _-'-_ ........ _""---_"'----'
Observed data
- MCMFIT fitted
- - - Two-site model simulated
o 20 40 60 80 100
Time (h)
Figure 3.10. Cd flow-interruption experiment reported by Tran et al. (1998a, with permission).
Circles are experimental data, and lines are fits of the reaction-advection-diffusion transport model.
the no-flow period causing an increase in the Cd concentration of the interstitial pore
water. The temperature and pH changes were small and insufficient to explain the ob-
served desorption. However, in some other studies on Cd, increases in effluent concen-
trations have been observed at the start of desorption in flow-through experiments (e.g.,
Dunnivant et al., 1992; Selim et al., 1992). This has generally been termed the "snow-
plow effect," which is not related to the nonequilibrium effect and can be caused by
factors such as changes in ionic strength of influent solutions. Tran et al. (1998a) fitted
transport model solutions to the experimental data; some results are shown in Figure
3.10. Note that the post-interruption limb of the data is fitted well by the two-site, non-
equilibrium transport model, which suggested that either non equilibrium sorption was
occurring, or that a nonlinear, equilibrium isotherm was active.
ModeJ Fitting
Sorption equilibrium or nonequilibrium during transport of a solute can be tested by
checking the fit of various equilibrium or nonequilibrium models to the data. It has com-
monly been observed that for a range of inorganic and organic solutes, a two-site non-
equilibrium model successfully described the observed BTCs (e.g., Valocchi, 1985;
Dunnivant et al., 1992; Gaber et al., 1992; Kookana et al., 1993, 1994). Similar sorption
behavior has been observed for Cd. Dunnivant et al. (1992), for example, noted that a
two-site nonequilibrium approach was necessary to describe Cd BTCs in aquifer col-
umns, particularly in the presence of dissolved organic carbon. Kookana et al. (1994)
reported that the two-site non equilibrium model fitted their observed breakthrough data
better than the equilibrium model in two soils. From experimental verification with
nonreactive solutes, they suggested that in an Oxisol, the observed nonequilibrium be-
havior of Cd might have been due to sorption-related nonequilibrium, whereas physical
nonequilibrium, perhaps due to the presence of immobile water, was more likely in a
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 77
Spada sol. Hinz and Selim (1994) studied Cd movement in two soils and used the ob-
served BTCs to test equilibrium sorption models based on eight different sorption iso-
therms. They observed that, while in one soil (Typic Udipsamment) transport models
based on equilibrium sorption isotherms were adequate to describe Cd and Zn BTCs,
none of the models could describe the BTCs in the other soil (Aquic Fragiudalf). The
authors suggested that possibly the finer texture of the latter soil may have caused the
nonequilibrium conditions.
In other studies, equilibrium sorption models have been found to be sufficient to de-
scribe the observed Cd sorption in column experiments (e.g., Gerritse, 1996; Bajracharya
et aI., 1996). Gerritse (1996) in his study, however, noted that with the equilibrium
sorption model, the dispersivity for Cd was greater than that for Cl, a nonreactive solute.
The increased dispersivity was ascribed to the heterogeneous distribution of sorption
sites in soil. Gerritse (1996) also showed that a BTC from a Spodosol, which showed
substantial tailing and fitted a two-site nonequilibrium model (Kookana et aI., 1994),
could be described by allowing a higher dispersivity in the equilibrium model. The in-
crease in dispersivity could result from both nonequilibrium as well as equilibrium sorp-
tion conditions. For example, the dispersion due to solute exchange between mobile and
immobile regions (physical nonequilibrium) can be lumped into a dispersivity term (e.g.,
Passioura, 1971; van Genuchten and Dalton, 1986). Similarly, pore scale heterogeneity
could lead to different dispersion coefficients for reactive and nonreactive solutes even
when local equilibrium conditions are maintained during flow (e.g., Sugita and Gillham,
1995). A clear discrimination between equilibrium and nonequilibrium processes should
be possible only through experiments in which Cd transport is studied by combining
varying flow rates with flow interruption tests.
Mass Balance Check for Complete BTCs
In cases where complete breakthrough is achieved (i.e., the concentration in the break-
through curve is the same as that in the feed solution), the concept of column holdup
(Huber and Gerritse, 1971; van Genuchten and Parker, 1984; Barry and Sposito, 1988;
Barry and Bajracharya, 1995; Bajracharya and Barry, 1997a), i.e., the mass of Cd re-
tained in column, can be used to check the results of model fitting. The holdup or the
total mass of Cd in the column at the instant when influent concentration equals that of
the effluent is (Barry and Bajracharya, 1995):
(4)
where So is the solid phase concentration existing in equilibrium with the influent liquid
phase concentration, Co, and Vo is the pore volume in the sand column. Note that this
expression is valid for either linear or nonlinear, and for both equilibrium and none qui-
librium sorption.
Holdup can be calculated in two ways, either directly from the breakthrough curves
or from the above equation using the fitted isotherm. Bajracharya et al. (1996) calcu-
lated H using both approaches, and found the results agreed very well except for the
single experiment where the model fits indicated nonequilibrium sorption to be active.
78 Fate and Transport of Heavy Metals in the Vadose Zone
In this one case, the holdup estimates based on the nonequilibrium model differed from
the experimental value by 20%. Their results show that some care is needed in interpret-
ing results of model fits to data, even in cases where good fits are obtained.
Causes of Sorption Nonequilibrium During Transport
Sorption nonequilibrium may arise due to the sorption reaction itself being slow (chemi-
cal kinetics), or if the soil solution is not well mixed (physical nonequilibrium), or both.
Skopp (1986) provided a critical analysis of time-dependent flow processes in soils. Given
that chemical kinetics of Cd are relatively fast and that the occurrence of sorption non-
equilibrium during transport tends to depend on soil type (as discussed above), it is
likely that in most cases reported above, the time-dependency of Cd sorption represents
physical nonequilibrium, such as incomplete mixing of solution between mobile and im-
mobile domains in soil.
Time-dependence of Cd sorption during transport might be due to film or intraparticle
diffusion, or diffusive transfer between mobile and immobile portions of the flow domain
(Barry and Li, 1994). In both cases, the nonequilibrium is caused by diffusive transport
within some stagnant liquid in the soil and is characterized by a transfer rate parameter,
a and /3, the fraction of pore space that is mobile. The parameters, a and /3, are known as
nonequilibrium parameters, and can arise in modeling either physical or chemical non-
equilibrium solute transport processes. The rate parameter, a, is best considered as an
apparent parameter. It has been shown to vary under variations in flow velocity (Schulin
et al., 1987; Griffioen et al., 1998). The parameter, /3, in general does not vary much with
water content or pore water velocity (Schulin et al., 1987; Bajracharya and Barry, 1997b;
Griffioen et al., 1998), although there are some reports indicating velocity dependence
(e.g., Nkedi-Kizza et al., 1983; Brusseau et al., 1994).
The main physical transport processes are characterized by the solute advection rate,
V [LT-
1
], and the dispersion coefficient, D [L
2
T-
1
]. Solutes that undergo time-depen-
dent reactions with soil minimally require two more parameters; viz., the equilibrium
partition coefficient, K [0 M-
1
], and the (in this case) chemical reaction rate, a [T-
1
].
Both K and a are usually determined from batch experiments, while V and D are deter-
mined using a nonreactive tracer in a flow experiment. This separation of physical and
chemical parameters is not always useful. Gerritse (1996), for example, found that the
dispersivity for transport of Cd was much greater than that for CI-, which is considered
to be a tracer. Other, similar, findings have been reported by Boekhold and Van der Zee
(1992), Kookana et al. (1994) and Hinz and Seiim (1994).
Cd Transport Under Field Conditions and Its Modeling
There are relatively few studies on the transport behavior of Cd under field conditions
and its model simulation (Sidle et al., 1977; Dowdy et al., 1991; Streck and Richter,
1997a). Slow displacement of heavy metal cations and therefore the need for long-term
experimentation, and the environmental concerns associated with any input of heavy
metals for experimental purposes may be some of the reasons for small-scale rather than
large-scale studies (Streck and Richter, 1997a). However, continuous intentional or un-
intentional applications of metal-rich sewage sludge or wastewater to land or the expo-
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 79
sure of sites to industrial wastes has provided a situation under which heavy metal trans-
port has been studied.
Generally, the downward movement of heavy metals has been found to be very slow,
restricting them within a few centimeters of the incorporation depth. However, in some
cases significant leaching of Cd and Zn has been observed (Dowdy et al., 1991). In the
study by Dowdy et al. (1991), however, massive amounts of sludge were applied every
year for 14 years on highly structured soils. Lund et al. (1976) reported elevated concen-
trations of Cd, and other heavy metals up to a depth of 3 m in coarse-textured soils under
sludge drying ponds in use for >20 years. In another study involving a sewage sludge
disposal site, Sidle and Kardos (1977) found that 6.6% of the applied Cd in the form of
sludge was recovered in the percolate at 120 em depth in the soil profile. In this study,
simulations with a simple transport model based on Freundlich adsorption isotherm of
cationic form of Cd predicted virtually no movement of Cd in the soil (Sidle et al., 1977).
They concluded that the complexed form of Cd as well as its movement through prefer-
ential pathways should be taken into account for any predictive simulation. Recently, in
a study by Streck and Richter (1997a), transport of Cd and Zn was investigated at field
scale following application of wastewater for 29 years on a sandy soil. They found that
Cd and Zn were partly displaced to a depth of 0.9 and 0.7 m, respectively. The simula-
tions of Cd leaching were carried out with various types of modeling approaches em-
ploying parallel soil column (PSC) and convective-dispersive (CD E) approaches using
either a grid model (points in the field) or Monte-Carlo model (Streck and Richter,
1997b). Sorption was described by an extended Freundlich equation capable of taking
the spatially variable organic carbon content and pH at the site into account. The ob-
served profile of Cd agreed well with that predicted with the PSC model, both with grid
and Monte-Carlo simulations. Therefore, it was concluded that the spatial variability of
sorption could adequately describe the field-scale dispersion of Cd. It is noteworthy that
the sorption data used in the simulations were measured in the presence of an electrolyte
matching the mean ionic strength of the wastewater used for irrigation and included CI
anion (0.0025 M CaCI
2
). The simulations with the CDE modeL however, could agree
with measured data only when the dispersivity parameter was adjusted to 0.29 m-a
value considered to be on the higher side of those commonly reported. As discussed in
the earlier section, increased dispersivity may result from the physicochemical heteroge-
neity or nonequilibrium conditions during sorption. The results from this study high-
light that spatial variability of key soil properties affecting Cd sorption, such as pH and
organic carbon content, has a greater bearing on transport behavior of Cd than perhaps
the nonidealilty of microscale processes, such as sorption nonequilibrium.
The importance of heterogeneity of soil properties and spatial variability in determin-
ing Cd transport through soil was also demonstrated through simulations with a simple
root zone model (using stochastic theory) by Boekhold and Van der Zee (1991) for a
sandy soil. In the example they considered (representing a site exposed to combined agri-
cultural and industrial activities with 50 glhafyr rate of Cd input), simulations showed
that after 40 years, the average Cd concentration in mean water flux can reach the Dutch
reference value for groundwater (1.5 J.lglL). They recommended that when groundwater
quality is of major concern, accurate knowledge of sorption parameters and input rates of
Cd are crucial for reliable results, because leaching rates are very sensitive to Cd input
rate and to flow and sorption parameters. They observed that large areas in the field may
80 Fate and Transport of Heavy Metals in the Vadose Zone
have high leaching rates, which may remain undetected by simulations with the average
behavior of Cd. Therefore, soil heterogeneity of both soil physical and chemical proper-
ties must be taken into account in an assessment of Cd leaching through soil profIles.
Desorption and Reversibility of Cd Sorption
Reports of desorption of Cd from soils are relatively fewer in the literature than those
of sorption, except those carried out using specifIc extractants to establish the solid phase
speciation of Cd (Tiller, 1996). In sorption-desorption experiments on Cd in soils, both
complete and partial reversibility of sorption have been reported in the literature. For
example, Christensen (1984b) studied desorption of Cd at low Cd concentrations (0.1 to
6 mg/g in soil) in two Danish soils (loamy sand and sandy loam) at pH 6.0. They ob-
served a full reversibility of Cd sorption in the loamy sand but only partial in the sandy
loam. Mayer (1978) also noted similar full reversibility of Cd sorption in an acid subsur-
face soil over a wide range of solution concentrations of Cd (1-10000 ~ / L ) . Complete
reversibility of sorbed Cd from poorly crystalline kaolinite was also reported by PuIs et
al. (1991). Similarly, Cd sorption was found to be completely reversible in both column
and batch experiments in an Australian Oxisol, whereas hysteresis was observed in AlfIsol
(Kookana et al., 1994; Naidu et al., 1997). The ambient pH of the Oxisol was closer to
the point of net zero charge and therefore the soil had a very low cation exchange capac-
ity. In contrast, Amacher et al. (1986), while studying the desorption behavior of Cd in
fIve different soils also found incomplete reversibility of several metals, including Cd.
Mter allowing the sorption reaction between metals and soils to proceed for 336 hr, they
carried out desorption in 0.005 M Ca(N0
3
h A signifIcant fraction of the sorbed metals
did not desorb from soils (Table 3.1) even after the long desorption period. In this study
the Ca was present in suffIciently high concentration as Ca(N0
3
)2 to replace the Cd on
exchange sites. It was suggested that the poor reversibility in these studies may have
been due to specifically sorbed Cd on metal oxides or organic matter, or due to forma-
tion of insoluble compounds or coprecipitated Cd. Recently, Kookana (unpublished data)
observed sorption desorption hysteresis in two AlfIsols from South Australia, as shown
by the data presented in Figure 3.llA,B. Tran et al. (1998b) reported Cd desorption
experiments carried out on a homogeneous sand medium at constant-pH of 5.5 and 6.5.
They compared sorption/desorption isotherms and found signifIcant hysteretic behavior
(Figure 3.11 C) at a pH of 6.5. The partial reversibility of Cd is likely to be linked to the
mechanism of sorption in soils, as discussed below.
Desorption of Specifically Sorbed Cd
Tiller et al. (1984b) in their desorption studies on clay-sized isolates from several soils
found that at a soil pH of 5, up to 85% of sorbed Cd was easily desorbed rapidly in 0.01
M Ca(N0
3
h However, at pH 7, they noted that the easily desorbed fraction (they
called it nonspecifically sorbed) was much lower, particularly from the clay-sized frac-
tion from Oxisol. The lower proportion of desorbed Cd at higher pH may be because the
surface may be highly undersaturated relative to the number of sorption sites available
for binding (Naidu et al., 1997). Thus, it appears that the reversibility of sorbed Cd is a
function of the nature of sorbent and the soil conditions determining the affinity of Cd
for soil (Kookana et al., 1997). As discussed earlier, Cd can form high affinity inner-
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 81
Table 3.1. Fraction of Sorbed Cd Released in Solution after 336 hr Desorption
a
Soil Type
Typic Hapludults
Typic Udifluvents
Aquic Fragiudalfs
Vertic Haplaquepts
Typic Udipsamments
pH
5.1
7.4
6.4
5.4
5.4
CEC Fe
2
0
3
Cd Released
cmol(+)/kg (%) (%)
3.72 10.2 50.3-74.9
6.20 0.44 5.4-34.1
8.31 1.14 2.2-15.6
31.3 0.94 3.6-9.4
1.20 2.20 24.7-51.5
a After Amacher et aI., 1986.
8.0,------------------,
A
6.0 f-
0.0 r- B
30.0 I-
I I I

OJ
20.0 I- 0-----
E /e
C 10.01- ./
__ __ __
o 0.0 ...
C/)
c
cp
//
I ,.
cp /'
I e
10 I I I
2.5 L__---'-__ __ _'__ __ '___--'
4.0 -
3.5 -
3.0 -
0.0 0.1 0.2 0.3 0.4 0.5
Solution concentration (rng/L)
Freundlich K
(n)
8.0 (0.89)
71.5 (0.81)
147.2 (0.85)
189.2 (0.92)
21.8 (0.84)
Figure 3.11. Cd adsorption and desorption batch isotherms for (A) an Alfisol, solution pH 6.1 0.1 ,
(B) an Alfisol pH 6.6 0.2 (Kookana, unpublished data), and (C) a homogeneous sand; solution
pH 6.5 (Tran et aI., 1997b, with permission). The ordinate represents the Cd solution phase
concentration, and the abscissa is the solid phase concentration.
sphere complexes as well as coprecipitates with certain minerals and from such sorbent
the desorption of Cd is unlikely to be fully reversible.
On certain sorbents, the desorption of Cd and other metals has been found to be
influenced by the time of metal-sorbent contact (Brummer et al., 1988; Backes et al.,
1995). Backes et al. (1995) studied the kinetics of Cd and Co desorption from synthetic
Fe and Mn oxides (at pH - 6) by both batch and flow methods. They concluded that not
only the oxide sorbed large amounts of Cd and Co but substantial proportions of sorbed
metals could not readily be des orbed in soil solution, especially from Mn oxides. The
82 rate and Transport of Heavy Metals in the Vadose Zone
rate of desorption from goethite became progressively slower with contact time between
sorbate and the sorbent. While, as the authors state, several mechanisms may cause such
effects, it is not clear if the physical or chemical changes were responsible for reduced
desorption with contact time.
Partial Reversibility of Cd Sorption from Calcite and Calcareous Soils
In batch experiments on calcite surfaces, sorption of Cd has been found to be only
partially reversible (Zachara et al., 1991). Similarly, in flow-through experiments on a
calcareous soil, Buchter et al. (1996) found that 35% of the applied Cd did not elute from
the column, indicating sorption hysteresis. They also reported that, in a batch study on
the same soil, pronounced hysteresis in Cd sorption-desorption was observed. Batch
desorption experiments on a calcareous soil from South Australia (Kookana, unpub-
lished data in Figure 3.11) also show the sorption hysteresis of Cd. The partial reversibility
from calcite may be due to dehydration of sorbed Cd and coprecipitation as suggested by
Davis et al. (1987). Cd reversibility may be time dependent and may be so slow that
during its transport in natural systems such as groundwater, nonequilibrium behavior
could become evident (Zachara et al., 1991).
((1 Desorption Kinetics
Desorption kinetics of Cd in soils is relatively little understood. However, from pub-
lished studies on natural soils and synthetic minerals it appears that desorption kinetics
of Cd depend on the sorbent properties as well as experimental conditions. In batch
experiments involving shaking, often the desorption equilibrium is achieved within hours.
For example, Tiller et al. (1984b) noted that, at pH 5, up to 85% of Cd sorbed on clay-
size fractions from soils was desorbed rapidly in one quick wash (5 min) with 0.01 M
Ca(N0
3
h Kookana et al. (1997) reported that, in two Australian soils, solution concen-
tration of Cd during desorption did not significantly change after 2 hours of shaking.
Similarly, Tran et al. (1998b) carried out a series of batch kinetic desorption experi-
ments at pH 6 which showed no significant difference in Cd solution phase concentra-
tion between 1 day and 10 days equilibration, indicating that desorption of Cd was not
time dependent over that time scale. In contrast, however, Amacher et al. (1986), in their
batch kinetic studies on Cd desorption in five soils, noted a rapid initial phase of desorp-
tion followed by a slower phase. However, they noted that although the overall reten-
tion/release reaction was not in equilibrium, the metal and soil reaction was almost
instantaneous.
On synthetic Fe and Mn oxides the desorption of Cd has been found to continue for
several days and may be diffusion controlled. Using synthetic goethite, Gerth et al. (1993)
found that extraction of metals (added at 10-6 M and sorbed during a reaction period of
21 d at 35C) with 0.7 M HN0
3
for 14 days at 35C released 89,72, and 60% Ni, Cd,
and Zn, respectively. This supported the observations of Brummer et al. (1988) that,
during the sorption process, metals become immobilized, possibly by diffusion into highly
specific binding sites in goethite micropores, which protect them against acid attack.
Recently, Backes et al. (1995) studied the desorption behavior of Cd and Co on Fe and
Mn oxide surfaces at 20C and compared the amounts desorbed from soils in contact
with metals for 1- and 15-week periods. In this study, the desorption of metals was in-
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 83
duced by continuous pumping of 0.0 1M Ca(N0
3
)2 through oxides at a flow rate of 2.5-
3 mLimin. The results showed a rapid and slower phase of desorption, the latter being
the dominant in terms of total desorption. The amount of Cd associated with the slower
phase increased with increased contact time between Cd and sorbent during sorption.
When sorption was allowed over 15 weeks, the amount of Cd desorbed (within 5 hr) was
found to be almost half of that desorbed when the sorption reaction time was only 1
week. Therefore, it appears that both the rate and extent of Cd release is dependent on
the nature of the sorbent and mechanism of sorption of Cd.
Sorption Reversibility in Flow-Through Experiments
A comparison of the sorption and desorption flanks of the breakthrough curves (BTC),
can provide clues about the reversibility of solute sorption during transport. For a solute
showing a linear sorption isotherm and symmetrical breakthrough curve, the desorption
flank of BTC when inverted and superimposed on the sorption flank should match if the
sorption is reversible. However, nonlinearity of sorption isotherm and kinetics of sorp-
tion-desorption reactions can influence the shape of sorption and desorption flanks of a
BTC and, hence, deviations on superimposition may be seen. In such cases it is impor-
tant is to assess the mass balance for the solute entering and eluting out of the column.
KookanCl. et al. (1994) conducted Cd transport experiments on an Oxisol to obtain both
sorption and desorption fronts of a BTC. When the sorption and desorption flanks of the
Cd BTCs were inverted and superimposed, some deviation between the two flanks of
the Cd BTC was noted, but the mass balance (the areas on the left of the two flanks of
the BTCs) were essentially the same during sorption and desorption phases (Naidu et
al., 1997). These results show that nearly all the Cd that was introduced into the soil was
recovered during desorption. Campbell et al. (1987) similarly carried out Cd sorption-
desorption studies on montmorillonite-humic acid mixtures using the miscible-displace-
ment technique. An examination of their BTCs and mass balance also shows that most of
bound Cd did elute from the columns.
Partial reversibility of sorption can influence the desorption flank of a BTC markedly,
as shown by Tran et al. (1998b). Column experiments were reported by Tran et al.
(1998b), in which a 1 mg/L Cd solution was introduced into short (approximately 5 cm)
columns containing homogeneous sand, which showed sorption desorption hysteresis
for Cd in batch experiments. After the sorption phase was complete (effluent concentra-
tion at 1 mg/L), the influent was switched to a Cd-free solution. The resultant break-
through curve is shown in Figure 3.12. Clearly the symmetry of the influent pulse is not
maintained in the breakthrough curve, the asymmetry reflecting the nonsingular or hys-
teretic sorption/desorption isotherm of Cd.
From the above discussion it is clear that Cd sorption in soils is not always reversible.
The reversibility is likely to be influenced by the nature of sorbent, pH, and composition
of soil solution. Currently the desorption process of Cd is poorly understood and war-
rants further research.
SUMMARY
Cd is sorbed by both specific and nonspecific interactions with soils, depending upon
the nature of mineral matter present in soil and soil solution composition. Similarly the
84 Fate and Transport of Heavy Metals in the Vadose Zone
;J'
bb
1.0
5 0.5
U
o . o ~ __ ...
o 20 40 60
Time (h)
80 100
Figure 3.12. Column effluent concentrations (e) showing adsorption then desorption of a Cd pulse
(Tran et aI., 1997b, with permission) at pH 6. The line is a model fit to the experimental data.
extent of Cd sorption is also influenced by the nature of the sorbent as well the composi-
tion of soil solution in terms of ionic strength, nature of competing and complexing ions.
Ca ions are particularly effective in competing with Cd for sorption sites even at lower
ionic strengths. It is therefore important to adequately account for such competition in
assessing the mobility of Cd in soils. Similarly, the presence of ligands, especially Cl, in
soil solution can markedly influence Cd sorption and mobility. Further research to im-
prove the understanding of the effects of organic ligands on Cd behavior is warranted.
In a heterogeneous and dynamic system such as the soil environment, several factors
together determine the nature and extent of retention reactions of Cd. While it is diffi-
cult to isolate the individual effects of various factors influencing Cd, some workers have
been able to develop relations which can quantitate the individual influence of Cd con-
centration, pH, and ionic strength of Ca, during sorption. Clearly, more efforts are needed
in this area to develop a better understanding of influences of various factors in multi-
variate systems.
Most research on Cd sorption desorption equilibria and kinetics has been carried out
in well-mixed batch systems. While such systems are easy, quick, and suitable for estab-
lishing the fundamental retention reactions, they do not always represent the realistic
conditions under which Cd mobility needs to be assessed; e.g., under flow conditions. Cd
sorption behavior has been found to be different in flow-through systems as compared to
batch systems. However, from the current work it is not possible to conclude whether
batch systems over- or underestimate the sorption of Cd. Clearly the sensitivity of Cd
sorption to several factors, and the fundamental differences between the techniques,
together with varying conditions in different experiments even with the same technique,
makes it very difficult to draw any meaningful conclusions.
Desorption of sorbed Cd from soil is probably more relevant for the assessment of its
mobility and potential adverse impact on the environment. However, it remains a poorly
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 85
understood phenomenon. From the limited work available in literature, it is concluded
that sorption of Cd is not always reversible. Indeed, the reversibility of Cd sorption
depends upon the nature of the sorbent as well as the desorbing solution. Cd interactions
with calcite, which is the essential component of calcareous soils and some aquifers, can
have significant implications for Cd mobility in the environment. It has been shown that
sorption of Cd followed by coprecipitation and dehydration on calcite can result in par-
tial reversibility and possible nonequilibrium conditions under flow conditions. Also, the
high pH in calcareous soils favors Cd retention and therefore its mobility is likely to be
limited in such soils. Ease of Cd desorption is likely to be linked to the binding affinity of
Cd to the sorbent and, therefore, under conditions where Cd shows feeble binding, its
sorption is likely to be reversible. Under such conditions, both sorption as well as des-
orption processes will favor greater mobility of Cd. Highly weathered acidic soils (e.g.,
some Australian Oxisols), or acidic sandy soils with inherently low cation exchange
capacities, are therefore likely to have greater availability of Cd in soil solution, which
may have implications for its plant uptake of leaching through the soil profile.
REFERENCES
AkratanakuL S., L. Boersma, and G.O. Klock. Sorption process in soils as influenced by pore
water velocity: II. Experimental results. SoiL Sci. 135, pp. 331-341, 1983.
Allison, J.D. and D.S. Brown. MINTEQA2/PRODEFA2-A Geochemical Speciation Model
and Interactive Preprocessor, in ChemicaL Equilibrium and Reaction Modeu, SSSA Special Pub-
lication 42, pp. 241-252. (Soil Science Society of America, Madison, WI), 1995.
Alloway, B.J. Cadmium, in Heavy Metau in Soiu. pp. 100-124. B.J. Alloway, Ed., John Wiley &
Sons, Inc., New York, 1990.
Al-Soufi, R.W. A method for simulating cadmium transport in soil: Model development and
experimental evaluation. J. Hydro!' 163, pp. 233-247, 1994.
Amacher, M.C., J. Kotuby-Amacher, H.M. Selim, and I.K Iskandar. Retention and release of
metals by soils-Evaluation of several models. Geoderma. 38, pp. 131-154, 1986.
Backes, C.A., R.G. McLaren, A.W. Rate, and R.S. Swift. Kinetics of cadmium and cobalt
desorption from iron and manganese oxides. SoiL Sci. Soc. Am. J. 59, pp. 778-785, 1995.
Bajracharya, K Transport of Cadmium in Soil. D. Eng. Thesis. Asian Institute of Technology,
Bangkok, Thailand, 1989.
Bajracharya, K and D.A. Barry. Accuracy criteria for linearised diffusion wave flood routing. J.
Hydro!' 195, pp. 200-217, 1997a.
Bajracharya, K and D.A. Barry. Nonequilibrium solute transport parameters and their physical
significance: Numerical and experimental results. J. Contam. HydroL. 24, pp. 185-204, 1997b.
Bajracharya, K, Y.T. Tran, and D.A. Barry. Cadmium adsorption at different pore water
velocities. Geoderma 73, pp. 197-216, 1996.
Barrow, N.J. Reactions with variable-charge soils. FertiLizer &d. 14, pp. 1-100, 1987.
Barry, D.A. and K Bajracharya. Optimised Muskingum-Cunge solution method for solute
transport with equilibrium Freundlich reaction. J. Contam. HydroL. 18, pp. 221-238, 1995.
Barry, D.A. and L. Li. Physical Basis of Nonequilibrium Solute Transport in Soil, in 15th
InternationaL CongrNJ of SoiL Science TranJactionJ, AcapuLco, Mexico, JuLy 10-16. International Soci-
ety of Soil Science & Mexico Society of Soil Science, 2a, pp. 86-105, 1994.
Barry, D.A. and G. Sposito. Application of the convection-dispersion model to solute transport
in finite soil columns. SoiL Sci. Soc. Am. J. 52, pp. 3-9, 1988.
86 Fate and Transport of Heavy Metals in the Vadose Zone
Bingham, F.T., G. Sposito, and J.E. Strong. The effect of sulfate on the availability of cadmium.
SoiL Sci. 141, pp. 172-177, 1986
Boekhold, A.E. and S.E.A.T.M. Van der Zee. Long-term effects of soil heterogeneity on cad-
mium behaviour in soil. J. Contam. HydroL. 7, pp. 371-390, 1991.
Boekhold, A.E. and S.E.A. T.M. Van der Zee. A scaled sorption model validated at the column scale
to predict cadmium contents in a spatially variable field soil. Soil Sci. 154, pp. 105-112, 1992.
Boekhold, A.E., E.J.M. Temminghoff, and S.E.A.T.M. Van der Zee. Influence of electrolyte
composition and pH on cadmium sorption by an acid soil. J. SoiL Sci. 44, pp. 85-96, 1993.
Boesten, J.J. T.1. and L.J. T. Van der Pas. Modeling adsorption/desorption kinetics of pesticides
in a soil suspension. SoiL Sci. 146, pp. 221-231, 1988.
Bolton, K.A., S. Sjoberg, and L.J. Evans. Proton binding and cadmium complexation constants
for a soil humic acid using a quasi-particle model. SoiL Sci. Soc. Am. J. 60, pp. 1064-1072, 1996.
Brummer, G., J. Gerth, and K.G. Tiller. Reaction kinetics of the adsorption and desorption of
Ni, Zn and Cd by goethite. I. Adsorption and diffusion of metals. J. SoiL Sci. 39, pp. 35-52,
1988.
Brusseau, M.L., Z. GerstL D. Augustijn, and P.S.C. Rao. Simulating solute transport in an
aggregated soil with the dual-porosity model: Measured and optimised parameter values. J.
HydroL. 163, pp. 187-193, 1994.
Brusseau, M.L., P.S.C. Rao, R.E. Jessup, and J.M. Davidson. Flow interruption: A method for
investigating sorption non-equilibrium. J. Contam. HydroL. 4, pp. 223-240, 1989.
Brusseau, M.L. and P.S.C. Rao. Sorption nonideality during organic contaminant transport in
porous media. CRC Crit. Rev. Environ. ControL 19, pp. 33-99, 1989.
Buchter, B., C. Hinz, M. Gfeller, and H. Fluhler. Cadmium transport in an unsaturated stony
subsoil monolith. Soil Sci. Soc. Am. J. 60, pp. 716-721, 1996.
Burgisser, C., A. Scheidegger, M. Borkovec, and H. Sticher. Transport and adsorption of
cadmium in columns. Deat. BOden!.:. GeddeLL. MilteiL. 66, pp. 283-286, 1991.
CampbelL G.D., H.F. Galcia, and P.W. Schindler. Binding of cadmium by montmorillonite-
humic acid mixtures: Miscible displacement experiments. AuAraL. J. SoiL Red. 25, pp. 391-403,
1987.
Chardon, W. Mobility of Cadmium in Soil (in Dutch). PhD Thesis. Agricultural University,
Wageningen, The Netherlands, 1984.
Christensen, T.H. Cadmium soil sorption at low concentrations: Effect of time, cadmium loading,
pH and calcium. Water Air SoiL PoLLut. 21, pp. 105-114, 1984a.
Christensen, T.H. Cadmium soil sorption at low concentrations: II. Reversibility, effect of
changes in solute composition, and effect of soil aging (die.). Water Air SoiL PoLLut. 21, pp. 115-
125, 1984b.
Christensen, T.H. Cadmium soil sorption at low concentrations: V. Evidence of competition by
other heavy metals. Water Air SoiL PoLLut. 34, pp. 293-303, 1987.
Davis, J.A. and J.O. Leckie. Surface ionization and complexation at the oxide/water interface.
J. CoLwiJ Interface Sci. 67, pp. 91-107, 1978.
Davis, J.A. Complexation of trace metals by adsorbed natural organic material. Geochimca et
COdmochimca Acta. 48, pp. 677-691, 1984.
Davis, J.A., C.C. Fuller, and A.D. Cook. A model for trace metal sorption process at the calcite
surface: Adsorption of Cd+
2
and subsequent solid solution formation. Geochimiea et COdmochimiea
Acta, 51, pp. 1477-1490, 1987.
Doner, H.E. Chloride as a factor in mobilities of Ni(I1), Cu(II) and Cd(II) in soil. SoiL Sci. Soc.
Am. J. 42, pp. 882--885, 1978.
Dowdy, R.H., J.J. LattereiL T.D. Hinesly, R.B. Grossman, and D.L. Sullivan. Trace metal
movement in an aeric Ochraqualf following 14 years of annual sludge applications. J. Environ.
QULlL 20, pp. 119-123, 1991.
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 87
Dowdy, RH. and V.V. Volk. Movement of Heavy Metals in Soils, in Chemical MObility and
Reactivity in Soil SY<ftel1k1, D.W. Nelson, Ed., Soil Science Society of America, Madison, Wis-
consin, 1983, pp. 229-240.
Dunnivant, F.M., P.M. Jardine, D.L. Taylor, and J.F. McCarthy. Co-transport of cadmium and
hexachlorobiphenyl by dissolved organic carbon through columns containing aquifer material.
Environ. Sci. Teclmol. 26, pp. 360-368, 1992.
Dzombak, D.A. and F.M.M. Morel. Sorption of cadmium on hydrous ferric oxide at high
sorbate/sorbent ratios: Equilibrium, kinetics and modeling. J. Colloid Interface Sci. 112, pp. 588-
598, 1986.
Elliott, H.A. and C.M. Denneny. Soil adsorption of cadmium from solutions containing organic
ligands. J. Environ. Qual. 11, pp. 658-662, 1982.
Farley, KJ., D.A. Dzombak, and F.M.M. Morel. A surface precipitation model for the sorption
of cations on metal oxides. J. Colloid Interface Sci. 106, pp. 226-242, 1985.
Forbes, E.A., A.M. Posner, and J.P. Quirk. The specific adsorption of divalent Cd, Co, Cu, Pb,
and Zn on goethite. J. Soil Sci. 27, pp. 154-166, 1976.
Fuller, C.C. and J .A. Davis. Processes and kinetics of Cd
2
+ sorption by a calcareous aquifer sand.
Geochimica et CO<fmochimca Acta. 51, pp. 1491-1502, 1987.
Gaber, H.M., S.D. Comfort, W.P. Inskeep, and H.A. EI-Attar. A test of local equilibrium
assumption for adsorption and transport of picloram. SoiL Sci. Soc. Am. J. 56, pp. 1392-1400,
1992.
Garcia-Miragaya, J. and A.L. Page. Influence of ionic strength and inorganic complex formation
on sorption of trace amounts of cadmium by montmorillonite. Soil Sci. Soc. Am. J. 40, pp. 658-
663, 1976.
Garcia-Miragaya, J. and A.L. Page. Influence of exchangeable cation on the sorption of trace
amounts of cadmium by montmorillonite. Soil Sci. Soc. Am. J. 41, pp. 718-721, 1977.
Garcia-Miragaya, J. and A.L. Page. Sorption of trace quantities of cadmium by soils with
different chemical and mineralogical composition. Water Air SoiL PoLLut. 9, pp. 289-299, 1978.
Gerritse, RG. Dispersion of cadmium in columns of saturated sandy soils. J. Environ. Qual. 25, pp.
1344-1349, 1996.
Gerth, J., G.W. Brummer, and KG. Tiller. Retention ofNi, Zn and Cd by Si-associated goethite.
Zeit<fchrift fur Pflanzenernahrung und Bodenlcunde 156, pp. 123-129, 1993.
Griffioen, J. W., D.A. Barry, and J.-Y. Parlange. Interpretation of the two-region model param-
eters. Water &<four. Re<f. 34, pp. 373-384, 1998.
Grolimund, D., M. Borkovec, P. Federer, and H. Sticher. Measurement of sorption isotherms
with flow-through reactors. Environ. Sci. Technol. 29, pp. 2317-2321, 1995.
Haas, C.1. and N.D. Horowitz. Adsorption of cadmium to kaolinite in the presence of organic
material. WaterAirSoiLPoLLut. 27, pp. 131-140, 1986.
Hachiya, K, M. Ashida, M. Sasaki, H. Khan, T. Inque, and T. Yasunaga. Study of the kinetics
of adsorption-desorption of Pb
2
+ on a y-A1203 surface by means of a relaxation technique. J.
Phy<f. Chem. 83, pp. 1866-1871, 1979.
Harter, R.D. and R. Naidu. Role of metal-organic complexation in metal sorption by soils. Adv.
Agron. 55, pp. 219-264, 1995.
Harter, RD. and R.G. Lehmann. Use of kinetics for the study of exchange reactions in soils. Soil
Sci. Soc. Am. J. 47, pp. 666-669, 1983.
Hayes, KF. and J.O. Leckie. Modeling ionic strength effects on cation adsorption at hydrous
oxide/solution interfaces. J. CoLloid Interface Sci. 115, pp. 564-572, 1987.
Hinz, C. and H.M. Selim. Transport of zinc and cadmium in soils: Experimental evidence and
modelling approaches. Soil Sci. Soc. Am. J. 58, pp. 1316-1327, 1994.
Hodgson, J.F., KG. Tiller, and M. Fellows. The role of hydrolysis in the reaction of heavy metals
with soil-forming materials. Proc. SoiL Sci. Soc. Am. 28, pp. 42-46, 1964.
88 Fate and Transport of Heavy Metals in the Vadose Zone
Holm, P.E., B.B.H. Anderson, and T.H. Christensen. Cadmium solubility in aerobic soils. Soil Sci.
Soc. Am. J. 60, pp. 775-780, 1996.
Homann, P.S. and RJ. Zasoski. Solution composition effects on cadmium sorption by forest soil
profiles. J. Environ. QuaL. 16, pp. 429-433, 1987.
Hornburg, V. and G.W. Brummer. Verhalten von Schwermetallen in Boden. 1. Untersuchungen
zur Schwermetallmobilitat. Zeitdchrift fur Pflanzenernahrung und BOdenkunde 156, pp. 467-477,
1993.
Huber, J.F.K and RG. Gerritse. Evaluation of dynamic gas chromatographic methods for the
determination of adsorption and solution isotherms. J. Chromatography 58, pp. 137-158, 1971.
Jardine, P.M. and D.L. Sparks. Potassium-calcium exchange in a multireactive soil systems. 1.
Kinetics. SoiL Sci. Soc. Am. J. 48, pp. 39-45, 1984.
Klamberg, H., G. Matthess, and A. Pekdeger. Organo-Metal Complexes as Mobility-Determin-
ing Factors of Inorganic Toxic Elements in Porous Media, in Inorganic Contaminantd in the
VadOde Zone, Ecological Studies, Vol. 74, B. Bar-Yosef, N.J. Barrow, and J. Goldsmith, Eds.,
Springer-Verlag, Berlin, pp. 3-17, 1989.
Kookana, RS. and R Naidu. Effect of Soil Solution Composition on Cadmium Transport
Through Variable Charge Soils, in Proceedingd Firdt InternationaL Conference Contaminantd and the
SoiL Environment-Feu. '96, AdelaUJe, AlldlraLia, pp. 29-30, 1996.
Kookana, RS. and R Naidu. Effect of soil solution composition on cadmium transport through
variable charge soils. Geoderma, 84, pp. 235-248, 1998.
Kookana, RS., R Naidu, and KG. Tiller. Sorption non-equilibrium during cadmium transport
through soils. Alldtrai. J. SoiL Ru. 32, pp. 635-651, 1994.
Kookana, RS., R Naidu, and KG. Tiller. Desorption of Cadmium is Determined by Its
Adsorption Affinity to Soils, in Proceedingd Fourth InternationaL Conference on the Biogeochemi.Jtry of
TraceElementd. Berkeley, CaLifornia, I.K Iskandar, S.E. Hardy, A.C. Chang, and G.M. Pierzynski,
Eds., 1997, pp. 399-400.
Kookana, RS., RD. Schuller, and L.A.G. Aylmore. Simulation of simazine transport through
soil columns using time-dependent sorption data measured under flow conditions. J. Contam.
HydroL. 14, pp. 93-115, 1993.
Krishnamurti, G.S.R, G. Cieslinski, P.M. Huang, and KC.J. Van Rees. Kinetics of cadmium
release from soils as influenced by organic acids: Implication in cadmium availability. J.
Environ. QuaL. 26, pp. 271-277, 1997.
Lieber, M., N.M. Perlmutter, and H.L. Frauenthal. Cadmium and hexavalent chromium in
Nassau County ground water. J. Am. Water Workd A<idOC. 56, pp. 734-740, 1964.
Lund, L.J., A.L. Page, and C.O. Nelson. Movement of heavy metals below sewage disposal
ponds. J. Environ. QuaL. 5, pp. 330-334, 1976.
Mayer, R Adsorptionsisothermen als RegelgroBen beim Transport von Schwermetalen in Boden.
Zeitdchrift fur Pjlanzenernahrung und BOdenkunde 141, pp. 11-28, 1978.
McLaughlin, M.J., L.T. Palmer, KG. Tiller, T.A. Beech, and M.K Smart. Increased soil salinity
causes elevated cadmium concentrations in field grown potato tubers. J. Environ. QuaL. 23, pp.
1013-1018, 1994.
McLaughlin, M.J., KG. Tiller, and M.K Smart. Speciation of cadmium in soil solutions of saline/
sodic soils and relationship with cadmium concentrations in potato tubers (Solanum tuuerodum
L.). Alldtrai. J. SoiL Ru. 35, pp. 183-198, 1997.
Miller, D.M., M.E. Sumner, and W.P. Miller. A comparison of batch- and flow-generated anion
adsorption isotherms. SoiL Sci. Soc. Am. J. 53, pp. 373-380, 1989.
Morisawa, S. and Y. Inoue. Mathematical models for estimating dynamic performance in the
saturated zone: Discussions on the mechanisms of groundwater pollution by cadmium. Proc.
Environ. Sanitation Eng. Ru. 21, pp. 43-55, 1985.
Sorption-Desorption Equilibria and Dynamics of Cadmium During Transport in Soil 89
Murali. V. and L.A.G. Aylmore. No-flow equilibrium and adsorption dynamics during ionic
transport in soils. Nature 283. pp. 467-469. 1980.
Naidu. R . N.J. deLacy. N.S. Bolan. RS. Kookana. and KG. Tiller. Effect of Inorganic Ligands
on Adsorption of Cadmium by Soils. in ProceedingJ of the 15th InternationaL Soil Science Society
Congrc.:l.:l, JuLy 1994. pp. 190-191. 1994b.
Naidu. R. N.S. Bolan. RS. Kookana. and KG. Tiller. Ionic strength and pH effects on surface
charge and Cd sorption characteristics of soils. J. SoiL Sci. 45. pp. 419-429. 1994a.
Naidu. R. RS. Kookana. M.E. Sumner. R.D. Harter. and KG. Tiller. Cadmium sorption and
transport in variable charge soils: A review. J. Environ. QuaL. 26. pp. 602-617. 1997.
Neal. RH. and G. Sposito. Effects of soluble organic matter and sewage sludge amendments on
cadmium sorption by soils at low cadmium concentrations. SoiL Sci. 142. pp. 164-172. 1986.
Nkedi-Kizza. P . J.W. Biggar. H.M. Selim. M.Th. van Genuchten. P.J. Wierenga. J.M. Davidson.
and D.R. Nielsen. On the equivalence of two conceptual models for describing ion exchange
during transport through an aggregated oxisol. Water ReJour. ReJ. 20. pp. 1123-1130. 1984.
Nkedi-Kizza. P .. J.W. Bigger. M.Th. van Genuchten. P.J. Wierenga. H.M. Selim. J.M. Davidson.
and D.R Nielsen. Modeling tritium and chloride 36 transport through an aggregated Oxisol.
Water ReJour. ReJ. 19. pp. 691-700. 1983.
OConnor. G.A .. C. OConnor. and G.R Cline. Sorption of cadmium by calcareous soils:
Influence of solution composition. SoiL Sci. Soc. Am. J. 48. pp. 1244-1247. 1984.
Parker. D.R. W.A. Norvell. and RL. Chaney. GEOCHEM-PC-A Chemical Speciation Program
for IBM and Compatible Personal Computers. in ChemicaL Equilibrium and Reaction Modeu. SSSA
Special Publication 42. Soil Science Society of America. Madison. WI. 1995. pp. 253-269.
Passioura. J.B. Hydrodynamic dispersion in aggregated media. Soil Sci. Ill, pp. 339-344. 1971.
PuIs. RW .. RM. Powell. D. Clark. and C.J. Eldred. Effect of pH. solid/solution ratio. ionic
strength. and organic acids on Pb anJ Cd sorption on kaolinite. Water Air SoiL PoLLut. 57-58. pp.
423-430. 1991.
Runnells. D.D.. T.A. Shepherd. and E.E. Angino. Metals in water: Determining natural back-
ground concentrations in mineralized areas. Environ. Sci. TechnoL. 26. pp. 2316-2323. 1992.
Schindler. P.W . P. Leichti. and J.C. Westall. Adsorption of copper. cadmium. and lead from
aqueous solutions to the kaolinite/water interface. NetherlandJ J. Agric. Sci. 35. pp. 219-230.
1987.
Schulin. R. P.J. Wierenga. H. Fliihler. and J. Leuenberger. Solute transport through a stony
soil. SoiL Sci. Soc. Am. J. 51. pp. 36-42. 1987.
Selim. H.M . B. Buchter. C. Hinz. and L. Ma. Modeling the transport and retention of cadmium
in soils: Multireaction and multicomponent approaches. SoiL Sci. Soc. Am. J. 56. pp. lO04-1015.
1992.
Selim. H.M. Prediction of contaminant retention and transport in soils using multireaction
models. Environ. HeaLth PerJpect. 39. pp. 69-75. 1989.
Shimojima. E. and M.L. Sharma. The influence of pore water velocity on transport of sorptive
and non-sorptive chemicals through an unsaturated sand. J. HydroL. 164. pp. 239-261. 1995.
Sidle. R.C. and L.T. Kardos. Transport of heavy metals in a sludge-treated forested area. J.
Environ. QuaL. 6. pp. 431-437. 1977.
Sidle. RC . L.T. Kardos. and M.Th. van Genuchten. Heavy metals transport model in a sludge-
treated soil. J. Environ. QuaL. 6. pp. 438-442. 1977.
Skopp. J. Analysis of time-dependent chemical processes in soils. J. Environ. QuaL. 15. pp. 205-
213. 1986.
Stipp. S.L.. M.F. Hochella. G.A. Parks. and J.O. Leckie. Cd2+ uptake by calcite. solid-state
diffusion. and the formation of solid-solution: Interface processes observed with near-surface
sensitive techniques (XPS. LEED. and AES). Geochimica et COJmochimca Acta. 56. pp. 1941-
1954. 1992.
90 Fate and Transport of Heavy Metals in the Vadose Zone
Streck. T. and J. Richter. Heavy metal displacement and in a sandy soil at the field scale: 1.
Measurements and parameterization of sorption. J. Environ. QuaL. 26. pp. 49-56. 1997a.
Streck. T. and J. Richter. Heavy metal displacement and in a sandy soil at the field scale: II.
Modeling. J. Environ. QuaL. 26: pp. 56-62. 1997b.
Sugita. F. and RW. Gillham. Pore scale variation in retardation factor as a cause of nonideal
reactive breakthrough curves. 3. Column investigations. Water Redour. Red. 31. pp. 121-128.
1995.
Tan. T.C. and W.K Teo. Combined effect of carbon dosage and initial adsorbate concentration
on the adsorption isotherm of heavy metals on activated carbon. Water Red. 21. pp. 1183-1188.
1987.
Temminghoff. E.J.M . S.E.A.T.M. Van der Zee. and F.A.M. De Haan. Speciation and calcium
competition effects on cadmium sorption by sandy soil at various pHs. European J. SoiL Sci. 46.
pp. 649-655. 1995.
Theis. T.L.. R Iyer. and L.W. Kaul. Kinetic studies of cadmium and ferricyanide adsorption on
goethite. Environ. Sci. Tec/moL. 22. pp. 1013-1017. 1988.
Tiller. KG. Heavy metals in soils and their environmental significance. Adv. SoiL Sci. 9. pp. 113-
142. 1989.
Tiller. KG. Soil Contamination Issues: Past. Present and Future. A Personal Perspective. in
ContaminantdandtheSoiLEnvironmentintheAuAralMiaPacificRegion. R Naidu. RS. Kookana. D.P.
Oliver. S. Rogers. and M.J. McLaughlin. Eds .. Kluwer Academic Publishers. 1996. pp. 1-28.
Tiller. KG . J. Gerth. and G. Brummer. The sorption of Cd. Zn. and Ni by a soil clay fraction:
Procedures for partition of bound forms and their interpretation. Geoderma. 34. pp. 1-16. 1984a.
Tiller. KG .. J. Gerth and G. Brummer. The relative affinities of Cd. Ni and Zn for different clay
fractions and goethite. Geoderma. 34. pp. 17-35. 1984b.
Tiller. KG . V.K Nayyar. and P.M. Clayton. Specific and non-specific sorption of cadmium by
soil clays as influenced by zinc and calcium. AwtraL. J. SoiL Red. 17. pp. 17-28. 1979.
Tran. Y.T . K Bajracharya. and D.A. Barry. Anomalous cadmium sorption in flow interruption
experiments. Geoderma, 84. pp. 169-184. 1998a.
Tran. Y.T . D.A. Barry. and K Bajracharya. Cadmium desorption in sand. Environ. SCt: TechnoL.
(in review). 1998b.
Valocchi A.J. Validity of the local equilibrium assumption for modelling sorbing solute through
homogenous soils. Water Redoar. Red. 21. pp. 808--820. 1985.
van Genuchten. M.Th. and F.N. Dalton. Models for simulating salt movement in aggregated field
soils. Geoderma 38. pp. 165-183. 1986.
van Genuchten. M.Th. and J.C. Parker. Boundary conditions for displacement experiments
through short laboratory soil columns. SoiL Sci. Soc. Am. J. 48. pp. 703-708. 1984.
Zachara. J.M. and S.C. Smith. Edge complexation reaction of cadmium on specimen and soil-
derived smectite. SoiL Sci. Soc. Am. J. 58. pp. 762-769. 1994.
Zachara. J.M . C.E. Cowan. and C.T. Resch. Sorption of divalent metals on calcite. Geochimica
et COdmochimica Acta. 55. pp. 1549-1562. 1991.
Zachara. J.M . S.C. Smith. C.T. Resch. and C.E. Cowan. Cadmium sorption to soil separates
containing layer silicates and iron and aluminum oxides. SoiL Set: Soc. Am. J. 56. pp. 1075-1084.
1992.
Zachara. J.M .. S.C. Smith. J.P. McKinley. and C.T. Resch. Cadmium sorption on specimen and
soil smectites in sodium and calcium electrolytes. Soil Sci. Soc. Am. J. 57. pp. 1491-1501. 1993.
Zasoski. RG. and RG. Burau. A technique for studying the kinetics of adsorption in suspen-
sions. Soil Sci. Soc. Am. J. 42. pp. 372-374. 1978.
Zhu. B. and A.K Alva. Differential adsorption of trace metals by soils as influenced by exchange-
able cations and ionic strength. SoiL Sci. 155. pp. 61-66. 1993.
CHAPTER ~
Modeling the Kinetics of Heavy Metals
Reactivity in Soils
H. Magdi Selim
INTRODUCTION
The reactivity, toxicity, and mobility of heavy metal in the soil system is of major
environmental concern. Thus, understanding of the complex interactions of heavy met-
als in the environment is a prerequisite in the effort to predict their behavior in the
vadose zone. To predict the fate and transport of heavy metals in the soil, models that
include retention and release reactions of solutes with the soil matrix are needed. Reten-
tion and release reactions in soils include ion exchange, adsorption/desorption, precipi-
tation/dissolution, and other mechanisms such as chemical or biological transformations.
Adsorption is the process where solutes bind or adhere to soil matrix surfaces to form
outer- or inner-sphere solute surface-site complexes.
Over the last 30 years, several models describing the retention behavior of heavy
metals in the soil environment have been developed (see Selim et aI., 1990, and Selim,
1992 for a review). Table 4.1 provides a listing of commonly used equilibrium and ki-
netic retention approaches. A disadvantage of several of such empirical approaches lies
in the basic assumption of local equilibrium of the governing reactions. Alternatives to
equilibrium-based approaches are the multisite or multireaction models which deal with
the multiple interactions of one solute species in the soil environment. Multiple interac-
tion approaches assume that a fraction of the total sites is time dependent; i.e., kinetic in
nature whereas the remaining fraction interacts rapidly or instantaneously with that in
the soil solution. Nonlinear equilibrium (Freundlich) and first- or nth-order kinetic re-
actions are the associated processes (for a review see Selim, 1992). Another two-site
approach was proposed by Theis et aI. (1988) for Cd mobility and adsorption on goe-
thite. They assumed that the nature of reactions for both sites was governed by second-
order kinetic reactions. The reactions were assumed to be consecutive where the second
reaction was irreversible in nature. Amacher et aI. (1988) developed a multireaction
model that includes concurrent and concurrent-consecutive processes of the nonlinear
91
92 Fate and Transport of Heavy Metals in the Vadose Zone
Table 4.1. Selected Equilibrium and Kinetic Type Models for Heavy Metal
Retention in Soils
Model
Equilibrium Type
Linear
Freundlich
General Freundlich
Rothmund-Kornfeld ion exchange
Langmuir
General Langmuir-Freundlich
Langmuir with sigmoidicity
Kinetic Type
First-order
nth-order
Irreversible (sink/source)
Second-order irreversible
Langmuir kinetic
Elovich
Power
Mass transfer
Formulation
a
S=K.JC
S = K.J C
b
S/S
max
= [w CI(l + w C ) ) ~
Sj/ST = ~ K (C/CT)n
S/S
max
= w CI[ 1 + we]
S/S
max
= (w C ) ~ 1[1 + ( w C ) ~ )
S/S
max
= w CI[l + wC + oK)
aS/at = k
f
(Sip) C - kb S
aS/at = k
f
(Sip) C
n
- kb S
aS/at = ks (Sip) (C - Cpl
aS/at = ks (Sip) C (Smax - S)
aS/at = k
f
(Sip) C (Smax - S) - kb S
aS/at = A exp(-BS)
aS/at = K (Sip) C
n
Sm
aS/at = K (Sip) (C - CO)
a A, B, b, C*, C
P
' K, K.J, ~ K ' kb' kf' k" n, m, Smax, W, /3, and (J are adjustable model
parameters, p is bulk density, S is volumetric soil water content, C
T
is total
solute concentration and ~ is total amount sorbed of all competing species.
kinetic type. The model was capable of describing the retention behavior of Cd and
Cr(VI) with time for several soils. Selim et al. (1992) developed a multicomponent ap-
proach which accounts for an equilibrium ion exchange and a specific sorption process
based on a second-order (Langmuir) kinetic reaction. The multicomponent model ad-
equately predicted the observed breakthrough results where a pronounced snowplow or
chromatographic effect was observed. Eff1uent peak concentrations were 3-5 fold that
of the input Cd pulse.
In this chapter, we focus on major features of kinetic retention processes and model-
ing approaches of heavy metals in the vadose zone. Retention approaches of the linear
and nonlinear reversible and irreversible kinetic types are first discussed. Models of the
multiple-reaction type, including the two-site equilibrium-kinetic models, and ion-ex-
change multicomponent models are also presented. Illustrative examples of Cu and Cd
isotherms and breakthrough results (BTCs) are discussed.
LINEAR RETENTION
For several heavy metals (e.g, Cr, Cu, Zn, Cd, and Hg), retention and release reac-
tions in the soil solution have been observed to be strongly time-dependent. Recent stud-
ies on the kinetic behavior of the fate of several heavy metals include Aringhieri et al.
(1985), and Amacher et al. (1986) among others. A number of empirical models have
been proposed to describe kinetic retention and release reactions of solutes in the solu-
tion phase. The first-order kinetic approach is perhaps the earliest single (linear) form of
reaction used to describe time-dependent sorption which may be expressed as
Modeling the Kinetics of Heavy Metals Reactivity in Soils 93
as
P a;- = kf e c - kb P S
(1)
where C is the heavy metal concentration in solution (llg/mL), S is the amount of heavy
metal sorbed or retained by the soil (Ilg/g soil), P is the soil bulk density (g/cm
3
) , e is the
volumetric soil-water content (cm
3
/cm
3
), and t is time (h). Reaction given in Equation 1
is fully reversible between the heavy metal species present in solution and that sorbed by
soil matrix surfaces where k
f
and kb represent the forward and backward rate coeffi-
cients (h-
1
). The first-order reaction was first incorporated into the classical (convec-
tive-dispersive) transport equation by Lapidus and Amundson (1952) to describe solute
retention during transport under steady water flow conditions.
Integration of the first-order reaction (Eq. 1) subject to initial conditions of C = C
i
'
and S = 0 at t = 0, yields a system of linear sorption isotherms. That is, for any reaction
time t, a linear relation between Sand C is obtained. However, linear isotherms are not
often encountered except for selected cations and heavy metals at low concentrations
(see Selim et aI., 1992). In contrast, nonlinear retention behavior is commonly observed
for several heavy metals as depicted by the nonlinear isotherms for CU for Cecil and
McLaren soils shown in Figure 4.1.
NONLINEAR RETENTION
As a consequence of nonlinear retention behavior such as depicted in Figure 4.1, the
single reaction given by Equation 1 has been extended to include the nonlinear kinetic
form,
(2)
where m is a dimensionless parameter commonly less than unity and represents the or-
der of the nonlinear reaction. For both single kinetic forms (Eqs. 1 and 2), the magnitude
of the rate coefficients dictates the extent of the kinetic behavior of the reaction. For
small values of k
f
and kb' the rate of retention is slow, and strong kinetic dependence is
anticipated. In contrast, for large values of k
f
and K
b
, the retention reaction is a rapid one
and should approach quasi-equilibrium in a relatively short time. In fact, at large times
(t ~ 00) equilibrium is attained and the rate of retention (aSia t) approaches zero and the
above equation yields
where
(3)
which is an analogous form to the Freundlich equilibrium equation where Kd is the dis-
tribution coefficient (cm
3
/g). Therefore, for linear or Freundlich isotherms, one may
regard Kd as the ratio of the rate coefficient for sorption (forward reaction) to that for
desorption or release (backward reaction).
94 Fate and Transport of Heavy Metals in the Vadose Zone
------- - - - - - - - - ~ - - - - - -
500
o 2 hours
4
~
0 400

12
(/)
.," 0"
~
Ol
300
6
"0
<D
.0
....
200
0
(/)
"0

::J
100
t
()
Cu Adsorption
McLaren Soil
0
0 10 20 30 40 50 60 70
250
o 2 hours
~
12
0
200
(/)
~
..
Ol
=t 150
.......-
"0
<D
.0
....
100
0
(/)
"0

::J
50
()
Cu Adsorption
Cecil Soil
0
0 20 40 60 80 100
Cu in Solution (llg/mL)
Figure 4.1. Cu adsorption isotherms at several reaction times in Cecil and McLaren soils.
LANGMUIR OR SECOND-ORDER KINETICS
Based on this approach, heavy metal retention mechanisms are assumed to follow a
second-order kinetic type reaction where the forward process is controlled by the prod-
uct of the solution concentration C (mg L-
1
) and the amount of unoccupied or unfilled
sites (<1 (Selim and Amacher, 1988). Specifically, the reaction may be expressed by the
following reversible process:
(4)
Therefore, the kinetic rate equations for heavy metal retention may be expressed by the
following kinetic Langmuir (or second-order) equation:
Modeling the Kinetics of Heavy Metals Reactivity in Soils 95
where k
f
and kb (h-
1
) are forward and backward rate coefficients for the retention sites.
If <I> is unity, the above equation yields a first-order kinetic retention reaction (Lapidus
and Amundson, 1952). However, a major disadvantage of first-order kinetic reactions is
that as the concentration in solution increases, a maximum heavy metal sorption is not
attained, which implies that there is an infinite heavy metal retention capacity of the soil or
infinite amount of exchange sites on matrix surfaces. In contrast, the second-order ap-
proach achieves maximum sorption when all unfilled sites become occupied (i.e., <I> ~ 0).
We recognize that the unfilled or vacant sites (<I are not strictly vacant. They are occu-
pied by hydrogen, hydroxyl, or other nonspecifically (e.g., Na, Ca, Cl, and N0
3
) or
specifically (e.g., P0
4
, As0
4
, and transition metals) adsorbed species. Vacant or unfilled
refers to vacant or unfilled by the specific solute species of interest. The process of occu-
pying a vacant site by a given solute species actually is one of replacement or exchange of
one species for another. However, the simplifying assumption on which this model is
based is that the filling of sites by a particular solute species need not consider the corre-
sponding replacement of species already occupying the sites.
At large reaction times, i.e., as t ~ =, one can assume that the sites achieve local
equilibrium. As a result, Equation 5 yields the following expression,
(6)
where w represents the equilibrium constant for the reaction associated with the reten-
tion sites. Moreover, this form (Eq. 6) is analogous to expressions for homovalent ion-
exchange equilibrium reactions. In this sense, the equilibrium constant w resembles the
selectivity coefficients for exchange reactions of soil matrix surfaces (Sposito, 1984).
However, a major difference between ion-exchange and the proposed second-order ap-
proach is that no consideration of other competing ions in solution or on matrix surfaces
is incorporated into the second-order rate equations. In a strict thermodynamic sense,
the above formulation should be expressed in terms of activities rather than concentra-
tions. However, we use the implicit assumption that solution phase ion activity coeffi-
cients are constant in a constant ionic strength medium. Further rearrangement of
Equation 6 yields the following expression at t ~ 00,
S wC
------
(7)
which is the widely recognized Langmuir isotherm equation. This formulation is the
oldest and most commonly encountered in soils. It was developed to describe the adsorp-
tion of gases by solids where a finite number of adsorption sites in the surface is as-
sumed. The terms wand Smax are adjustable parameters. Here W (mL Ilg-
1
) is a measure
of the bond strength of molecules on the matrix surface and Smax (Ilg g-l of soil) is the
maximum sorption capacity or total amount of available sites per unit soil mass. In an
96 Fate and Transport of Heavy Metals in the Vadose Zone
attempt to classifY the various shapes of sorption isotherms, it was recognized that the
Langmuir isotherm is the most commonly used and is referred to as the L-curve iso-
therm (Sposito, 1984). Moreover, Langmuir isotherms were used successfully to de-
scribe Cd, Cu, Pb, and Zn retention in soils. Figure 4.2 shows experimental and fitted
isotherm examples of the use of the Langmuir equation to describe Cu retention in Cecil
and McLaren soil after 192 h of reaction.
HYSTERESIS
Adsorption-desorption hysteretic behavior has been observed by several scientists.
Examples of hysteretic behavior for Cu adsorption-desorption for a McLaren soil is
shown in Figure 4.3. Cu desorption shows significant hysteresis or nonsingularity be-
havior which was apparent at high initial concentrations. Based on the hysteresis behav-
ior for several initial concentrations (not shown), desorption results suggest that part of
the adsorbed Cu was not easily desorbed or becomes nondesorbable by forming strong
interaction with the soil matrix. Selim et al. (1976) showed that singularity or hysteresis
may be partly the result from failure to achieve equilibrium during adsorption and des-
orption. If adsorption as well as desorption were carried out for times sufficient for equi-
librium to be attained, or if the kinetic rate coefficients were sufficiently large, such
hysteretic behavior would be minimized.
IRREVERSIBLE REACTIONS
Irreversible processes of various heavy metals in the soil account for various (sink!
source) reactions, including precipitation/dissolutions, mineralization, immobilization,
biological transformations, volatilization and radioactive decay, among others. First-or-
der kinetic reactions have been utilized to quantifY the irreversible retention processes
by several authors. Models that account for first-order kinetic and sequential first-order
(irreversible) decay reactions include Rasmuson and Neretienks (1981) and Amacher et
al. (1988). The first-order irreversible retention form is
(8)
where k is the irreversible rate of reaction (h-'). This mechanism has been used to
lrr
describe various heavy metals adsorption/precipitation, and radio-nuclide decay. De-
scription of precipitation reactions that involve secondary nucleation is not an easy task,
and it is often difficult to distinguish between precipitation and adsorption.
SPECIFIC SORPTION
It has been postulated that specific sorption on high affinity sites may be regarded as
an irreversible process. A second-order approach was modified to describe irreversible
or weakly reversible retention when the rate of desorption or release is small, i.e., when
the backward rate coefficient kb approaches zero (see Selim et al., 1992). Therefore, a
second-order irreversible process can be expressed as
Modeling the Kinetics of Heavy Metals Reactivity in Soils 97
500 . - - - - - - - - - - - - - - - - - - - - - - - - - - - _ ~ - ~ - ~
---
~ 400
E
-
Cl 300
w
CD
a: 200
o
(/)
:J 100
o
---
&---
--McLaren

Cecil
20 40 60 80
Cu CONCENTRATION (mg/L)
Figure 4.2. Retention isotherms for Cu after 8 days of reactions for Cecil and McLaren soils. Solid
curves are calculated isotherms using equilibrium Langmuir model.
5 0 0 . - - - - - - - - - - - - - - - - - - - - - ~
:=-
o 400
(J)
~
-
"0
Q)
.0
300
(; 200
(J)
~
:J
o
100
Desorption
r ~ - -----------
, -'
I //
Adsorption
o ,
,.'


Cu - McLaren Soil
10 20 30 40 50
Cu in Solution (J.Lg/mL)
60
Figure 4.3. Cu adsorption and desorption isotherms for McLaren soil.
as
pa;-=kf e (ST -S)C (9)
where only two parameters, ST and kf' are required to account for irreversible retention.
The term ST represents the total amount of specific sites (J.Lg g-l of soil). For several
metal ions including Cd, Ni, Co, and Zn, specific sorption has been shown to be depen-
dent on time of reaction. Therefore, the use of a kinetic rather than an equilibrium sorp-
tion mechanism is recommended. A major advantage of the formulation of irreversible
reaction (Equation 7) is that a sorption maximum is achieved when all unfilled sites
become occupied (i.e., S ~ ST)' In contrast, for the first-order type Equation 6, as metal
ion concentration increases, an irreversible sorption maximum is not attained.
98 Fate and Transport of Heavy Metals in the Vadose Zone
MULTIPLE RETENTION
Several studies showed that retention of heavy metals in several soils was not ad-
equately described by use of a single reaction of the equilibrium or kinetic types. The
inadequacy of single reactions is not surprising since they only describe the behavior of
one species with no consideration to the simultaneous reactions of others in the soil
system. Multicomponent models consider a number of processes governing several spe-
cies in the soil solution including ion exchange, complexation, dissolution/precipitation,
and competitive adsorption. However, multicomponent models often consider the local
equilibrium assumption (LEA) to be valid. On the other hand, multisite or multireaction
models deal with the multiple interactions of one species in the soil environment. Such
models are empirical in nature and based on the assumption that a fraction of the total
sites are highly kinetic, whereas the remaining fraction of sites interact slowly or instan-
taneously with that in the soil solution (Selim et al., 1976; Jardine et al., 1985). Non-
linear equilibrium (Freundlich) and first- or nth-order kinetic reactions were the associated
processes.
The two-site approach proved successful in describing observed extensive tailing of
breakthrough results and has been used by several scientists including Jardine et al.
(1985), Parker and Jardine (1986), among others. The model proved successful in de-
scribing the retention and transport of several dissolved chemicals including aluminum,
phosphorus, potassium, cadmium, chromium, and methyl bromide. However, there are
several inherent disadvantages of the two-site model. First, the reaction mechanisms are
restricted to those that are fully reversible. Moreover, the model does not account for
possible consecutive type solute interactions in the soil system. Another two-site ap-
proach was proposed by Theis et al. (1988) for Cd mobility and adsorption on goethite.
They assumed that the nature of reactions for both sites to be governed by second-order
kinetic reactions. The reactions were assumed to be consecutive where the second reac-
tion was irreversible in nature. Other empirical approaches include the consecutive (equi-
librium-kinetic) model of Barrow and Shaw (1979). Here an adsorbed (surface) phase
was assumed to be in direct equilibrium with that in the solution phase and slowly (and
reversibly) with an absorbed (internal) phase within the soil matrix. Another approach
is that of van der Zee and van Riemsdijk (1986), in which a reversible reaction for P
sorption-desorption was governed according to Langmuir kinetic model (Eq. 5). In ad-
dition, precipitation reaction was accounted for as a (kinetic) diffusion process of P
through a thin layer of metal phosphate coating that surrounds metal oxides. Metal ox-
ide is assumed to be converted to metal phosphate in the reaction zone by a precipita-
tion-like reaction.
Amacher et al. (1988) and Selim et al. (1989) proposed a simplified model which
accounts for multiple reactions of solutes during transport in soils. In addition to the
soil solution phase (C) of a solute in the soil, four other phases representing solute
retained by the soil matrix (Se' S\, S2' S3' and Sire> were also considered. A schematic
of the multireaction model is shown in Figure 4.4. We assume Se to be governed by an
equilibrium Freundlich reaction, whereas S\ and S2 were governed by nonlinear ki-
netic reactions,
(10)
Modeling the Kinetics of Heavy Metals Reactivity in Soils 99
Figure 4.4. Schematic diagram of the multi reaction model.
aS
l
_ k e C
n
- k S
- 1 - 2 1
at p
(11)
(12)
where kl to k6 are the associated rates coefficients (h-
1
). These two phases (Sl and S2)
may be regarded as the amounts sorbed on surfaces of soil particles and chemically bound
to Ai and Fe oxide surfaces or other types of surfaces.
Amacher et al. (1988) pointed out that it is not necessary to have a prior knowledge of
the exact retention mechanisms for these reactions to be applicable. Moreover, these
phases may be characterized by their kinetic sorption and release behavior to the soil
solution and thus are susceptible to leaching in the soil. In addition, the primary differ-
ence between these two phases not only lies in the difference in their kinetic behavior but
also on the degree of nonlinearity as indicated by the parameters nand m. The consecu-
tive reaction between S2 and S3 represents slow reaction as a result of further rearrange-
ments of solute retained on matrix surfaces. Incorporation of S3 in the model allows the
description of the frequently observed very slow release of solute from the soil. As a
result, this strongly retained phase was represented by
(13)
The multireaction model also considers irreversible solute removal via a retention sink
term Q in order to account for irreversible reactions such as precipitation/dissolution,
mineralization, and immobilization, among others. We expressed the sink term as a first-
order kinetic process in a similar fashion to that given by Equation 6, where k
irr
is the
associated rate coefficient (h-
1
).
The multireaction model (Eqs. 6 to 10) was incorporated into the classical convective-
dispersion transport equation (CDE) which can be expressed as (Selim et al., 1989)
100 Fate and Transport of Heavy Metals in the Vadose Zone
(14)
where D is the hydrodynamic dispersion coefficient (cm
2
h-
I
), v is Darcy's water flux
density (cm h-
I
), and z is soil depth (cm). The term R is dimensionless, referred to here
as the Freundlich retardation coefficient which is dependent on C,
(15)
For the linear case where the exponent b in Equation 7 is unity, the well-known retarda-
tion factor R is obtained
(16)
which is constant. The multireaction model was capable of describing the kinetic reten-
tion behavior of P, Cd, Cr, Zn, and Hg based on batch data sets for several soils (see
Amacher et al., 1988; Selim et aI., 1990).
ION EXCHANGE RETENTION
This mechanism is commonly considered as a rapid reaction of the nonspecific sorp-
tion type. The mechanism is a fully reversible reaction between ions in the soil solution
and those retained on charged matrix surfaces. The exchange reaction for two compet-
ing ions i and j, having valencies Vi and Vj' respectively, may be written as
T
K
IJ
(17)
where TK;j denotes the thermodynamic equilibrium constant and a and a" (omitting the
subscripts) are the ion activity in soil solution and on the exchanger surfaces, respec-
tively. For the case of a binary homovalent ion, a generic selectivity coefficient K;j (Rubin
and James, 1973) or a separation factor for the affinity of ions on exchange sites is often
used. Examples of calculated and measured homovalent ion exchange isotherms are il-
lustrated in Figure 4.5 for Cd-Ca for two soils (Selim et aI., 1992). For :KcJCa > 1, sorp-
tion of ion Cd is preferred and the isotherms are convex, whereas for K0ICa < 1, sorption
affinity is apposite and the isotherms are concave. The capability of the ion exchange
approach in describing multiple pulse applications is illustrated in Figure 4.6. Here three
input pulses of Cd having a concentration (Co) of 10 mg L -I were consecutively applied
to a Windsor soil column (Selim et aI., 1992). The ion exchange model well predicted the
position of the BTC peaks. Moreover, the assumption of equilibrium ion exchange with
the selectivity coefficient (:KcJCa) based on Figure 4.5 adequately predicts the observed
Modeling the Kinetics of Heavy Metals Reactivity in Soils 101
>< 1.0 ,....----------.-----'::;00
-
c
w
al
a: 0.8
o
en
"C
o
LL 0.6
o
z
o
t; 0.4

a:
LL
w 0.2
...J
o
:E
Cd-Ca ISOTHERM
- Kc.tc.= 1
- - Kc.tc. = 0.5
_.- !<cdC. = 2
WINDSOR SOIL
o EUSTIS SOIL
........
o 0.2 0.4 0.6 0.8 1.0
Cd CONCENTRATION (C/C
o
)
Figure 4.5. Cadmium-calcium exchange isotherm for Windsor and Eustis soils. Solid and dashed
curves are simulations using different selectivities (I<cdca)'
8
6
2
o
20
Cd - WINDSOR (Co = 10 mg/L) I
:
.....

70
io
'(- .......
120
VNo
...
t




j
t-...
. .
170 220
Figure 4.6. Measured (closed circles) and predicted breakthrough curves in Windsor soil column
for three Cd pulses of Co = 10 mg L -1. The solid curve is prediction using equilibrium ion exchange.
chromatographic effect where the effluent concentrations exceed that of the input pulse
solution (C/C
o
> 1). Selim et al. (1992) obtained equally good predictions for multiple
input pulse applications with Co of 100 mg L-
I
.
The assumption of equilibrium ion exchange reaction has been employed to describe
sorption of heavy metals in soils by several investigators (Abd-Elfattah and Wada, 1981;
Harmsen, 1977; Bittel and Miller, 1974; Selim et al., 1992; Hinz and Selim, 1994). In
general, the affinity of heavy metals increase with decreasing heavy metal fraction on
exchanger surfaces. Using an empirical selectivity coefficient, it was shown that Zn af-
finity increased up to two orders of magnitude for low Zn surface coverage in a Ca
background solution (Abd-Elfattah and Wada, 1981). The Rothmund-Kornfeld approach
incorporates variable selectivity based on the amount adsorbed (s) or exchanger com-
102 Fate and Transport of Heavy Metals in the Vadose Zone
position. The approach is empirical and provides a simple equation that incorporated the
characteristic shape of binary exchange isotherms as a function of Sj as well as the total
solution concentration in solution (Cr). Harmsen (1977) and Bond and Phillips (1990)
expressed the Rothmund-Kornfeld as
(Sir: = R
K
. [(Cir:]n
(sY' 'J (C.)Vl
J J
(18)
where n is a dimensionless empirical parameter associated with the ion pair i-j, and R ~ j
is the Rothmund-Kornfeld selectivity coefficient. The above equation is best known as a
simple form of the Freundlich equation which applies to ion exchange processes. As
pointed out by Harmsen (1977), the Freundlich equation may be considered as an ap-
proximation of the Rothmund-Kornfeld equation valid for Sj Sj and Cj Cj' where
s. = RK.(c.)n
, 'J'
(19)
Based on best-fit predictions, Hinz and Selim (1994) showed strong Zn and Cd affinity
(compared to Ca) at low Zn and Cd concentrations based on parameter estimates of ion-
exchange isotherms using the Rothmund-Kornfeld approach.
Kinetic Ion Exchange
Recently, Sparks (1989) compiled an extensive list of cations (and anions) that exhib-
ited kinetic ion exchange behavior in soils; e.g., AI, NH
4
, K, and several heavy metal
cations. According to Ogwada and Sparks (1986a,b,c), kinetic ion exchange behavior
was probably due to mass transfer (or diffusion) and chemical kinetic processes. The
proposed approach was analogous to mass transfer or diffusion between the solid and
solution phase such that, for ion species i,
as 0
-' = a(s - s,.)
at '
(20)
where at any time t, the symbol Sj denotes the amount sorbed, where SjG is the amount
sorbed at equilibrium, and a is an apparent rate coefficient Cd-I) for the kinetic-type
sites. In Equation 17, the amount sorbed at equilibrium SjG is calculated using the respec-
tive isotherm relations similar to Equation 14. Expressions similar to Equation 17 have
been used to describe mass transfer between mobile and immobile water as well as chemi-
cal kinetics (Parker and Jardine, 1986). For large a, Sj approaches SjG in a relatively
short time and equilibrium is rapidly achieved, whereas for small a, kinetic behavior
should be dominant for extended period of time.
Case Study
We investigated Cu retention by monitoring its concentration in the soil solution with
time for a wide range of input concentrations. Cu transport was studied using miscible
Modeling the Kinetics of Heavy Metals Reactivity in Soils 103
displacement methods. Two different soils were studied. Cecil soil was chosen as a bench-
mark and McLaren soil was obtained from a site near an abandoned Cu mine on Fisher
Mountain, Montana. Isotherms were measured using standard ion exchange methods
where ten-gram samples of soil were equilibrated with Cu and Mg at varying ratios. The
samples were shaken for 24 h on a reciprocal shaker with 30 mL of various proportions
of CUS04 and MgS04 solutions. The solutions were then centrifuged and decanted. For
the first two steps (24 h each step) total concentration was 0.5 N followed by four time
steps at 0.01 N. Triplicate samples were used for each solution ratio. Adsorbed cations
were removed by three extractions with 1 N NaOAc and corrections were made for the
entrained solution. Cu and Mg in solution and extractant solution were analyzed by
ICP. Based on these Cu-Mg exchange isotherm experiments, selectivity for Cecil and
McLaren soils were obtained. The transport of Cu in the Cecil and McLaren soils were
investigated using the miscible displacement technique. Plexiglas columns (6.4 cm i.d. X
10 cm) were uniformly packed with air-dry soil and were slowly water-saturated. Upon
saturation, the fluxes were adjusted to the desired flow rates. A Cu pulse of 100 mg L-
1
was introduced into each column after it was totally saturated with 0.01 N MgS0
4
or
Mg(CI0
4
h as the background solution. Perchlorate as the background solution was
used to minimize ion pair formation. The Cu pulse was eluted subsequently with 0.005
M MgS04 solution. The ionic strength was maintained nearly constant throughout the
experiment. In other soil column experiments, similar conditions were used except that
no background solution was used in the Cu pulse input solution. This resulted in a
condition of variable total ionic concentrations or ionic strength during input pulse ap-
plication and the subsequent leaching solution. A fraction collector was used to collect
column effluent.
Figures 4.7 and 4.8 show the effect of total concentration or ionic strength of the input
pulse solution on Cu breakthrough results. When Cu was introduced in Mg background
solution with minimum change in ionic strength, Cu breakthrough curves (BTCs) ap-
pear symmetrical in shape with considerable tailing and peak concentration of 40 mg L-
1

Mg BTC shows an initial increase in concentration due to slight increase in ionic strength
followed by a continued decrease during leaching. When Cu was introduced in the ab-
sence of a background solution, the total concentration considerably decreased from
0.005 to 0.0015 M. As shown in Figure 4.8, Cu BTC showed a sharp increase in concen-
tration due to chromatographic (or snowplow) effect (Selim et al., 1992). The Cu peak
concentration was 94 mg L-
1
and the corresponding Mg concentration in the effluent
decreased due to depletion of Mg during the introduction of Cu. Mg concentration in-
creased, thereafter, to a steady state level during subsequent leaching, however. This
snowplow effect is a strong indication of competitive ion exchange between Mg and Cu
cations. The amount of Cu recovered in the effluent was 53% of that applied in the
presence of MgS04 as the background solution, whereas only 38% was recovered when
no background solution was used. For McLaren soil (Figure 4.9), the snowplow effect
was pronounced as shown in Figures 4.7 and 4.8 due to changes in total concentration of
input solutions with a recovery of 60% of that applied. Therefore, miscible displacement
or transport experiments indicated that there was strong ion exchange between Cu and
Mg cations which was also affected by the counter ion used. Effluent peak concentra-
tions were 3-5 fold that of the input Cd pulse, which is indicative of pronounced chro-
matographic effect.
104 Fate and Transport of Heavy Metals in the Vadose Zone
----- ---- - - ~ - - - - ~ - . - - - - -- ----- - ~ - - - - ~ - - ' -
200
-
Cecil Soil Column I
...J
- Sulfate Ol
E
-
150
C
0
:;::
Mg
~
....
...
c:
100
Q)
0
c:
0
U
...
c:
50
Q)
:J
::::
w
Cu
0
0 10 20 30 40 50 60
Pore Volume
Figure 4.7. BTCs of Cu and Mg for Cecil soil (Column I).
200
-...J
-
Column II en
Cecil Soil
E
Sulfate
- 150
c
0
Mg
+:i
cu
...
-
C
100
(I)
u
c
0
()
-
50
c
(I)
::::I
tt:
w
Cu
0
0 10 20 30 40 50 60
Pore Volume
Figure 4.8. BTCs of Cu and Mg for Cecil soil (Column /I).
In summary, we presented an overview of several models which are used for the de-
scription of the retention of heavy metals in soils. Single reactions models were classified
into equilibrium and kinetic types. A general purpose multireaction kinetic and trans-
port model was also presented. Major features of multi reaction models are that they are
fiexible and are not restricted by the number of solute species present in the soil system
Modeling the Kinetics of Heavy Metals Reactivity in Soils 105
-..J
-
C)
E
-
c
o
~
~
+"
C
Q)
(.)
C
o
(.)
+"
C
Q)
::::J
e
W
200
150
100
50
0
McLaren Soil
0 10 20
,..

30
Column V
Sulfate
40
Pore Volume
Figure 4.9. BTCs of cu and Mg for McLaren soil (Column V).
Mg
Cu
50 60
nor the governing retention reaction mechanisms. This includes reversible and irrevers-
ible reactions of the linear and nonlinear kinetic types. Moreover, these models can in-
corporate concurrent as well as consecutive type retention reactions which may be
equilibrium or kinetic in nature. Ion exchange mechanisms of the instantaneous and
kinetic types were also presented. Case studies of Cd and Cu isotherms as well as trans-
port in soil columns provided illustrations of model applications.
REFERENCES
Abd-Elfattah, A. and K. Wada. Adsorption oflead, copper. zinc. cobalt. and calcium by soils that
differ in cation-exchange materials. J. SoiL Sci. 32. pp. 271-283. 1981.
Amacher. M.C . H.M. Selim. and I.K. Iskandar. Kinetics of chromium (VI) and cadmium
retention in soils; A nonlinear multireaction model. Soil Sci. Soc. Am. J. 52. pp. 398-408. 1988.
Amacher. M.C . J. Kotuby-Amacher. H.M. Selim. and I.K. Iskandar. Retention and release of
metals by soils-evaluation of several models. Geooerma. 38. pp. 131-154. 1986.
Aringhieri. R.. P. Carrai. and G. Petruzzelli. Kinetics of Cu and Cd adsorption by and Italian soil.
SoiL Sci. 139. pp. 197-204. 1985.
Barrow. N.J. and T.C. Shaw. Effects of solution and vigor of shaking on the rate of phosphate
adsorption by soil. J. Soil Sci. 30. pp. 67-76. 1979.
BitteL J.E. and R.J. Miller. Lead. cadmium and calcium selectivity coefficients of a montmoril-
lonite. illite and kaolinite. J. Environ. QuaL. 3. pp. 250-253. 1974.
Harmsen. K. Behavwr of Heavy Metau in Soiu. Centre for Agriculture Publishing and Documenta-
tion. Wageningen. The Netherlands. 1977.
Hinz. C. and H.M. Selim. Transport of Zn and Cd in soils: Experimental evidence and modelling
approaches. SoiL Sci. Soc. Am. J. 58. pp. 1316-1327. 1994.
106 Fate and Transport of Heavy Metals in the Vadose Zone
Jardine, P.M., J .C. Parker, and L.W. Zelazny. Kinetics and mechanisms of aluminum adsorption
on kaolinite using a two-site nonequilibrium transport model. Soil Sci. Soc. Am. J. 49, pp. 867-
873, 1985.
Lapidus, L. and N.L. Amundson. Mathematics for adsorption in beds. VI. The effect of longitu-
dinal diffusion in ion exchange and chromatographic column. J. PhYd. Chem. 56, pp. 984-988,
1952.
Ogwada, RA. and D.L. Sparks. A critical evaluation on the use of kinetics for determining
thermodynamics of ion exchange in soils. Soil Sci. Soc. Am. J. 50, pp. 300-305, 1986a.
Ogwada, R.A. and D.L. Sparks. Kinetics of ion exchange on clay minerals and soil. I. Evaluation
of methods. Soil Sci. Soc. Am. J. 50, pp. 1158-1162, 1986b.
Ogwada, RA. and D.L. Sparks. Kinetics of ion exchange on clay minerals and soil. II. Elucida-
tion of rate-limiting steps. Soil Set: Soc. Am. J. 50, pp. 1162-1166, 1986c.
Parker, J.C. and P.M. Jardine. Effect of heterogeneous adsorption behavior on ion transport.
Water Ruour. Ru. 22, pp. 1334-1340, 1986.
Rasmuson, A. and I. Neretienks. Migration of radionuclides in fissured rock. The influence of
micropore diffusion and longitudinal dispersion. J. GeopPyd. Ru. 86, pp. 3749-3758, 1981.
Rubin, J. and RV. James. Dispersion-affected transport of reacting solution in saturated porous
media, Galerkin method applied to equilibrium-controlled exchange in unidirectional steady
water flow. Water Ruour. Ru. 9, pp. 1332-1356, 1973.
Selim, H.M. Prediction of contaminant retention and transport in soils using kinetic multireaction
models. Environ. HeaLth Perdpec. 83, pp. 69-75, 1989.
Selim, H.M. Modeling the transport and retention of inorganics in soils. Adv. Agron., 47, pp. 331-
384, 1992.
Selim, H.M. and M.C. Amacher. A second-order kinetic approach for modeling solute retention
and transport in soils. Water Ruourc. Ru. 24, pp. 2061-2075, 1988.
Selim, H.M., B. Buchter, C. Hinz, and L. Ma. Modeling the transport and retention of cadmium
in soil: Multireaction and multicomponent approaches. Soil Sci. Soc. Am. J. 56, pp. 1004-1015,
1992.
Selim, H.M., J.M. Davidson, and R.S. Mansell. Evaluation of a 2-site Adsorption-Desorption
Model for Describing Solute Transport in Soils. Proc. Summer Computer Simulation Con! Wadh-
ington, DC, 12-I4Ju!y, 1976, La Jo!la, CA, Simulation Councils Inc., La Jolla, CA. 1976, pp. 444-
448.
Selim, H.M., M.C. Amacher, and I.K. Iskandar. Modeling the Transport of Heavy Metals in
Soils. CRREL Monograph 90-2, U.S. Army Corps of Engineers, 1990, p. 158.
Sparks, D.L. Kinet0 of Soil Chemica! ProceddU. Academic Press, San Diego, CA, 1989.
Sposito, G. The Surface Chemidtry of Soil!. Oxford University Press, New York, 1984.
Theis, T.L., R Iyer, and L.W. Kaul. Kinetic studies of cadmium and ferricyanide adsorption on
goethite. Environ. Sci. Techno/. 22, pp. 1032-1017, 1988.
van der Zee, S.E.A. T.M. and W.H. van Riemsdijk. Sorption kinetics and transport of phosphate
in a sandy soil. Geoderma 38, pp. 293-309, 1986.
CHAPTER S
Copper Retention as Affected by
Complex Formation with
Tartaric and Fulvic Acids
Alexander A. Ponizovsky, T.A. Studenikina, and E.V. Mironenko
INTRODUCTION
Soil solutions from humic horizons and surface waters contain inorganic and organic
compounds capable of forming complexes with heavy metals (HM). Complexation can
influence mobility of metals and their availability to living organisms. Many experimen-
tal studies of the HM retention in soils have been conducted without taking into account
complexing ligands. Copper is a trace metal commonly present in many surface waters
and soil solutions in detectable amounts. The influence of complex formation on copper
behavior in soils and waters is the subject of this chapter.
Copper(lI) Retention by Soils, Oxides, and Clays
Copper retention in soils is a rather complex process even without any substances
able to associate with the metal ions in solutions. McLaren and Crawford (1973a, 1973b)
suggested separating the following pools of natively available copper in soils:
present in soil solution and exchangeable Cu (extractable by 0.05 M CaCI
2
)
specifically bound by inorganic sites (soluble in 2.5% acetic acid)
specifically bound with organic matter (soluble in K-pyrophosphate)
occluded by free oxides (soluble in acid solution of ammonium oxalate)
residual (bound in the lattices of minerals, soluble in HF).
These pools are not strictly related to types of chemical bounds. The total amount of
copper retained in soil may not be in equilibrium with the Cu(II) in the soil solution, and
may not be exchangeable and extractable with salt solutions. Zhang and Sparks (1996)
reported that the quantity of Cu extracted by Na2EDTA from the montmorillonite used
10"'7
108 Fate and Transport of Heavy Metals in the Vadose Zone
in Cu-retention experiments was only 1.0 to 1.2% of the Cu-exchange capacity. Copper
may be adsorbed by soil colloids in amounts in excess of their conventional exchange
capacities, assuming adsorption as the Cu
2
+ ion (e.g., McLaren and Crawford, 1973a,
1973b). This phenomenon has been termed "specific adsorption" and is shown to occur
on clays (Bingham et aI., 1964), organic matter (DeMumbrum and Jackson, 1956), and
free oxides (Grimme, 1968).
Copper sorption has been actively studied on soil components, such as Mo, Fe, AI
oxides, or bentonite. Murray (1975) studied Cu
2
+ retention on manganese oxide and
found that adsorption of one Cu
2
+ ion leads to the release of 1.2 H+ ions. An inequivalence
of proton released to retained metal was also found for Ca
2
+, Mg2+, Zn
2
+ and some other
metals. This was explained by the replacement of a proton on the surface by a divalent
metal ion. Shindler et al. (1976) found that the number of protons released per Cu
2
+
adsorbed increases from 1 to 2 with the enhancement of pH. They attributed such a
phenomenon to the retention of [CuOHr complexes instead of free Cu
2
+ ions.
Infrared spectroscopic studies by Parfitt and Russel (1977) showed that sorption of
Cu
2
+ on goethite at pH 7 did not lead to direct interaction between copper ions and
surface OH groups. According to Rodda et al. (1996), experimental data on adsorption
on goethite can be fitted by a model in which monomeric Cu(OH)+ and dimeric
CU2(OH)22+ complexes compete for surface sites, with the dimer more strongly adsorbed
to the surface. The number of protons released per Cu(II) ion adsorbed at pH 5.0 in-
creased from 1.76 at 25
D
C to 1.92 at 70
D
C.
Bower and Truog (1941) reported enhanced retention of Cu(II) by bentonite from
chloride background medium. This was explained by the CuCI+ and [CuOHr ion pair
involvement in the ion exchange and/or copper hydroxide precipitation. Sposito et al.
(1981) also found that copper sorption by Na-bentonite in chloride solutions exceeds
that in perchlorate solution. They suggested that monovalent CuCI+ complex was re-
sponsible for this effect, though this ion pair is known to be unstable and its concentra-
tion in solutions low. In these experiments, the solution pH was not stabilized and
decreased with the increase in exchangeable copper content. Such results were not con-
firmed in the study of Zhang and Sparks (1996). They obtained values of Cu-exchange
capacity for Na-montmorillonite similar for CI-, (CI0
4
)-, (N0
3
)-, and (S04)2- back-
ground at 0.25 M Cu(II) solution concentration and pH from 4.31 to 4.54. This concen-
tration level was rather high, and they suggested that some other relations could be
observed at a much lower Cu concentration, when specific adsorption occurs mainly at
nonexchangeable sites on clay edges.
H-montmorillonite was shown to retain copper from the chloride and acetate solu-
tions at pHd in amounts equal to the standard cation exchange capacity (CEC) (Bingham
et aI., 1964). In acetate solutions at pH>3, the amount of Cu
2
+ retained exceeded the
CEC. Acetate was sorbed by clay, but the sorbed amount was not equal to the excess of
Cu
2
+ retained.
Soils selectively adsorb Cu
2
+. Harter (1992) studied this phenomenon and revealed
that the amount of Cu
2
+ adsorbed exceeded that of the Ca
2
+ desorbed. He hypothesized
that this discrepancy could be attributed either to nonexchange sorption processes or to
the of both free ITletal and their charged complexes, e.g., [CuOHr which
results in maximum metal retention greater than CEC. McLaren and Crawford (1973b)
suggested that specific and nonspecific adsorption of copper by soils and montmorillo-
Copper Retention as Affected by Complex Formation 109
nite take place simultaneously and independently. The specific adsorption exceeds non-
specific adsorption at low copper concentrations, while the relation can be opposite as
the concentration increases. They assumed that nonspecific adsorption isotherms can be
described by the Vanselow equation, whereas the Langmuir adsorption equation was
found suitable for the "specific adsorption" isotherms within solution concentration ranges
of 1-5 mg mL-
1
(0.015 to 0.075 mmol L-
1
). Sposito et al. (1981) described sodium-cop-
per exchange isotherm on bentonite in the perchlorate solution by the Vanselow equa-
tion and obtained the selectivity coefficient independent of exchangeable Na/Cu ratio.
The same equation was applied to describe Cu-Na exchange on montmorillonite at pH
from 5.17 to 6.50 and total concentration of metal ions in the solution of 0.02 mole L-
1
(Zhang and Sparks, 1996).
Solution Complex Formation and Cu(1I) Adsorption
Copper(II) forms complexes with a number of anions commonly found in natural
waters (Yatsimirsky and Vasilyev, 1959). In water solutions free of other ligands, Cu(II)
is present mainly as a complex [Cu(H
2
0)6J 2+, a Lewis acid with pK=6.8, and
[Cu(OH) (H
2
0hr, a Lewis base. Simulations of the Cu(II) speciation in fresh waters
indicate a possible presence of a variety of complexes with different charges, depending
on the solution pH (Vuceta and Morgan, 1978).
The discrimination between copper complexes in natural waters remains a challeng-
ing analytical problem. A number of direct electrochemical techniques have been used in
an attempt to distinguish between chemical forms of metals. Some of the existing meth-
ods are based on the separation of the different species. Dialysis, ultrafiltration, and
centrifugation have been used to separate "free" metals in natural waters from metals
associated with colloid material. Extraction with chloroform and chelating with ion ex-
change resins have been applied to separate "bound" copper from the free ions. Several
studies were carried out on sea water (e.g., Florence and Batley, 1977) and sewage efflu-
ents. In sewage effluents Bender et al. (1970) using gel filtration chromatography
(Sephadex G-50) found no free Cu
2
+ ions. About 13% of the copper in the sample was
associated with molecular weights of 10
4
or greater, while the remaining molecular weights
ranged from 500 to 1000.
A ligand that can be sorbed on a surface can enhance or inhibit retention of Cu (Davies
and Leckie, 1978). Negatively charged Cu
2
+ complexes with organic acids can be only
slightly sorbed on the surface of layer silicate minerals (Bloomfield et al., 1976). How-
ever, glutamic acid increases retention of Cu(II), the influence of salicylic acid is negli-
gible, and picolinic acids inhibits copper retention by hydrous oxides (Davies and Leckie,
1978). Pampura (1993) found that from 1 to 3 mmol L-
1
citric, malic, and salicylic acids
decreased the Cu(II) sorption by soil. Such an effect can be explained by the formation
of complex ions of various charge and with different abilities to be adsorbed on charged
surfaces. Janvion et al. (1995) found that the concentration of acetate affected the distri-
bution coefficient of Cu(II) between the acetate buffer in the mobile and stationary phases
having both cation- and anion-exchange groups. This was explained by suggesting that
copper retention involves not only Cu
2
+ cation exchange, but also anion exchange of
copper anionic complexes, e.g., with 2,3-pyrazinedicarboxilic acid (2,3-PDCA),
[Cu(PDCA)2f-.
ilO Fate and Transport of Heavy Metals in the Vadose Zone
Elliot and Huang (1979) reported on a survey of a number of studies on the influence
of several chelating agents on the adsorption of various metal ions by some oxides. They
presented results on adsorption of Cu(II) complexes with glycine, nitrilotriacetic, and
aspartic acids on y-Al203. These authors also studied adsorption of metal-amino acid
complexes by y-Al203' which is a relatively hydrophilic and hydrophobic activated car-
bon, and adsorption of Cu(II) complexes with various ligands by several alumosilicates
with different surface charges and ion-exchange properties (Elliot and Huang, 1980,
1981). Their results indicated that several adsorption mechanisms were affecting the
retention. Besides hydrophylic complexes, hydrophobic complexes probably could be
sorbed on the hydrophobic parts of particle surfaces. This was proved by Baffi et al.
(1994), who studied Cu(II) adsorption on inorganic fractions of marine sediments in the
presence of aminoacids. Stumm et al. (1976) assumed that the interaction of the surface
functional groups with the metal ions to be similar with the complexation with ligands in
the solution. That is probably the reason why the modification of the surface of clay
minerals, e.g., by polyphosphates, enhances their ability to sorb trace metals (Klimova
and Tarasevitch, 1992).
Cu (II) adsorption by hydrous oxides was found to decrease in the presence of a
complexing ligand (citric acid and EDT A) in a manner that suggests competition be-
tween the ligand and oxide surface for complexation of the metal ion (Davies and Leckie,
1978). They studied the influence of several complexing ligands, capable of adsorbing
on the surface of amorphous iron oxide on the trace metal uptake by oxide surface. The
adsorption edge for Cu(II) was found in the pH range 5 to 6. Adsorption of metal ions on
various oxide surfaces increased abruptly in this pH range, where hydrolysis products
became a significant fraction of the dissolved metal. The authors suggested that adsorbed
CuOH+ was the main form of copper(II) on the surface. The location of the adsorption
edge for Cu(II) was influenced by ionic strength and total Cu(II) concentration. Cu(II)
binding by amorphous Fe(OH)3 was not affected by the addition of salicylic and
protocatecholic acids in the experiments of Davies and Leckie (1978). Adsorption was
enhanced by glutamic acid and 2,3-PDCA. Picolinic acid effectively prevented copper(II)
removal from the solution. The authors attributed these effects to the adsorption of
glutamic acid and 2,3-PDCA on Fe(OHh This adsorption should increase the binding
strength of the surface for Cu(II). Adsorbed picolinate could not function as a complexing
ligand for metals due to the fact that coordinating groups were bound with the surface
and unavailable to the metal ion.
Complexes of CI.I(II) with Fulvic Acids
Fulvic acids (FA) are important components of surface waters and soil solutions. The
complexing ability of FA is well known but rather difficult to be described quantitatively
since FA are high molecular weight substances of complex composition with a great
number of functional groups. Complex formation between Cu(II) and FA was studied
by different techniques. A set of the methods was applied to calculate stability constants
of soil organic matter with Cu(II) (see, e.g., Young et aI., 1982; Hirose et al., 1982;
Perdue and Lytle, 1983; Ephraim and Marinsky, 1990; and a review by Bizri et aI.,
1984). These methods provide some conditional values, related to definite solution ionic
strength, copper and FA concentrations, etc. For example, stability constants calculated
Copper Retention as Affected by Complex Formation 111
by Ruzie and Scatchard plots, obtained by ion selective electrodes and potentiometric
stripping analysis, decreased continuously with the degree of site occupation (Soares
and Vasconcelos, 1994).
Influence of FA and Humic Acids (HA) on the Retention of
Cu(lI) by Solid Phases
Humic substances can be sorbed on mineral surfaces containing hydroxylated Al, Fe,
or Mn sites (Parfitt et al., 1977; Murphy et al., 1990), thereby changing the mineral sur-
face properties. Surface bound humic substances increase the adsorption of some metal
cations on single mineral solids by contributing additional potentially high affinity com-
plexation sites. For copper which forms strong complexes with humic substances, these
substances enhance metal binding to Fe and AI oxides at lower pH (e.g., < 6) by co-
sorption and decrease metal ion retention at higher pH by formation of nonadsorbing
aqueous complexes (Tipping et al" 1983; Allard et al., 1989). For other combinations of
metals and sorbents, mineral-bound humic substances exhibit variable effects on metal
binding (Xu et al., 1989). It depends on the complexation ability of humic substance for
the metaL the distribution of the humic substance between the solution and solid, and the
variation of both of them with respect to pH and ionic strength (Zachara et al., 1994).
It has not been estimated whether the mineral-bound humic substance exhibits com-
parable complexation properties for metals to that in solution. Both Davies (1984) and
Laxen (1985) concluded that metal-humic substance interactions were stronger on sur-
faces than in solution. Tipping et al. (1983) suggested that when humic substances bind
to goethite, new high affinity metal complexation sites are formed.
Dissolved FA decreased Cu(II) retention by montmorillonite, though the effect was
less than the influence of citrate and EDTA (Vuceta and Morgan, 1978). Clay in Cu-
form adsorbs FA and HA (Theng and ScharpenseL 1975). The authors attributed this
phenomenon to the precipitation of slightly soluble copper salts of FA and HA due to
release of exchangeable Cu(II). Chakrabarti et al. (1984) mentioned that dissolved FA
decrease the velocity of the dissolved copper retention by the solid ion-exchange phase.
This was perhaps caused by the low velocity of dissociation of the Cu-FA complex.
Wershaw et al. (1983) suggested that Cu(II), forming charge transfer complexes with
F A, in some conditions coordinates not one, but two or more FA molecules, and this
leads to the aggregation of molecules. Such a point of view coincides with the existing
conception of humic substances aggregation by metal ions (Gamble et al., 1984). It was
found that, while extracting bound copper from bentonite, the amount of Cu(II) re-
leased was dependent on the concentration of ligand and the stability constant of the
metal complex. The ratio of metal release to complexing sites decreased in the order:
EDTA> humic acid> tannic acid (Guy and Chakrabarti, 1976).
COPPER RETENTION BY SOil (A CASE STUDY)
Kinetics of Cu(lI) Retention
A case study of the Cu(II) retention was carried out on a silty clay Typic Haplustoll
soil (leached chernozem) sampled in the Tula region, Russia, from the A horizon (0-20
cm depth). Soil organic C content was 27.5 g kg-I, and exchangeable cation contents,
112 Fate and Transport of Heavy Metals in the Vadose Zone
determined by modified Pfeffer method (Khitrov, 1982) (extraction by 0.1 M NH
4
CI
solution in 70% alcohol), were 27.4, 1.7,0.13, and 0.70 cmol
e
kg-I ofCa
2
+, Mg2+, Na+, and
K+, respectively, with total cation exchange capacity 29.9 cmol
e
kg-I. pH measured in
water extract with 1: 1 soil: water ratio was 5.77. Each of these values is an average of
three subsamples.
The soil sample was air dried and ground to pass through a 2-mm sieve. To remove
carbonates that can be present in the leached chernozem in trace amounts, the sample
was leached with 0.1 M HCI until the leachate pH was about 3. Then sediment was
washed out with water to remove the excess HCl.
Then sample was saturated with Ca2+ by treatments with 0.1 M CaCI
2
, to diminish to
a trace level an exchangeable H+ content, estimated by 1 M KCI extraction with subse-
quent KOH titration. To remove extra CaCl
2
the sediment was washed with water, until
the final Ca
2
+ concentration in leachates became 2.5 mM (further decrease leads to pep-
tization of the sediment). Then, the solution was decanted, the sample was air dried,
ground in a mortar and sieved to pass a 1-mm mesh screen.
Soil samples were placed into the flasks and suspended in 3 mM Ca(N0
3
)2 solution
as a background. Cu(N0
3
h solution was added and pH value was adjusted by adding
HN0
3
or KOH. The flasks were shaken at 251 C for 90 days. The supernatant solu-
tions were decanted, filtered through 0.2 pm membrane filters and copper concentration
was measured with atomic adsorption spectroscopy (AAS). The amount retained was
taken as the difference between the amount added and the amount recovered in the
equilibrium solution.
Equilibrium of Cu(n) retention by soil was found to be obtained for the period from
4 to 24 hours (Figure 5.1). Sorption velocity was highest at pH 6 (about 95% of the
maximal copper retention was observed already in 1 hour) and lowest at pH 4. Mter 24
h no increase in soil copper content was observed.
Cu(1I) Retention Isotherms and Cation Balance
Soil samples were suspended in 3 mM Ca(N0
3
h Then some amounts of the 0.1 1\ 1
CU(N0
3
)2 solution were added, and pH value was adjusted, adding HN0
3
or KOB
The isotherms were obtained for pH 4,5, and 6. The flasks were shaken at 251 C for I
day. In some intervals the pH value was corrected by titration with HN0
3
or KOH. The
suspensions were centrifuged and supernatant solutions were analyzed. Retained cop-
per was calculated as above. Amount of H+ displaced from soil by Cu
2
+ was evaluatec
from the amount of HN0
3
or KOH used to adjust the pH value.
Isotherms of Cu(n) displacement by Ca
2
+ from the contaminated soil were obtainec
by treatment of the soil residue in the flasks by 3 mM Ca(N0
3
h with the pH adjustmen:
as mentioned above. The suspensions were shaken for 1 day, centrifuged, and analyzec
Cu(n) retention isotherms at different pH are presented on Figure 5.2. Increase i:-
pH leads to the enhanced copper sorption. At pH 6 the sorption is much higher, and the'
shape of the isotherm is different from those at pH 4 and 5.
Cu(n) sorption in all cases was accompanied by the release of both Ca
2
+ and H+ ions
from soil to solution. To maintain pH level, KOH was added to the suspension and K-
was partly retained by soil.
Copper Retention as Affected by Complex Formation 113
60
50
""'"
-
-
40
";
~
-
-. -.
r- - -
r
CI
..w:
"5
E 30
E
______ pH = 4
I--
_____ pH = 5
:,
0
--A-- pH = 6
en
20
10
o
o 20 40 60 80 100 120 140 160 180
time, h
Figure 5.1. Kinetics of Cu(11) retention by soil.
350
300
250
~
200
a
E
E
:, 150
0
en
100
50
0
0 1
_ pH=4, Cu retention
___ pH=5, Cu retention
~ pH=6, Cu retention
-cr- pH=4, Cu displacement
-0- pH=5, Cu displacement
2
meu' mM
3
Figure 5.2. Isotherms of Cu(1I) retention by soil without ligands at pH 4, 5, and 6.
114 Fate and Transport of Heavy Metals in the Vadose Zone
250 ,---------------------------------------,
";"0 200
~
_I>
o
E
E 150
~
Q)
c::
iii
i! 100
....
o
"5l
VI
m 50
~
VI
c::
.2
B 0
80 100 120 140 160 180 200 220 240 260 280 300
Cu sorbed, mmol
c
kg
1
Figure 5.3. Cation balance at Cu(1I) retention, pH=5.
It was found that the relations of released and retained ions amounts, e.g., at pH 5
(Figure 5.3) in the studied range can be expressed as
Ca
2
\el. = 0.765 CUsorb.
H+ reI. = 0.292 CUsorb.
K\orb. = 0.0332 CUsorb.
Here Ca
2
+
re
l.' H+rel.' K\orb., and CUsorb. are amounts of Ca
2
+, H+ released and K+, Cu(II)
retained (mole kg-
1
soil), respectively. Thus total cation balance for copper retention is
Taking into account measurements errors, it could be concluded that amount of iom
released is equal to the amount retained. Balance between Cu(II) retained and Ca
2
+ and
H+ released at pH 4, 5, and 6 is presented in Table 5.1. Impact of K+ retention or dis-
placement on cation balance is rather small. In all studied cases Cu(II) displaced from
soil both Ca
2
+ and H+. At pH 4 and 6 (Ca
2
+
re
l. + H+rel)/Cu2+sorb. ratio was 0.89 and 0.98,
respectively, i.e., not so much different from the value of 1.05 at pH 5. Variation of thi;;
ratio, caused probably by the measurements errors, doesn't allow rejecting the hypoth-
esis on the exchange equivalence.
The "reverse" isotherm of copper displacement by Ca
2
+ at pH 4 does not diverge
significantly from the direct one (Figure 5.2). The increased distinction at pH 5 can be
Copper Retention as Affected by Complex Formation 115
Table 5.1. Cation Balance at Cu(I1) Retention by Soil (mole per mOle Cu(I1)) NL - no Ligands, (1) -
Cu(1I) Retention, (2) - Cu(I1) Displacement
pH=4 pH=5 pH=6
NL TA FA NL TA FA NL TA FA
(1) (2) (1) (1 ) (1 ) (1) (1) (1 ) (1 ) (1)
Ca
2
+
0.64 0.64 0.54 0.63 0.76 0.43 0.39 0.21 0.16 0.19
H+
0.25 0.32 0.29 0.18 0.29 0.34 0.19 0.77 0.76 0.77
(Ca
2
++W) 0.89 0.96 0.83 0.81 1.05 0.77 0.58 0.98 0.92 0.96
attributed to (i) displacement of Cu(II) mainly by Ca
2
+ and not by H+ ions; (ii) the
differences in the solution Ca
2
+ concentrations in direct and reverse runs; (iii) slower
velocity of the reverse process comparatively with the direct one.
Thus the Cu
2
+ _(Ca
2
+ + H+) exchange seems to be a reversible process, though it could
be difficult to displace all the copper retained due to the higher soil selectivity with re-
spect to Cu (II).
Evaluation of Na
2
EDTA Ability to Extract Retained Copper
To extract retained copper, contaminated soil was 3 times extracted with portions of
0.02 mole L-
1
Na2EDTA. To prevent peptization and to improve copper displacement,
lO mM Ca(N0
3
)2 was added to the Na2EDT A solution.
As it is shown on Figure 5.4, copper was quantitatively revealed from the contami-
nated soil by 0.02 mole L -I Na2EDTA extraction. Though CuEDT A is a very stable
complex (pK = 18.9) (Sillen and Martell, 1970), first treatment revealed only about 80%
of the total copper retained amount, and 3 subsequent extractions by Na2EDTA solu-
tion were necessary for complete copper displacement.
Ca(N0
3
)2 + Na2EDTA solution was shown to extract Cu(II) more efficiently and did
not peptisize soil as it was observed for Na2EDT A.
Effect of Tartrate and Fulvic Acid on Cu(lI) Retention Isotherms
Tartaric acid (T A) and FA were taken as samples of soluble organic compounds of
soil solutions. Tartaric acid (HOOC-CH(OH)-CH(OH)-COOH) is a low-molecular-
weight dicarbonic hydroxy acid found in root exudates of many plant species (Riviere,
1960; Smith, 1969; Ivanov, 1973; Hale and Moore, 1978), and in leachates from decom-
posed leaf litter (Nykvist, 1963). The anion derived from dissociation of tartaric acid
forms relatively stable complexes with many metals. Therefore, the presence of tartrates
may affect the mobility of heavy metals in the root zone and their uptake by plants.
Fulvic acids are probably the most common high molecular weight compounds present
in soil solutions. Being polyelectrolytes with a variety of functional groups, they are able
to form chelate complexes with HM. Baker (1973) has proposed that the metal trans-
port mechanism in soils involves mainly complexes with fulvic and humic acids.
For our case study, FA was extracted from the soil samples by 0.1 M KOH with the
subsequent sorption on Amberlite XAD-8, as it was recommended by IHSS (Kuwatsuka
et aI., 1992), and was purified by dialysis with dialysis membrane "Film 100"
1 16 Fate and Transport of Heavy Metals in the Vadose Zone
150 ---,--------------------;;,
CU,eleased =4.29+0,921 *Cu'otoIned (r=0.989)
120
";'
Cl
~
'0
E
E
'ti'
Q)
90 tJ)
<II
~
~
...
Q)
c..
c..
8
60
3 0 ~ - - - - . - - - - _ . - - - - _ , - - - ~
30 60 90 120 150
Cu sorbed, mmol kg-
1
Figure 5.4. Efficiency of the retained Cu(lI) extraction from the soil by 3 treatments with 0.02 male
L-
1
Na
2
EDTA.
CPolymersintez," Vladimir, Russia). The solution was freed of salts by passing through
cation-exchange resin KU-2 in H-form. Then, the sample was freeze dried. The stock
solution used for the experiments contained 2 g L -1 dry FA.
Tartrate added forms a number of complexes with Cu(II) (Ponizovsky et aI., 1997)
decreasing its activity in the solution. That can be illustrated by the data obtained by
adding tartaric acid to 0.001 M CU(N0
3
)2 solution, measuring Cu
2
+ activity by ion selec-
tive electrode, and calculating the activity coefficient as a ratio of activity to concentra-
tion (Figure 5.5).
To evaluate the influence of TA and FA on Cu(II) retention, tartaric acid, or FA
solution was added to the soil suspension in 3 mM Ca(N0
3
h Then some amounts of 0.1
M CU(N0
3
)2 solution were added, pH value was adjusted adding HN0
3
or KOH, and
the suspensions were shaken at 25 1 C for 1 day. In some intervals the pH value was
corrected by titration with HN0
3
or KOH. The suspensions were centrifuged and su-
pernatant solutions were analyzed. Retained copper was calculated as above.
In soil suspensions at copper concentration less than 0.5 mM, tartrate diminished
Cu(II) retention by soil at all the pH studied (Figure 5.6 a, b, c). At higher copper
concentration, the difference between the retention in containing tartrate and tartrate-
free suspensions is less, and at some Cu(II) level we observed even increase in retention,
probably due to the precipitation of copper tartrate.
Fulvic acid decreased copper retention at low Cu(II) concentration (Figure 5.6 a, b,
c). That can be also attributed to complex formation. At higher Cu(II) concentration an
increase in retention was observed, probably due to precipitation of Cu-fulvate.
Copper Retention as Affected by Complex Formation 117
0.9 ....,..-------------------,
0.8
~ 0.7
u
== CD
8
~ 0.6
. ~
U
m
.t.
8 0.5
0.4
0.3 -;-----.----,----,-----,-----,--------1
0.0 0.2 0.4 0.8 1.0 1.2
Figure 5.5. Cu
2
+ ion activity coefficient as a function of tartrate to Cu(lI) ratio in solution at mcu=l
mM (calculated from the measured Cu
2
+ activity).
100,--------------------,
80 (a)
~ 60
"0
E
E
en"a 40
20
-e- without ligands
___ with tartrate added
----y- with FA added
O ~ - _ . - - , _ - ~ - _ . - - , _ - _ r - ~ - ~
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Figure 5.6a. Copper retention isotherms at pH 4.
118 Fate and Transport of Heavy Metals in the Vadose Zone
160.-------------------------------------,
140
120
~
100
Cl
..>0::
(5
E 80
E
:.
u
CJ)
60
40
20
O ~ - - - - _ r - - - - _ , - - - - _ . - - - - - - r _ - - - - , _ - - - - -
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Figure 5.6b. Copper retention isotherms at pH 5.
600.-------------------------------------,
500
400
~
Cl
..>0::
(5
E 300
E
:.
u
CJ)
200
100
o ~ - - - . - - - - . _ - - _ r - - _ . - - - - , _ - - _ . - - _ . - - ~
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
Figure 5.6(. Copper retention isotherms at pH 6.
Copper Retention as Affected by Complex Formation 119
As it can be seen from the cation balance (Table 5.1), in the presence of tartrate and
FA, copper also replaced both exchangeable Ca
2
+ and H+, and within the experimental
error the exchange seems to be equivalent.
Modeling of Cu(lI) Retention (Exchange) by Soil
Copper(II) retention isotherms were described by Freundlich and Vanselow equa-
tions.
The Freundlich equation was used for heavy metal retention by Van Riemsdijk and
Van der Zee (1989), Selim (1992), etc., though this equation was derived not for ion
exchange but for sorption on energetically nonhomogeneous surface, and does not take
into account the pH dependency of the process. The authors use it as
(1)
or
(2)
Here Scu' mcu, and acu are amount of copper sorbed (mmol kg-I), Cu(II) concentration
(mmol L-
I
), and activity, respectively; K
I
, K
2
, b, and d are coefficients. The Vanselow
approach was applied using the following equations:
Scu _ k acu
Sca - Cu-ca aCa
SCu k acu
-S2 = Cu-H -2-
H aH
2<:fS
cu
+ 2<:fSea + SH = Q
Here Sea and SH are amounts of Ca
2
+ and H+ adsorbed, and Q is some operational value
of cation exchange capacity (mmol
e
kg-I). kCu-ea and kCu_H are Cu-Ca and Cu-H ion
exchange selectivity coefficients. Q differs from CEC in the value of SH. Copper and
hydrogen activities were measured by ISE, and aea calculated based on the solution
composition using the ion association model described by Mironenko et al. (1996). This
model takes into account association of inorganic cations both with inorganic anions and
organic anions derived from the dissociation of tartaric or fulvic acids, and precipitation
of slightly soluble salts.
Freundlich plots of Cu(II) retention isotherms on logarithmic scale are really very
close to linear at all the pH values studied (Figure 5.7). Coefficients values are presented
in Table 5.2. It can be seen that both Equations 1 and 2 are approximately equally suit-
able to describe the isotherms. The slopes of the lines for pH 4 and 5 are similar, and the
intercepts differ from one another.
A pH-dependent Freundlich isotherm was used by several authors to describe ad-
sorption of bivalent heavy metal in acid soils (Lexmond, 1980; Van Riemsdijk and Van
der Zee, 1989)
120 Fate and Transport of Heavy Metals in the Vadose Zone
-3.0 -3.0 -,---------------,
-3.5
-3.5
'8
VJ
-4.0
""-'
Cl
-4.0
..Q
-4.5

pH=4

pH=5
-4.5
A pH=6
-5.0 -+----.---..,----.----.---' -5.0 -+---..,----,----,----,----'
-6 -5 -3 -2
-6 -5 -3 -2
Figure 5.7. Freundlich equation plot for copper retention isotherms: (a) Equation 1, and (b)
Equation 2.
Table 5.2. Freundlich Equation Parameters for Cu(1I) Retention Isotherms; r - Correlation Coef-
ficients
Equation 1 Equation 2
pH Ligand logK b r log K b r
4 Nl 2.94 0.339 0.991 3.02 0.350 0.997
TA 1.92 0.417 0.994 2.02 0.344 0.995
FA 1.88 0.345 0.987 1.94 0.330 0.992
5 Nl 3.09 0.339 0.993 2.11 0.340 0.995
TA 2.18 0.511 0.978 2.31 0.389 0.985
FA 2.04 0.338 0.962 2.11 0.292 0.991
6 Nl 7.02 1.18 0.912 3.27 0.630 0.993
(3)
where Q(M) is amount of heavy metal adsorbed (mol kg-I), (H) and (M) are the proton
and heavy metal activity, respectively, and K
3
, m, and n are constants. Taking K2 as a
function of pH, one can obtain the equation similar to Equation 3 to calculate Cu reten-
tion in some pH range, but the validity of this approximation will be limited, and related
only to some level of Ca
2
+ and other cations concentrations in the solution. Such approxi-
mation could be useful for some applications, but it needs data for some more pH values.
The Vanselow equation can be applied to describe Cu(II) retention if assuming that
(i) retention is a reversible ion exchange process; (ii) Cu(II) displaces from soil both
exchangeable H+ and Ca
2
+; (iii) the exchange is equivalent, i.e., 1 mole Cu(II) dis-
places 1 mole of (H+ + Ca
2
+). Vanselow equation parameters were evaluated based on
the obtained isotherms (Table 5.3). The precision of the data does not enable rejecting
the hypothesis that parameters values are independent on the pH. To describe copper
Copper Retention as Affected by Complex Formation 121
retention, this equation is probably the most universal, but it needs a lot of experimen-
tal data to estimate all the parameters, and some model to calculate the ion activities in
the solutions.
Isotherms of Cu(II) retention in the presence of tartrate and FA also can be fitted by
the Freundlich equation (Table 5.2) but the obtained parameters probably are related
only to the studied levels of Cu(II), tartrate or FA, Ca
2
+, and other ion concentrations.
Application of the Vanselow equation for description of Cu
2
+ retention in the presence
of complex forming substances could be more efficient, but in solutions containing tar-
trate available models does not allow calculating precise values of Ca
2
+ and Cu
2
+ activi-
ties, as far as a number of forming complexes is rather large and their properties are not
well known (Ponizovsky et al., 1997). So we could only evaluate the parameters of this
equation (Table 5.3). That does not enable rejecting the hypothesis of their identity with
the ones obtained for the exchange without ligands.
The model designed by Mironenko et al. (1996) allows evaluating Cu
2
+ activity in the
solutions containing FA, but the precision of Ca
2
+ activity calculation is rather low. So
the attempts to use the Vanselow equation to calculate Cu(II) retention by soil can be
only semiquantitative (Table 5.3). Available data do not allow estimating whether the
parameters of this equation are different for ion exchange with and without ligands.
SUMMARY
Copper(II) retention in soil in all cases can be suggested as accompanied by simulta-
neous displacement of both Ca
2
+ and H+ ions. Estimated ratios of amounts Cu(II) re-
tained and (Ca
2
+ + H+) displaced at pH 4,5, and 6 both in the absence and in the presence
of T A and FA were from 0.77 to 1.05 molc per molc. That doesn't enable rejecting the
hypothesis of equivalent exchange. Thus "nonequivalent" retention of copper, e.g., on
montmorillonite, mentioned by Harter (1992) and Sposito et al. (1981), could be caused
by displacement by Cu
2
+ of not only exchangeable Na+, but H+. Traditional methods
used to estimate the ion exchange capacity and exchangeable cations contents in clays
and soils do not enable evaluating the content of exchangeable H+. However, Cu
2
+ ions
can easily displace them, so the amount of copper retained should be higher than amounts
of Ca
2
+ or Na+ released. This effect did not appear in the experiments of Bingham et al.
(1964), who found that copper retention by H-montmorillonite in the concentration range
from 10--5 to 10-3 M at pH<3 was equivalent to H+ displaced, probably because they
observed Cu
2
+-H+ and not Cu
2
+-Na+ exchange.
Decrease in Cu(II) activity due to complexation with tartraric and fulvic acids leads
to diminishing of Cu(II) retention in soil at concentrations below 0.5 mM. Nevertheless,
at higher copper concentration the retention in tartrate and FA containing suspensions
becomes even higher than in the suspensions without these substances, probably due to
the precipitation of copper tartrate and fulvate. The Freundlich equation can be applied
for approximation of retention isotherms, but its parameters values can be valid only in
a very narrow range of pH, cations concentrations in the solution, and available complex
forming substances.
The Vanselow equation is much more universal to describe Cu(II) retention in soil
even in the presence of some complexes, though its abilities are restricted by the lack of
the models to calculate the ion activities in solutions. Increased retention of Cu
2
+ at el-
122 Fate and Transport of Heavy Metals in the Vadose Zone
Table 5.3. Vanselow Equation Parameters for Cu
2
+ - Ca
2
+ - H+ Ion Exchange
Isotherms
pH
4
5
Ligand
NL
TA
FA
NL
TA
43.9
53.6
28.3
72.4
82.3
k CU-H
4.83*10-
6
6.23*10-
6
4.60*10-
5
5.03*10-
8
1.94*10-
7
k Cu-Ca
0.996*10
4
1.15*10
4
9.44*10
4
1.27*10
4
5.24*10
4
evated pH can be caused by the shift of Cu
2
+-H+ exchange equilibrium. At pH 6 prob-
ably some additional mechanisms of retention are involved.
Single extraction even by N ~ E D T A solution doesn't enable releasing all the Cu(II)
amount retained, probably as far as it is bound in a strong surface complex. However, 3
treatments with the solution containing N ~ E D T A and Ca(N0
3
)2 enable displacing all the
copper retained. This extraction can be suggested for direct estimation of exchangeable
Cu(II) content in the experiments on copper exchange and in the contaminated soils.
REFERENCES
Allard, B., V. Moulin, L. Basso, M.T. Tran, and D. Stammore. Americium sorption on alumina
in presence of humic materials. Geoderma 44, pp. 181-187, 1989.
BaHI, F., M.C. Ianni, M. Ravera, and E. Magi. Study of the influence of free dissolved amino acids
on copper(II) adsorptionlremobilization from inorganic fractions of marine sediments using a
reversed phase liquid chromatographic procedure. AnaL. Chim. Acta. 294, pp. 127-134, 1994.
Baker, W.E. The role of humic acids from Tasmanian podzolic soils in mineral degradation and
metal mobilization. Geochim. COdmochim. Acta. 37, pp. 269-281, 1973.
Bender, M.E., W.R Matson, and RA. Jordan. On the significance of metal complexing agents
in secondary sewage effluents. Env. Sci. TechnoL. 4(6), pp. 520-521, 1970.
Bingham, F.T., A.L. Page, and J.R Sims. Retention ofCu and Zn by H-montmorillonite. Soil Sci.
Soc. Amer. Proc. 28, pp. 351-354, 1964.
Bizri, y', M. Cromer, and J.P. Scharff. Constantes de stabilite de complexes organo-mineraux.
Interactions des ions plombeux avec les composes organiques hydrosolubles des eaux gravitaires
de podsol. Geochim. COdmochim. Acta. 48, pp. 227-234, 1984.
Bloomfield, C., W.1. Kelso, and G. Pruden. Reactions between metals and humified organic
matter. J. SoiL Sci. 27, pp. 16--31, 1976.
Bower, C.A. and E. Truog. Base exchange capacity determination as influenced by nature of
cation employed and formation of basic exchange salts. SoiL Sci. Soc. Am. J. 20, pp. 86-89, 1941.
Chakrabarti, C.L., Lu Yanjia, D.C. Gregoire, M.H. Back, and W.H. Schroeder. Kinetic studies
of metal speciation using Chelex cation exchange resin: application to cadmium, copper and
lead speciation in river water and snow. Environ. Sci. Techno!. 28(11), pp. 1957-1967, 1984.
Davies, J.A. Adsorption of Natural Organic Matter from Fresh-Water Environments by Alumi-
num Oxide, in Contaminantd and Sedimentd, 2, R.A. Baker, Ed., Ann Arbor Science, Ann Arbor,
MI, 1984.
Davies, J.A. and J.O. Leckie. Effect of adsorbed complexing ligands on trace metal uptake by
hydrous oxides. Environ. Sci. TechnoL., 12, pp. 1309-1315, 1978.
DeMumbrum, L.E. and M.L. Jackson. Copper and zinc exchange from dilute neutral solutions
by soil colloidal electrolytes. SoiL Sci. 81, pp. 353-357, 1956.
Copper Retention as Affected by Complex Formation 123
Elliot, H.A. and C.P. Huang. The adsorption characteristics of Cu (II) in the presence of chelating
agents. J. ColloiJ Inter/. Sci. 70, pp. 29-45, 1979.
Elliot, H.A. and C.P. Huang. Adsorption of some copper(II) - amino acid complexes at the solid-
solution interface. Effect ofligand and surface hydrophobicity. Environ. Sci. TechnoL. 14, pp. 87-
93, 1980.
Elliot, H.A. and C.P. Huang. Adsorption characteristics of some Cu(II) complexes on alumino-
silicates. Water lW. 15, pp. 849-855, 1981.
Ephraim, J.H. and J.A. Marinsky. Ultrafiltration as a technique for studying metal-humate
interactions: studies with iron and copper. Anal. Chim. Acta. 232, pp. 171-180, 1990.
Florence, T.M. and G.E. Batley. Determination of the chemical forms of trace elements in natural
waters, with special reference to copper, lead, cadmium and zinc. Tabnta 24, pp. 151-158, 1977.
Gamble, D.S., C.H. Langford, and A.W. Wonderdown. The Interrelationship of Aggregation
and Cation Binding of Fulvic Acids, in Complexation 0/ Trace Metau in Natural Waterd, C.J.M.
Kramer and J.C. Duinker, Eds., Martinus Nijhoff/Dr. W. Junk, The Hague, 1984.
Grimme, H. Die Adsorption von Mn, Co, Cu, und Zn durch Goethit aus verdunnten Losungen.
Z PJbntzenernahr. und Bodenlc. 121, pp. 58-65, 1968.
Guy, R.D. and D.L. Chakrabarti. Studies of metal-organic interactions in model systems pertain-
ing to natural waters. Can. J. Chem. 54, pp. 2600-2611. 1976.
Hale, M.G. and L.D. Moore. Factors affecting root exudation. Adv. Agron. ll(31), pp. 93-124,
1970-1978.
Harter, R.D. Competitive sorption of cobalt, copper, and nickel ions by a calcium-saturated soil.
Soil Sci. Soc. Am. J. 56, pp. 444-449, 1992.
Hirose, K., Y. Dokiya, and Y. Sugimura. Determination of conditional stability constants of
organic copper and zinc complexes dissolved in sea water using ligand exchange method with
EDTA. Marine Chem. ll, pp. 343-354, 1982.
Ivanov, V.P. Pbnt Exudated and Their Role in the Life 0/ PhytocenOded. Nauka, Moscow, 1973 (in
Russian).
Janvion, P., S. Motellier, and H. Pitsch. Ion-exchange mechanisms of some transition metals on
a mixed-bed resin with a complexing eluent. J. Chromatography A., 715, pp. 105-ll5, 1995.
Khitrov, N.B. Evaluation of the Pfeffer method modified by Molodtsov and Ignatova to deter-
mine exchangeable cations content in a single soil sample. Pochvovyedeniye. 6, pp. 105-111, 1982
(in Russian).
Klimova, G.M. and Yu.1. Tarasevitch. Absorption of heavy metals ions from water by sorbents
on the basis of layered silicates modified by polyphosphates. Khimiya i Telchnologiya Vody.
14(12), pp. 929-936, 1992 (in Russian).
Kuwatsuka, S., A. Watanabe, K. Itoh, and S. Arai. Comparison of two methods of preparation
of humic and fulvic acids, IHSS method and NAGGY method. Soil Sci. Pbnt Nutr. 38, pp. 23-
30, 1992.
Laxen, P.H. Trace elements adsorption co-precipitation on hydrous ferric oxide under realistic
conditions. Water lW. 19, pp. 1229-1236, 1985.
Lexmond, Th.M. The effect of soil pH on copper toxicity to forage maize grown under field
conditions. Neth. J. Agric. Sci. 28, pp. 164-183, 1980.
McLaren, R.G. and D.V. Crawford. Studies on soil copper. I. The fractionation of copper in soils.
J. Soil Sci. 24, pp. 172-181. 1973a.
McLaren, R.G. and D.V. Crawford. Studies on soil copper. II. The specific adsorption of copper
by soils. J. Soil Sci. 24, pp. 443-452, 1973b.
Mironenko, E.V., A.A. Ponizovsky, and T.A. Studenikina. Modeling of Chemical Equilibria in
Copper Contaminated Soil Containing Low Molecular Weight Organic Acids and Fulvic
Acids, in Heavy Metau in the Environment: Proceedingd 0/ International Sympodium, PUdhchino, RUddia,
1996. ONTI, Pushchino, 1996, pp. 153-154.
124 Fate and Transport of Heavy Metals in the Vadose Zone
Murphy, E.M., J.M. Zachara, and S.c. Smith. Influence of mineral-bound humic substances on
the sorption of hydrophobic organic compounds. Environ. Sci. Techno!. 24, pp. 1507-1516,
1990.
Murray, J.W. The interaction of metal ions at the manganese dioxide-solution interface. Geochim.
COdm. Acta. 39, pp. 505-519, 1975.
Nykvist, N. Leaching and decomposition of water-soluble organic substances from different
types of leaf and needle. Litter StuJia Fourtolia Succia. Stockholm, 3, pp. 27-31, 1963.
Pampura, T.V. The Influence of Low Molecular Weight Organic Acids on Cu and Zn Sorption
by Chernozem, in PhYdical ChemiJtry of Madd-&cchange ProcNdN in Soiu: Proceedingd of International
Conference, PU.Jhchino, 1992. aNTI, Pushchino, 1993, pp. 85-90.
Parfitt, R.L. and J.D. Russel. Adsorption on hydrous oxides. III. Adsorption of various ions on
goethite. J. Soil Sci. 28, pp. 297-305, 1977.
Parfitt, RL., A.R. Fraser, and V.C. Farmer. Adsorption on hydrous oxides. III. Fulvic acid and
humic acid on goethite, gibbsite, and imogolite. J. Soil Sci. 28, pp. 289-296, 1977.
Perdue, E.M. and C.R Lytle. Distribution model for binding of protons and metal ions by humic
substances. Environ. Sci. Techno!. 17, pp. 654-660, 1983.
Ponizovsky, A.A., E.V. Mironenko, and T.A. Studenikina. Complex formation in calcium and
copper tartrate solutions. Zh. Neorgan. Khim. 42, pp. 632-637, 1997 (in Russian).
Riviere, J. Etude de la rhizosphere de ble. Ann. Agron. 11, pp. 397-440, 1960.
Rodda, D.P., J.D. Wells, and B.B. Johnson. Anomalous adsorption of copper(II) on goethite.
J. Collow Inteff, Sci. 184, pp. 564-569, 1996.
Selim, H.M. Transport and Retention of Solutes in Soils: Multireaction and Multicomponent
Models, in Engineering Adpectd of Metal- Wadte Management, I.K Iskandar and H.M. Selim, Eds.,
Lewis Publishers, Boca Raton, FL, 1992.
Shindler, P.W., B. Furst, R Dick, and P.U.J. Wolf. Ligand properties of surface silanol groups.
I. Surface complex formation with Fe
3
., Cu
2
+, Cd
2
., and Pb
2
+. J. ColLow Inteif. Sci. 55, pp. 469-
475, 1976.
Sillen, L.G. and A.E. Martell. Stability COndtantd of Metal-Ion Comp!exN. Special Publ. No. 25,
Suppl. No.1 to Special Publ. No. 17, The Chemical Society, London, 1970.
Smith, W.H. Release of organic materials from the roots of tree seedlings. For. Sci. 15, pp. 138-
143, 1969.
Soares, H.M.V.M. and M.T.S.D. Vasconcelos. Study of the lability of copper(II)-fulvic acid
complexes by ion selective electrodes and potentiometric stripping analysis. Anal. Chim. Acta.
293, pp. 261-270, 1994.
Sposito, G., KN. Holtzclaw, C.T. Johnston, and C.S. Le Vesque-Madore. Thermodynamics of
sodium-copper exchange on Wyoming bentonite at 298 K Soil Sci. Soc. Am. J. 45, pp. 1079-
1084, 1981.
Stumm W., H. HohL and F. Dalang. Interaction of metal ions with hydrous oxide surfaces.
Croatica Chemica Acta. 48, pp. 491-504, 1976.
Theng, B.KG. and H.W. Scharpensel. The Adsorption of 14C-Labelled Humic Acid by Mont-
morillonite, in Proceedingd of the International Clay Conference, Mexico City. Applied Science Pub-
lishers, Barking, UK, 1975, pp. 643-653.
Tipping, E., J.R Griffith, and J. Hilton. The effect of adsorbed humic substances on the uptake
of copper(II) by goethite. Croatica Chem. Acta. 56, pp. 613-621, 1983.
Van Riemsdijk, W.H. and S.E.A.T.M. Van der Zee. Multicomponent transport modelling of
enhanced metal leaching using synthetic ligands. Geoderma. 44, pp. 143-158, 1989.
Vuceta, J. and J.J. Morgan. Chemical modelling of trace metals in fresh waters: role of
complexation and adsorption. Environ. SCi: Technol. 12, pp. 1302-1309, 1978.
Wershaw, RL., D.M. McKnight, and D.J. Pinkney. The Speciation of Copper in Natural Water
Systems. I. Evidence of Presence of Copper(II)-Fulvic Acid Charge Transfer Complex, in
Copper Retention as Affected by Complex Formation 125
ProceedingJ of the Second InternationaL SympoJium on Peat in AgricuLture and HorticuLture, K.M.
Schallinger, Ed., Hebrew University, Jerusalem, 1983.
Xu, H., J. Ephraim, A. Ledin, and B. Allard. Effects of fulvic acid on the adsorption of Cd(II)
on alumina. Sci. Tot. Environ. 81/82, pp. 653-660, 1989.
Yatsimirsky, K.B. and V.P. Vasilyev. Stability COnJtantJ of Complex CompOUndJ. Academic Science
Publishers, Moscow, 1959 (in Russian).
Young, S.D., B.W. Bache, and D.J. Linehan. The potentiometric measurement of stability
constants of soil polycarboxylate-Cu
2
+ chelates. J. SoiL Sci. 33, pp. 467-479, 1982.
Zachara, J.M., C.L. Resch, and S.C. Smith. Influence of humic substances on Co
2
+ sorption by
subsurface mineral separate and its mineralogic components. Geochim. COJmochim. Acta. 58, pp.
553-566, 1994.
Zhang, Z.Z. and D.L. Sparks. Sodium-copper exchange on Wyoming montmorillonite in chlo-
ride, perchlorite, nitrate, and sulfate solutions. SoiL Sci. Soc. Am. J. 60, pp. 1750-1757, 1996.
CH{\PTER 6
Copper Mobility and Bioavailability in
Relation with Chemical Speciation in
Sandy Soil
E.J.M. Temminghoff, M.P.J.C. Marinussen, and S.E.A.T.M. Van der Zee
INTRODUCTION
In many urban and rural areas, heavy metals have accumulated in soil. Whereas some
heavy metals such as Cu and Zn may be micronutrient elements, others are not essential
for growth (e.g., Cd, Pb). In case the heavy metals become abundant in soil, adverse
effects may occur. Due to heavy metal toxicity, adverse effects may concern both soil
organisms and plants in situ, which threatens soil ecology and agricultural production.
However, also agricultural product quality may be adversely affected without observ-
able effects on yield, whereas leaching may cause a deterioration of the quality of ground-
water; e.g., used for drinking or industrial water. Consequently, much research has been
devoted to quantifying the effects of heavy metal contamination with regard to soil as a
resource and with regard to the transfer of these compounds into the food chain.
It has been well established that heavy metals may be retained by soil due to chemical
interactions with the solid phase. This retention is a major factor that controls heavy
metal biological availability as well as mobility. The key variable appears to be the heavy
metal concentration in the soil solution which is directly related with the displacement in
the flowing soil solution. Also for plants, the supply of heavy metals to the root system is
strongly affected by the transport in solution toward roots and root hairs. Chemical
interactions between the solid and the solution phases of soil control the level at which
metal concentrations are buffered in the solution. This has led to the use of unbuffered
salt extractants; e.g., Ca(N0
3
h CaCI
2
, NaN0
3
, NH
4
N0
3
solutions, to provide a mea-
sure of the bioavailable fraction for, e.g., crops (Sanders et aI., 1987; Novozamsky et aI.,
1993; Aten and Gupta, 1996). Unfortunately, it is hard to predict the extractable quan-
tities for a range of soil types and contaminant levels because these selective extractants
have not been interpreted quantitatively with regard to the involved soil chemical pro-
cesses. Therefore, understanding of bioavailability and mobility still necessitate the un-
derstanding of chemical behavior of the soiIJextractant solution-system.
127
128 Fate and Transport of Heavy Metals in the Vadose Zone
The occurrence of heavy metals in soil solution and solid phase in the form of different
chemical species is known as chemical speciation. Speciation depends on the heavy metal
of interest, the composition of the soil solution and of the solid phase. Thus, pH, ionic
strength, and the presence of cations, anions, organic and inorganic ligands are impor-
tant parameters with regard to the soil solution and types and quantities of primary,
clay, and metal(hydr)oxide minerals. Chemical precipitates as well as organic matter are
important with regard to the solid phase. In general, these parameters are spatially vari-
able in natural soils. For this reason, a practically adequate chemical speciation model
should enable the prediction of behavior, taking such spatial variability into account.
This implies that a model tested in laboratory experiments should be appropriate for
describing observations made in the field. It is our scope to consider in more detail the
chemical speciation of copper in sandy soils (for which organic matter is the main sor-
bent in the acid pH-range) and to use this understanding for the assessment ofbioavail-
ability and mobility.
The aim of this contribution is to present two heavy metal sorption models that ad-
equately capture the main phenomena for the binding by organic matter. Mter param-
eterization of these models for copper sorption by purified humic acid, we show how
sorption by other types of organic matter [i.e., solid (soil) and dissolved organic matter]
can be described with a plausible adaptation of only two parameters. The applicability of
the developed models to describe copper retention and mobility is then assessed. In par-
ticular, we show that the models that have been parameterized with laboratory experi-
ments is adequate in predicting the retention in a soil profile in the field. Since the
agreement between the extracted quantity of metals for a neutral unbuffered salt extrac-
tant (e.g., 0.01 M CaCI
2
) has been shown in the literature for plants, the bioavailability of
copper is discussed for earthworms on the basis of some recent publications. In the soil-
ecotoxicological literature, it is generally agreed that observations made in laboratory
experiments cannot be readily extrapolated to field situations, among others, in view of
soil heterogeneity. We show that in principle such an extrapolation may be feasible, as no
essential differences were observed in copper accumulation at a field site and in labora-
tory studies. However, spatial variability may necessitate a large experimental effort.
SORPTION MODELS
Both empirical and semimechanistic models have been developed to describe heavy
metals binding by soil, by dissolved organic matter, and by organisms. Recently, a
semimechanistic Non-Ideal Competitive Adsorption model (NICA) has been published
which accounts for adsorption by heterogeneous surfaces such as (soil) organic matter
(Van Riemsdijk, 1994; Koopal et al., 1994; Benedetti et aI., 1995). This model can be
characterized by the following isotherm equation
(1)
Copper Mobility and Bioavailability in Sandy Soil 129
CR. c f;,2
+ Q 1,2 1
i(max,2) L - n
(K 2
C
') ),2
), )
where Qi and Qi max are the adsorbed quantity and the maximum adsorption of compo-
nent i (mol/kg), K
j
is the median of the affinity constants (K) of component;', where;'
includes component i. The parameter n reflects the nonideal behavior of component;,
and p reflects the intrinsic heterogeneity of the heterogeneous surfaces. By Cj we denote
the "free" concentration of component;' (mollL). Subscript 1 is considered to be related
to the first type groups (e.g., "carboxylic") and subscript 2 to the second type groups
(e.g., "phenolic"). Equation 1 has the advantage that it is valid over a wide pH range (2-
10) and free metal ion concentration range (10-
2
_10-
14
mollL) and that it can easily be
extended to the multimetal binding case.
At median pH ranges and low heavy metal concentrations the NICA equation can be
simplified to a Two Species Freundlich equation (TSF) if we take only proton competi-
tion with one heavy metal into account (Van Riemsdijk, 1994; Temminghoff et al., 1997)
(2)
where we defined
W = Q. (it )n; (K )nH(p-l)
1 l,max 1 H
(3)
and
(4)
In the case of heavy metal binding by sandy soils, where organic matter has chemically
the most reactive surface, Qi can be written as
(5)
where Q.soil is the adsorbed quantity of component i by soil (mol/kg soil), and fOM the
fraction organic matter of the soil (kg/kg).
PARAMETER ASSESSMENT SORPTION MODELS
As a model substance for the binding studies onto organic matter we used purified
humic acid that was extracted from forest floor material taken from the Tongbersven
forest (near Oisterwijk, NL). It has been used in previous studies and is described by
130 Fate and Transport of Heavy Metals in the Vadose Zone
Van Dobben et al. (1992) and Mulder et al. (1994). The upper 2 cm of the podzol B
horizon from the forest floor material was collected in plastic bags. Upon return to the
lab, the sample was homogenized, sieved 2 mm), and stored at 4C (field moist). The
humic acid extraction was carried out according to the recommendation of the Interna-
tional Humic Substance Society (IHSS). The obtained purified humic acid was freeze
dried and stored until use. Before use, each sample was dissolved in demi water. Data
for proton and copper binding by purified humic acid (dissolved organic matter) were
measured potentiometrically using a pH electrode and/or a copper ion selective elec-
trode (Cu-ISE) and a double junction calomel reference electrode, all connected to a
programmable Wallingford titrator (Kinniburgh et al., 1995). All experiments were per-
formed at 25C in 0.003 M NaN0
3
background electrolyte. The Cu-ISE was linear for
pCu between 3 and 13 (r2 = 0.9996), using ethylenediamine for low Cu activities (Avdeef
et al., 1983; Benedetti et al., 1995). Copper binding was measured at constant pH values
of 4, 6, and 8. To keep the pH at a certain level we used the pH stat mode of the titrator.
The total Cu concentration in solution was calculated by the buret additions. The Cu
binding by humic acid can be calculated by subtracting the Cu
2
+ from the total Cu con-
centration in solution. Above pH 6, Cu hydroxide species should also be taken into ac-
count (Benedetti et al., 1995). A detailed description has been given by Temminghoff et
al. (1997). Copper binding by purified humic acid is shown in Figure 6.1a for pH 4, 6,
and 8 on logarithmic scales. Copper binding depends nonlinearly on pH and Cu
2
+ con-
centration and is well described by both the (bimodal) NICA model (Eq. 1) and the TSF
model (Eq. 2). The fitted parameters are given in Table 6.1 for both models. The NICA
model described the Cu adsorption slightly better because it accounts for an adsorption
maximum. Proton binding by the purified humic acid is shown in Figure 6.1 b where the
relative charge is given as a function of the pH. The fit of the proton binding data by
NICA (solid line) was excellent (r2 = 0.99998). The fitted NICA parameters for protons
are similar to those determined by Benedetti et al. (1995) despite the fact that their
humic acid is from a different location and that they used a different ionic strength.
COPPER SPECIATION IN A COPPER CONTAMINATED SOIL
With the parameters derived for copper binding by purified humic acid, we tried to
describe copper desorption by a copper contaminated sandy soil, where the most impor-
tant sorption component is solid organic matter (in terms of reactivity and abundancy).
Soil samples were taken from an arable field near Wageningen in The Netherlands
(Wildekamp site) which has been used previously to study copper sorption and toxicity
effects on earthworms, nematodes, and collembola (Temminghoff et al., 1994, 1997;
Marinussen et al., 1997a; Korthals et al., 1996; Bruus Pedersen et al., 1997). The soil was
a Spodosol in a slightly loamy moderately fine sand. The field contains a randomized
block design of four copper concentrations and four pH adjustments in 6 X 11 m plots.
Four pH levels coded A-D for pH
KC1
between 4.0 and 6.1 were established in 1982 using
calcium carbonate or sulfur, followed by establishment of four total Cu levels coded 1-4
between 0 and 750 kglha, added as CUS04' at each pH-level (Lexmond, 1980). We
assumed that an equilibrium between the copper pollution and the soil has been more or
less established since the "pollution" occurred more than 14 years ago. Soil samples were
collected from the plots with the highest (code D) and lowest (code A) pH, for each pHI
Copper Mobility and Bioavailability in Sandy Soil 1 31
0.50
NICA
TSF
0.00
-0.50
~
CY
b.O -1.00
o
.....
o
+ +
-1.50
-12 -10 -8 -6 -4
log[Cu 2+] (moIlL)
4
0 data
3
HNICA
,-...
g
~
'-"
)
2
~
.c:::
()
~
.....
)
'0
o L-________ .-________ -. __________ .-________ -. ________ ~
2 4 6 8 10 12
pH
=igure 6.1. (a) Copper binding by DOC (purified humic acid) as a function of log(Cu
2
+) at pH 4, 6,
md 8 for data, the NICA and the TSF model at 1=0.003 (0.003 M NaN0
3
); (b) Proton binding by
)OC (purified humic acid), calculated as delta charge, as function of pH [-log(W)] for data and the
\lICA model at 1=0.003 (0.003 M NaN0
3
).
:::'u combination, at a depth of 0-0.20 m. All soil samples were air-dried and sieved 2
nm). Typical soil properties are given in Table 6.3. Neutral unbuffered salt solution
132 Fate and Transport of Heavy Metals in the Vadose Zone
Table 6.1. Parameters for the NICA and the TSF Model to Describe the
Copper Binding by Purified Humic Acid at I = 0.003
NICA TSF
Site 1 Site 2
log KH 5.12 9.78
ncu
0.36
log K Cu 5.40 10.82 m
H
-0.40
n
H
0.81 0.83 log Ki -0.348
ncu
0.46 0.46
P
0.57 0.58
Qcu,max (mol/kg) 2.18 1.54
NICA: H; coefficient of determination (r2) = 0.99998.
Cu and H; coefficient of determination (r2) = 0.98.
TSF: coefficient of determination (r2) = 0.97.
extractions are carried out with 0.001 M Ca(N0
3
)2 and 0.01 M CaCl
2
(soil: solution
ratio 1:10; wN). Three grams of air-dried soil was equilibrated with 30 mL 0.001 M
Ca(N0
3
)2 and 0.01 M CaCl
2
for 20 h in an end-over-end shaker. Soil pH was deter-
mined in this suspension with a glass-calomel electrode. Mter centrifugation at 2,000 g
for 15 min, Cu concentration in the supernatant (Cu
ex
) was measured on a flame atomic
absorption spectrometer (Instrumental Laboratory AN AE spectrophotometer S 11). Total
extractable soil Cu content (CUT) was determined by equilibration of 3 g of air-dried soil
with 30 mL 0.43 M HN0
3
for 20 h in an end-over-end shaker.
Since DOC coagulate in 0.01 M CaCI
2
, we used 0.001 M Ca(N0
3
)2 for the Cu mobil-
ity experiments. The concentration of copper extracted from the soil [Cu
ex
; 0.001 M
Ca(N0
3
hJ extraction varied from 0.18 0.04 to 10.3 0.2 J..lmollL and corresponds to
"free" copper (Cu
2
+), copper bound by dissolved organic matter (CuDOC) and copper
bound by inorganic species. The Cu binding by the inorganic species (e.g., N0
3
) is neg-
ligible. To assess the Cu
2
+ and CuDOC fractions, speciation techniques are required
(Temminghoff et aI., 1994). Assuming that Cu binding by dissolved organic matter be-
haves as Cu binding by humic acid, the "free" Cu
2
+ can be calculated at every pH, CU
ex
and DOC concentration by either the NICA or the TSF model. In agreement with
Kinniburgh et al. (1996), we assumed that Ca competition is negligible for the copper
binding by organic matter. Dissolved organic carbon concentrations for all the 0-0.20 m
top layer soil samples were equal (8.9 1.4 mg/L). The calculated pCu
2
+ concentrations
in the top layer varied from 5.1 (soil4A) up to 9.9 (soillD). For sandy soils, free copper
(Cu
2
+) has been shown to be in equilibrium with Cu bound by solid organic matter of the
soil and by dissolved organic matter of the soil solution (Temminghoff et aI., 1994). The
Cu
2
+ binding by the (soil) solid organic matter in the top layer 0-0.20 m was also de-
scribed by the NICA and TSF models where all parameters, except Qcu,max and the
intrinsic heterogeneity (p), were kept equal to the case of Cu
2
+ binding by the humic
acids (Table 6.2). By fitting only these two parameters for NICA and the two corre-
sponding parameters for TSF (Table 6.2) we described the Cu
2
+ binding by soil for the
layer 0-0.20 m (Figure 6.2). The maximum Cu sorption capacity for (soil) solid organic
matter had to be reduced to 30% of the Qcu,max of the purified humic acid for both the
TSF and the NICA model. This is plausible because the number of reactive sites (sites
Copper Mobility and Bioavailability in Sandy Soil 133
Table 6.2. Adjustable Parameters for the NICA and the TSF Model to
Describe the Copper Binding by Soil Organic Matter at I = 0.003
p
Qcu,max (mol/kg)
NICA TSF
Site 1 Site 2
0.50
0.65
0.50
0.46
NICA: CU and H; coefficient of determination (r2) = 0.97.
TSF: coefficient of determination (r2) = 0.93.
-0.58
-1.77
density) is probably smaller for solid organic matter than for dissolved organic matter
since solid organic matter is less humified (Gooddy et aI., 1995). For the NICA model
the factor p had to be adjusted from 0.57 and 0.58 to 0.50, which indicates larger intrinsic
heterogeneity of (soil) solid organic matter than purified humic acid. A change in Qcu,max
and parameter p gives a change in K/ and mH (see Eqs. 3 and 4) for the TSF model. The
NICA and the TSF model described the "free" copper desorption data excellently. Only
for mean pH 5.6 and Cu
2
+ concentrations of approximately 10-
10
moliL, the TSF model
shows slight discrepancies.
MOBILITY
DOC Mobility Enhanced Copper Mobility
To show the effect of DOC concentration on Cu mobility, column experiments were
carried out with top layer soils of plot 4A (1.610 mmollkg Cu total) and 3D (1.887 mmoli
kg Cu). We used perspex columns 4.5 cm in diameter and about 20 cm long which were
filled with 300 g of both soils. The top of the column was connected via a tube to the
influent solution (0.001 M Ca(N0
3
)2; pH 5.7). A fraction collector (ISCO fraction col-
lector) was used to collect the effluent samples of about 15 mL. Besides total Cu, we
measured DOC and the pH in the effluent. More detailed information is given by
Temminghoff et al. (1997).
In Figure 6.3 the total copper and DOC concentration and pH in the effluent of the
columns are given as a function ofleached pore volumes 0INo) for soil4Al, with mean
pH 3.9 (Figure 6.3a), and soil 3D 1, with mean pH 6.6 (Figure 6.3b). The effluent pH for
both columns is given in Figure 6.3c. The effluent pH differs from the soil pH deter-
mined by batch experiment, especially at large pH. At large pH the difference is about 1
pH, while at small pH the difference is negligible. The Cu breakthrough curve (BTC) at
pH 6.6 shows one Cu peak that occurs simultaneous with the DOC peak (Figure 6.3b).
At pH 3.9 two Cu peaks are visible, the first one at the same moment as the DOC peak,
followed by a second Cu peak that corresponds with the decline of pH during the first
pore volumes (Figure 6.3c). The Cu concentration at pH 3.9 reached a maximum of
approximately 25 IlmoliL Cu after 7 pore volumes and decreased slowly to 8 IlmoliL
after 45 pore volumes. At pH 6.6 the Cu concentration was 11 IlmoliL (after 1 pore
volume) and decreased fast below 5llmol/L after 5 pore volumes. Maximum DOC con-
centration in the effluent was found after 1 pore volume both for pH 3.9 (83 mg/L DOC)
and for pH 6.6 (120 mg/L DOC) and decreased fast to concentrations below 10 mg/L.
134 Fate and Transport of Heavy Metals in the Vadose Zone
-2.00
0
data

TSF
-2.50
.. NIeA
1'*
,-.,
pH 5.6
~ ~
"
bIl
~
~ , ~
-
-3.00
.. 0
S
pH 3.9
'-'
~ . ~
8
0 -3.50
bIl

0
-



c
c

-4.00
-4.50
-12 -10 -8 -6 -4
log[Cu 2+] (moIlL)
Figure 6.2. Log (copper binding by soil) as a function of log ("free" copper soil solution) at two
average soil pH levels; data, NICA and TSF model at 1=0.003 [0.001 M Ca(N0
3
H
:::J
;:::,
~
......,
C
"
::I
E
"
:;'
S:!.
40 ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~ 120
a) soil column 4A 1
0 Cu data
100
30
80
20 60
40
o .........
~ - ... --- ... ---------...:. . ............. . ... - ..,... ............... .
o ~ - - - - - - - - - - ~ - - - - - - - - - - - . - - - - - - - - - - - - ~ - - - - - - - - - - ~
o
40 o 10 30
:::J
'bb

U
0
Cl
Figure 6.3a. Copper and DOC concentration and pH in the effluent as a function of pore volumes
for (a) soil column 4A 1, (b) soil column 301, and (c) effluent pH for 4A 1 and 301 for 0.001 M
Ca(N0
3
h as influent (1=0.003). Line and dashed line is prediction of copper by NICA and TSF model,
respectively.
The solid and the dashed lines are the Cu concentration predictions by the NICA and
the TSF model in the effluent that take both pH and DOC effects into account. For the
NICA and the TSF calculations, Cu
2
+ desorption by soil organic matter and Cu
2
+ bind-
Copper Mobility and Bioavailability in Sandy Soil 135
------------ -- -----
:3
:::.
!
d
"
'"
!
:;-


120
30
20
10
0
0
8
7
..
6
4
3
2
o
..
.... ..
10
..
..... ..
b) soil column 301 0 Cu data

DOC data
CuNICA
CuTSF
w
20 30
VNo
c) pH effluent soil columns
..
......
....
.... ..
".... ...
... ... .., ......... ......... ...
.. .. ..
soil column 3Dl

... .........
100
80
60

20
0
.. ..
..
..
::J
E
8
... __ a;, .............. ., ..,v .V 99 Vq. V9V VV v v'" v

soil column 4Al
10 30 40
Figure 6.3b and c. Copper and DOC concentration and pH in the effluent as a function of pore
volumes for (a) soil column 4A 1, (b) soil column 3Dl, and (c) effluent pH for 4A 1 and 3Dl for 0.001
M Ca(N0
3
b as influent (1=0.003). Line and dashed line is prediction of copper by NICA and TSF
model, respectively.
ing by DOC was calculated at each effluent pH and DOC concentration with the use of
the parameters of Tables 6.1 and 6.2. The Cu effluent concentration is the sum of Cu
2
+
and CuDOC. The Cu effluent concentration is predicted rather well for both models.
The first peak for both pH values is slightly overestimated, possibly due to nonequilib-
rium during the first pore volume percolation in the columns. At pH 3.9 after about 10
pore volumes, approximately 70% of the total Cu in the effluent solution is Cu
2
+ and
about 30% is present as CuDOC, as calculated with the NICA and/or the TSF model.
During the first ten pore volumes, Cu
2
+ varied from 20 to 70% due to changes in DOC
136 Fate and Transport of Heavy Metals in the Vadose Zone
and pH. At pH 6.6 only a small part of total Cu is present as Cu
2
+ (0.01 %) and most is of
the form CuDOC (> 99%), which shows that the effect of DOC on Cu mobility is of
greater importance at higher pH than at lower pH. Quite different conditions than in
batch/titration studies are found in column experiments. Still, the model and parameter-
ization for a model substance (purified humic acid) yield an excellent description of Cu
in leachate. Provided DOC is mobile itself, we observe that facilitated Cu-transport
occurs that leads to faster Cu-Ieaching.
Field Site Accumulation in Soil
If the previous speciation modeling captures the main phenomena, not only for condi-
tioned laboratory experiments but also for field situations, we should be able to predict
the extractable copper concentration in the soil solution at different depth. From two
plots (4A and 3D), soil was sampled from the layers 0-0.20,0.20-0040,0040-0.60,0.60-
0.80, and 0.80-0.90 m. All soil samples were air-dried and sieved 2 mm). Dissolved
organic matter (DOC) varied as a function of depth in the field between 16 and 1 mg/L,
solid organic matter (SOC) content between 37.8 and 3.8 g/kg, pH between 3.83 and
5.91, and total Cu content between 1.89 and 0.013 mmollkg. In Figure 604 total Cu
content and pH (Figure 6Aa) and solid and dissolved organic matter (Figure 6Ab) are
given as a function of depth. Using parameters of Tables 6.1 and 6.2 for describing Cu
binding by DOC and Cu binding by (soil) solid organic matter (parameters only deter-
mined for the top layers 0-0.20 m), we predicted the extractable Cu concentration in the
soil solution at each depth between 0 and 0.90 m with the NICA and the TSF models. In
Figure 6.5 the predicted CU
ex
concentration is given as a function of the measured CU
ex
concentration. The agreement is good for both models. For the NICA model the predic-
tion was slightly better than the TSF model since the correlation coefficients were 0.97
and 0.93, respectively.
At the original moment of contamination (1982), soils of plot A and D with the same
number had the same (total) Cu content in the plow layer, but currently differences are
observed. For soil4A (small pH) the total Cu content in 1994 was already smaller than
in soil 3D (large pH) in layer 0-0040 m although the added Cu in 1982 was higher for soil
4A (750 kg/ha) than for soil 3D (500 kg/ha). However, below a depth of 0040 m the
copper content in soil 4A is much larger than in soil 3D in 1994. The extractable Cu
concentration (CueJ, determined via 0.001 M Ca(N0
3
)2 extraction and the models, is
for soil 4A (layer 0-0.20 m), about five times larger than for soil for 3D (large pH),
which illustrates the increase in Cu mobility at small pH.
BIOAVAILABILlTV
Bioavailability for Soil Organisms
Another objective of our research is to determine whether the mobile fraction corre-
sponds with the fraction that is available for organisms. In soil fertility as well as envi-
ronmental studies involving plant uptake it has been shown that the 0.01 M CaCl
2
extractable contents in soil is a better indicator of the fraction that is available for uptake
than total extractable contents (Novozamsky et al., 1993). Thus, it may be useful as an
indication of availability for soil dwelling organisms too. Marinussen et al. (l997a) ex-
,-..
eo
~
0
S
S
'-'
::I
u
Copper Mobility and Bioavailability in Sandy Soil 137
2.0 ,.------------------,
1.5
1.0
0.5
0.0
o
0-20
0-20 20-40 40-60 60-80 80-90
depth (em)
20
-15
-10
,
20-40 40-60 60-80
depth (em)
7
6
5
4
3
CJ Cu3D
.. Cu4A
:I::
p..
pH 3D
----
pH4A
- -EJ- - SOM 3D
;J' -0- SOM 4A
"&
.......
g --.-- DOC 3D
o
-e- DOC4A
Figure 6.4. Total copper content (mmollkg) and pH as a function of soil depth (Figure 6.4a) and
total organic matter and DOC as a function of soil depth for field plot 4A and 3D (Figure 6.4b).
posed earthworms (LumbriclM rubeLLw) to soil plots 2A, 2D, 3A, and 4 C that were sampled
from the Wildekamp site as described above (Table 6.3). The earthworms were exposed
under laboratory conditions for 1, 7, 14,28, or 56 days to study tissue Cu accumulation.
After sampling, earthworms were rinsed with distilled water and kept for three days in a
petri dish on moist filter papers to empty their gut. Earthworms were killed by immer-
sion in liquid nitrogen and dried in an oven at 105C. Earthworms were individually
digested in 5 mL 65% HN0
3
and 4 mL 20% H
2
0
2
. These solutions were analyzed for
eu on a furnace AAS.
138 Fate and Transport of Heavy Metals in the Vadose Zone
-4
] -5
u
t
;:s
-6
g
:
bO
. -7
o NICA
+ TSF
-8
4
log[Cu] (moVL) measured
Figure 6.5. Extractable copper concentration calculated with the TSF and the NICA model in the
Spodosol soil profile as a function of the measured extractable copper concentration for five layer
soil 4A and 3D up to 0.9 m depth at 1=0.003 [0.001 M Solid line is the 1:1 line.
Table 6.3. Several Characteristics of the Soil Used in the Experiments [pH CaCI
2
,
Clay (%), Organic Matter (% C), and 'Total" Cu with Standard Deviation (mmol/kglI
Plot Clay C CUT
Code pH-CaCI
2
(%) (%) (mmol/kg)
1A 3.80 2.9 1.98 0.42 0.02
2A 3.91 2.2 2.20 0.98 0.02
3A 3.84 1.7 2.16 1.69 0.05
4A 3.77 3.4 2.15 1.66 0.09
18 4.33 2.4 1.89 0.27 0.01
28 4.24 2.9 2.02 1.13 0.01
38 4.18 1.9 2.14 1.49 0.02
48 3.85 2.4 2.02 1.93 0.02
1C 4.73 5.0 2.26 0.38 0.01
2C 4.24 3.8 2.34 1.18 0.03
3C 4.18 3.3 1.87 1.77 0.01
4C 3.85 5.0 2.07 2.26 0.04
10 5.24 3.9 2.21 0.16 0.005
20 5.28 3.8 2.24 1.23 0.04
3D 4.95 4.5 2.18 1.77 0.01
40 5.12 3.3 2.10 2.64 0.06
Mean 3.3 2.11
Std 1.0 0.13
Copper Mobility and Bioavailability in Sandy Soil 139
Tissue Cu concentrations increased as a function of time and proportionally with total
extractable soil Cu content (r2=0.9). Neither the correlation between accumulation and
soil pH, nor between accumulation and CU
ex
(0.01 M CaC1
2
), however, was significant.
Marinussen et al. (l997a) observed a large mortality of L. rubeLLIM in soils that were
high in CU
ex
(0.01 M CaC1
2
). Their data enable us to demonstrate the relationship be-
tween mortality and extractable soil Cu content. Speciation of Cu in the soil solution
plays a possible role in mortality but this was not investigated. Therefore, we show in
Figure 6.6 the average mortality rate in the five contamination levels. It is obvious that
the average mortality rate correlates better to CU
ex
than to CUT. From these observations
we conclude that the exposure route for uptake differs from the exposure route that
causes mortality. The latter seems to correlate with Cu concentration in soil solution
which depends on total soil Cu content and soil pH, as illustrated by Equation 2.
Field Site Accumulation by Earthworms
Generally, soil contamination is spatially variable. Hence, in contaminated field sites,
exposure of soil-dwelling organisms to soil contamination varies as a function of time.
The larger the spatial variability is, the more variation in exposure should be anticipated.
Marinussen and Van der Zee (1996) showed that effects of such variation on accumula-
tion of contaminants in organisms depends on the degree of spatial variability and the
size of the home-range of the organism.
Earthworms are organisms with a limited home range. The mobility of earthworms is
influenced by the earthworms' ecology and by environmental conditions; e.g., soil hu-
midity, soil temperature, and food availability (Sims and Gerard, 1985). Mobility and
spatial variability are complicating factors for predicting accumulation of heavy metal in
earthworms exposed to spatially variable soil contamination. However, this may be a
matter of effort rather than of principle. The involved effort may be the reason why field
studies on effects of spatial variability of soil contamination on exposure of organisms
are rare. Marinussen et al. (l997a,b,c) studied heavy metal accumulation in earthworms
under both laboratory and field conditions. The main objective of their studies was to
determine whether data on heavy metal accumulation obtained by laboratory studies
can be used for predicting heavy metal accumulation under field conditions.
Marinussen et al. (l997a) introduced in each of the four differently contaminated
plots (Table 6.3), 500 specimens of the earthworm L. rubeLIm. To determine tissue Cu
accumulation under field conditions, earthworms were sampled at three times (14, 28,
and 70 days after introduction). In Figure 6.7, we show that the decline in tissue Cu
concentration between the first and the second sample time coincided with relatively low
soil temperatures. These low soil temperatures may have caused a downward migration
of L. ruheLLIM. As shown in Figure 6.4a, soil in the upper 40 cm layer is considerably more
contaminated than soil at greater depth. Hence, earthworms that move downward are
exposed to less contaminated soil and therefore accumulate less copper, which is in agree-
mentwith the data (Figure 6.7). Marinussen et al. (1997a) found that tissue Cu accumu-
lation was significantly correlated with Cup whereas it was correlated neither with soil
pH nor with CU
ex
(0.01 M CaCI
2
). They also observed large mortality in plots where soil
was high in CU
ex
These results are in agreement with the laboratory experiments, de-
scribed above.
Figure 6.6. Average mortality rate of earthworms exposed to soil samples obtained from the
Wildekamp site (see text for details) as a function of CUT (6A) or CU
ex
(0.01 M C a C I ~ (68).
"--------
1.5
temperature
Oil
~ 1.2
0
8
plot 4C ..
d
0.9 0
plot2A
0.g
15
0
0
plot 2D
d
0.7
0
0
;:l
U
-----
plot 3A
0
;:l
'"
'" '.0
0.1
Copper Mobility and Bioavailability in Sandy Soil 141
,---------------------------.10
8
6
4
2
~ - - - - ~ ~ T - - - - - - - - - - - - - - - ~ O
o 14 28 42 56
exposure time [days]
IT
o
........
i
8
-
'0
'"
Figure 6.7. Tissue Cu accumulation under field conditions in earthworms Lumbricus rubel/us
exposed to contaminated soil at the Wildekamp site. See text for details.
Marinussen et al. (1997b,c) exposed earthworms Dendrohaena veneta to heavy metal
(Cu, Pb, Zn) contaminated soils under both laboratory and field conditions. The soil
samples used in the laboratory experiments were obtained from a contaminated field site
in Doetinchem, NL [sandy loam soil, 7% clay, 3% organic matter (loss on ignition)].
After homogenization of the soil, subsamples were taken and analyzed for CUT (12.8
1.8 mmollkg), CU
ex
(0.01 M CaCI
2
; below detection limit = 0.6 IlmollL), and pH-CaCI
2
(7.0 0.06). Dendrohaena veneta were exposed to the soil for 1, 2, 3, 7, 14,28,56, or 112
days to study tissue heavy metal accumulation under laboratory conditions. To study
heavy metal excretion, D. veneta were transferred to uncontaminated soil after exposure
to the contaminated soil for 112 days.
Both accumulation and excretion of Cu appeared to be fast processes (Marinussen et
aI., 1997c). An equilibrium in tissue Cu accumulation was achieved 14 days after intro-
duction in the contaminated soil. Three days after being transferred to uncontaminated
soil, earthworms lost about 70% of the accumulated Cu. Lead was accumulated to a very
small extent, and Pb excretion stagnated at 40% 56 days after transferring to uncontami-
nated soil. Zinc was not accumulated. From these data, we conclude that tissue Cu con-
centration in D. veneta adapt rapidly to changes in exposure which are common in spatially
variable soil contamination.
In another study of Marinussen et al. (l997b), D. veneta were exposed for 14 days to
10 differently contaminated soil samples (field site in Doetinchem, NL) to determine the
relationship between soil Cu content and tissue Cu concentrations. They found that in
soil containing 0.16 to 1.57 mmollkg Cu, the earthworm tissue Cu concentration in-
creased proportionally to the total extractable soil Cu content (CUT)' Earthworms seemed
to achieve a maximum tissue Cu concentration (Figure 6.8).
ILl2 Fate and Transport of Heavy Metals in the Vadose Zone
1.4 r-------------------------.
j
c
o
'.0
t
u
::s
U
1.2
1
0.8
0.6
g 0.4
'"
'" '.0
0.2




o ~ - - - ~ - - - - ~ - - - - ~ - - - - ~ - - - - ~ - - - - ~ - - - - ~ - - - - ~
o 2 4 6 8 10 12 14 16
0.43 M HN0
3
extractable eu [mmol/kg]
Figure 6.S. Tissue Cu concentrations in earthworms Dendrobaena veneta exposed to Cu contami-
nated soil under laboratory conditions.
Additionally, Marinussen et al. (1997b) introduced about 100 specimens of D. veneta
at each of 20 homogeneously distributed locations in the field site in Doetinchem. The
spatial variability of soil Cu contamination in this field site was considerable (Figure
6.9). At three different times, earthworms were sampled and analyzed for tissue Cu
concentration (procedure is described above). For each earthworm, the Cu concentra-
tion factor was calculated (CFcu is the ratio tissue Cu concentration to the total extract-
able soil Cu content). An accurate estimation of the total extractable soil Cu content at
each location was obtained by geostatistical interpolation (in CadU disjunctive kriging).
They also calculated CF CU for earthworms exposed under laboratory conditions (de-
rived from Figure 6.S). It appeared that CF CU values under field conditions were in good
agreement with CFcu values under laboratory conditions (Figure 6.10). This successful
extrapolation from laboratory to field scale was a result of a high soil sampling density.
They took Sl soil samples in the top layer (0-20 cm) in an Sl m
2
experimental plot. The
large variation in CF CU under field conditions may be explained by a considerable de-
crease of soil Cu contamination as a function of depth at this site (Figure 6.11). Since D.
veneta moves up and down through the upper layer of soil, the latest exposure level is
uncertain as a result of this kind of spatial variability.
The field studies by Marinussen et aI. (1997 a,b) show that earthworm heavy metal
accumulation under field conditions can be predicted using relationships between soil
heavy metal contamination and tissue heavy metal concentrations determined under labo-
ratory conditions. However, a high soil sampling density is required to obtain accurate
estimations of exposure levels of individual specimens.
Copper Mobility and Bioavailability in Sandy Soil 143
Figure 6.9. Copper contamination at the field site in Doetinchem, NL. The vertical axis is the total
extractable copper in soil (mmollkg). Soil samples were taken from the top layer (0 to 0.20 m).
The soil sampling scheme was a squared 8 by 8 grid with 1.0 m node distance, resulting in 81
soil samples at 64 m2.
SUMMARY
We presented two models, the NICA and the TSF models that have been developed
to describe, among others, heavy metal sorption by soil. For both models, data on pH-
dependent copper binding by purified humic acid were fitted. The description was good
and the determined parameters are in agreement with literature for other humic acids
and conditions such as ionic strength.
For natural solid soil organic matter it is plausible that the reactivit;y is smaller, whereas
the sorption site heterogeneit;y is larger than for dissolved purified humic acid. Adapting
144 Fate and Transport of Heavy Metals in the Vadose Zone
2.5 r--------------------------,
2
1-

*
....
0
u
~
1.5 l-
e::
0
'.;::l

o:s
b
e::
(l.l
u
e::
1 I-
0
u
;::l
u
0.5 t-
0.43 M HN0
3
extractable eu in soil [mmol/kg]
Figure 6.10. Copper concentration factors in earthworms Dendrobaena veneta exposed to contami-
nated soil under laboratory conditions (circles) or field conditions (asterisks).
100
eil
~
0

~
;::1 10
u
(l.l
::0
CIl
....
g
~
(1)
,.,
0
~
:E
M
~
0
0.1
V IV VI III VII II VIII IX
Figure 6.11. Total extractable soil Cu content (CUT) in four consecutive layers of 5 cm thickness
at 9 spatially distributed locations in the Doetinchem field site; open = 0-5 cm, hatched = 5-10
cm, cross-hatched = 10-15 cm; fine-hatched = 15-20 cm.
Copper Mobility and Bioavailability in Sandy Soil 145
only the two parameters that are related with reactivity (sorption maximum) and het-
erogeneity, the two models describe pH-dependent copper binding by natural organic
matter also.
To further ascertain the applicability of the models, the agreement between model
prediction and measured data was considered for two rather different situations. The
first of these concerned the leaching of copper for two sandy topsoil columns that have
different pH and DOC levels. The other situation was the retention that is apparent in a
field soil for depths up to 0.90 m, with significant variations in dissolved and soil organic
matter, total copper, and pH. In both cases the agreement between model predictions
and observations was good. This indicates that for sandy soil, the two models capture the
main phenomena. Hence, with regard to both mobility and the chemical interpretation
of a neutral unbuffered salt extraction, an interpretation using the NICA and the TSF
models may improve our understanding.
The chemical speciation modeling may be necessary to be able to predict, e.g., copper
uptake by plants. However, the available information with regard to accumulation of
copper earthworms indicates that bioavailability for earthworms depends on the total
copper content rather than the fraction that can be extracted with a mild extractant. The
latter fraction does appear to control the short-term toxicity of copper for earthworms,
possibly due to oral uptake. Copper accumulation by earthworms may be controlled
mainly by dermal uptake and is not strongly related with short-term toxicity effects. As
the total copper levels in field soils are often spatially variable, copper accumulation in
field situations may be difficult to predict. For a field site, we showed that it is in prin-
ciple feasible to predict copper accumulation using observations from laboratory experi-
ments and a reliable map of total copper contents of the involved site. Unfortunately, to
obtain a reliable map may require a high density of soil sampling. In summary, we con-
clude that tools have been developed to translate laboratory data such that they have
relevance for soil in situ. To apply these tools for practical predictions of bioavailability
and mobility of heavy metals is currently under investigation.
REFERENCES
Aten, C.F. and S.K. Gupta. On heavy metals in soil; rationalization of extractions by dilute salt
solutions, comparison of the extracted concentrations with uptake by rye grass and lettuce, and
the possible influence of pyrophosphate on plant uptake. Sci. Tot. Environ. 178, pp. 45-53, 1996.
Avdeef, A., J. Zabronsky, and H.H. Stuting. Calibration of copper ion selective electrode
response to pCu 19. AnaL. Chem. 55, pp. 298-304, 1983.
Benedetti, M.F., C.J. Milne, D.G. Kinniburgh, W.H. Van Riemsdijk, and L.K. Koopal. Metal ion
binding to humic substances: Application of the non-ideal competitive adsorption model.
Environ. Sci. TechnoL. 29, pp. 446-457, 1995.
Bruus Pedersen, M., E.J.M. Temminghoff, M.p.J.e. Marinussen, N. Elmegaard, and C.A.M.
van Gestel. Copper uptake and fitness of Fouomia candUJa willem in a copper contaminated
sandy soil as affected by pH and soil moisture AppL. SoiL &oL. 2(6), pp. 135-146, 1997.
Gooddy, D.C., P. Shand, D.G. Kinniburgh, and W.H. Van Riemsdijk. Field-based partition
coefficients for trace elements in soil solutions. European J. SoiL Sci., 46, pp. 265-285, 1995.
Kinniburgh, D.G., C.J. Milne, and P. Venema. Design and construction of a personal-computer-
based automatic titrator. SoiL Sci. Soc. Am J. 59, pp. 417-422, 1995.
146 Fate and Transport of Heavy Metals in the Vadose Zone
Kinniburgh. D.G . C.J. Milne. M.F. Benedetti. J.P. Pinheiro. J. Filius. L.K. KoopaL and W.H.
Van Riemsdijk. Metal ion binding by humic acid: Application of the NICA-Donnan model.
Environ. Sci. Techno!. 30. pp. 1687-1698. 1996.
Koopal, L.K.. W.H. Van Riemsdyk. J.C.M. De Wit. and M.F. Beneditti. Analytical isotherm
equations for multicomponent adsorption to heterogeneous surfaces. J. Coll. Interface Sci. 66.
pp. 51-60. 1994.
Korthals. G.W . A.D. Alexiev. T.M. Lexmond. J.E. Kammenga. and T. Bongers. Long-term
effects of copper and pH on the nematode community in an agroecosystem. Environ. ToxicoL.
Chem. 15. pp. 979-985. 1996.
Lexmond. Th.M. The effect of soil pH on copper toxicity for forage maize as grown under field
conditions Neth. J. Agric. Sci. 28. pp. 164-183. 1980.
Marinussen. M.P.J.C. and S.E.A.T.M. Van der Zee. Conceptual approach to estimating the
effect of home-range size on the exposure of organisms to spatial variable soil contamination.
&ol. MOdelling. 87. pp. 83-89. 1996.
Marinussen. M.P.J.C . S.E.A.T.M. Van der Zee. and F.A.M. De Haan. Cu accumulation in
LumbriclM rubelllM under laboratory conditions compared with accumulation under field con-
ditions. &otox. Environ. Safety 36. pp. 17-26. 1997a.
Marinussen. M.P.J.C . S.E.A.T.M. Van der Zee. and F.A.M. De Haan. Cu accumulation in the
earthworm Dendrobaena veneta in a heavy metal (Cu. PB. Zn) contaminated site compared to
Cu accumulation in laboratory experiments. Environ. PoLL. 96(2). pp. 227-233, 1997b.
Marinussen. M.P.J.C .. S.E.A.T.M. Van der Zee. F.A.M. De Haan. L.M. Bouwman. and M.M.
Hefting. Heavy metal (Copper. Lead and Zinc) accumulation and excretion by the earth-
worm. Dendrobaena veneta. J. Environ. Qual. 26. pp. 278-284. 1997c.
Mulder. J . D. Van den Burg. and E.J.M. Temminghoff. Depodzolization Due to Acid Rain:
Does Aluminium Decomplexation Affect to Solubility of Humic Substances? In Humic Sub-
dtancN in the GWbal Environment and Implicationd on Humic Health. N. Senesi and T.M. Miano,
Eds . Elsevier Science. 1994. pp. 1163-1168.
Novozamsky. 1.. Th.M. Lexmond. and v.J.G. Houba. A single extraction procedure of soil for
evaluation of uptake of some heavy metals by plants Int. J. Environ. Anal. Chem. 55. pp. 47-58,
1993.
Sanders. J.R . S.P. McGrath. and Mc.M. Adams. Zinc. copper and nickel concentrations in soil
extracts and crops grown on four soil treated metal loaded sewage sludges. Environ. PoLL. 44,
pp. 193-2lO. 1987.
Sims. R.W. and B.M. Gerard. Earthwormd, Linnean Society Synopses of the British Fauna (Ne'w
Series) No. 31. London and Leiden. E.J. Brill/Dr. W. Backhuys. 1985.
Stevenson. F.J. HumUd ChemiJtryj GenNiJ, CompOdition, Reactiond. John Wiley & Sons. Canada.
1982. Ch. 14.
Temminghoff. E.J.M . S.E.A.T.M. Van der Zee. and M.G. Keizer. The influence of pH on the
desorption and speciation of copper in a sandy soil. Soil Sci. 158. pp. 398-408. 1994.
Temminghoff. E.J.M . S.E.A.T.M. Van der Zee. and F.A.M. de Haan. Copper mobility in a
copper contaminated sandy soil as affected by pH. solid and dissolved organic matter. Environ.
Sci. Techno!. 31(4). pp. 1109-1115. 1997.
Tipping. E . A. Fitch. and F.J. Stevenson. Proton and copper binding by humic acid: application
of a discrete-site/electrostatic ion-binding model. Eur. J. Soil Sci. 46. p. 95. 1995.
Van Dobben. H.F . J. Mulder. H. Van Dam. and H. Houweling. In Impact of AcwAtmodphm;
Depodition on the BiogeochemiJtry of Moorland Poou and Surrounding Terredtrial Environment. Pudoe
Scientific Publishers. Wageningen. 1992. Chapter 2.
Van Riemsdijk. W.H. Keynote Lecture. 15th World CongrNJ of Soil Science. Acapulco. Mexico; The
International Society of Soil Science. Madison. WI. Vol. 1, 1994. p. 46.
CHAPTER 7
Selenium Speciation in Soil Water:
Experimental and Model Predictions
Katta J. Reddy
INTRODUCTION
Selenium (Se) occurs naturally in soils. The main geological source of Se in soils is
cretaceous shales. The common range ofSe in soils is between 0.01 and 2 mg/kg-
1
(Lakin,
1972). However, in seleniferous soils Se concentrations can be as high as 1200 mg/kg-
1
(Adriano, 1986). Selenium is a required micronutrient for humans and animals. Its re-
quirement, however, for plants is not clearly understood. Human activities introduce Se
into soils in many ways. These include burning fossil fuels (coal), disposal of coal combus-
tion by-products, mineral extraction activities, and application of fertilizers (Nriagu, 1989).
Selenium as a naturally occurring element is gaining national and international atten-
tion because of its potential deficiency and toxicity problems to humans and animals.
For example, in China two types of Se human diseases, cardiomyopathy (Se deficiency)
and selenosis (Se toxicity) were reported (Yang et al., 1983). In another case, disposal of
agricultural drainage water into wetlands of Kesterson National Wildlife Refuge in Cali-
fornia, caused bioaccumulation of Se by plants, fish, waterfowl, and animals at levels
that were harmful (Ohlendorf, 1989).
In soil water Se may exist in different oxidation states. These include Se (+6), Se (+4),
Se (0), and Se (-2). Among these, the Se (+6) and Se (+4) oxidation states are thermody-
namically stable under the pH and redox conditions that are found in most soils (Elrashidi
et al., 1987). However, in low redox environments Se (0) and Se (-2) species may be
expected. The Se (+6) and Se (+4) oxidation states in soil water may be comprised of
free ions and complexes. These include SeO/-, HSe04-' H
2
SeO/, CaSe04o, MgSe040
and SeOl-, HSe03-' H2Se03o, CaSe030, and MgSe03o. Additionally, soil water con-
tains dissolved organic carbon (DOC) due to the plant, animal, and biological activity;
therefore, DOC-Se complexes are expected.
Very little information exists on the speciation of dissolved Se in soil water because it
is difficult to separate Se (+6) and Se (+4) oxidation states without destroying their
147
148 Fate and Transport of Heavy Metals in the Vadose Zone
natural distribution. However, research in surface and groundwater suggest that dis-
solved Se consists of not only Se042- and Se032- but also metal-Se complexes and DOC-
Se complexes (Siu and Berman, 1989; Cooke and Bruland, 1987; Tanzer and Heumann,
1991; Wang et al., 1994; Reddy et al., 1995a).
Similarly, we can expect different dissolved Se species in soil water because soil water
contains higher ionic strength than surface water or groundwater due to an increased
concentration of dissolved salts. Thus, isolation, extraction, and measurement of dis-
solved Se species in soil water are important. Such information may help in predicting
the fate (availability, toxicity, adsorption, and precipitation) and transport (mobility) of
dissolved Se species in soil vadose zones (Reddy, 1998).
To date, there is little documentation on the quantification and model verification of
dissolved Se speciation in soil water. Therefore, in this chapter we review procedures
proposed for the speciation of dissolved Se, compare experimentally measured dissolved
Se speciation with a model prediction, and identiry further research needs in the specia-
tion of Se in soil water. The emphasis will be placed on SeOl- because it is predominant
and more toxic than SeOl- in natural environments.
SPECIATION OF DISSOLVED Se
The dissolved Se speciation in an aqueous solution can be performed with analytical
methods and/or geochemical models. Several methods including hydride generation atomic
absorption spectrometry (HG-AAS), fluorometry, (FM), high pressure liquid chroma-
tography (HPLC), and ion chromatography (IC) are available for the speciation of Se.
Among these methods, HG-AAS is the most commonly used method for the speciation
of dissolved Se in aqueous solutions because it can detect as low as 1 pg L-
1
of Se.
The HG-AAS method measures dissolved Se in an aqueous solution as Se (+4), from
which Se (+6) and DOC-Se can be measured by altering the pretreatment steps as de-
scribed below (Cutter, 1978; Workman and Soltanpour, 1980). The concentration of Se
(+4) in a sample is measured by generating H
2
Se with a NaBH4 solution and 7 M HCl
(undigested). Another aliquot of sample is heated for 20 minutes at 85C with 7M HCl
to reduce Se (+6) to Se (+4). The concentration of Se in this solution is considered as the
sum of Se (+6) and Se (+4) (digested). Difference between the concentration of Se in
digested and undigested samples is considered as the concentration of Se (+6). The undi-
gested or digested samples do not include DOC-Se.
The total Se concentration in an aqueous solution is measured by oxidizing organic
matter with H
2
0
2
for 20 minutes at 85C and then digesting with 7M HCl for another
20 minutes at 85C. This total Se is the sum of Se (+6), Se (+4), and DOC-Se. The
difference between total Se and digested Se is considered as DOC-Se species. How-
ever, the concentrations of Se (+6) and Se (+4) in soil water, determined with HG-AAS,
may consist of SeOl- and SeOl- species and their solution complexes. If SeOl- and
Se032- can be isolated and extracted directly from soil water, then they can be measured
using HG-AAS.
A geochemical model (e.g., MINTEQA2, GEOCHEM, WATEQFC) could be used
to calculate the speciation of dissolved Se in soil waters. Chemical data such as dissolved
concentrations of cations and anions, pH, and redox potential of the soil water are re-
quired by these models to calculate the chemical speciation (i.e., activity or concentra-
Selenium Speciation in Soil Water: Experimental and Model Predictions 149
tion of free ions and complexes) by solving a series of mass-balance equations through
an iterative procedure.
However, models are based on the assumption of equilibrium; therefore, soil water
should be close to a steady-state condition. The mass-balance equations for each dis-
solved species should contain all possible solution species to ensure accurate calculation
of the speciation; omission of any significant solution species from the mass-balance
equation will cause overestimation of the activity of dissolved free ions. Reported values
for the equilibrium constants for solution species might vary and the constants for some
species that may be present are not known, thus the species cannot be included in the
model. All these factors could lead to the misinterpretation of the speciation of dissolved
chemicals in soil water. Excellent discussions on this topic are presented by Amacher
(1984), Baham (1984), and Sposito (1994).
Few studies have examined the speciation of dissolved Se in soil water (See et aI.,
1995; Fio and Fujii, 1980; Reddy et aI., 1995a), despite its importance in understanding
Se solubility, availability, toxicity, and mobility.
Reddy et al. (l99Sa) reported that increasing the pH of CuCl
2
solution containing
SeOl-, SeOl-, and sulfate (SOl-) by adding NaOH precipitates cupric oxide (CuO).
The zero point of charge (ZPC) for CuO occurs at pH 9.S. As illustrated in Figure 7.1,
CuO particles adsorb SeOl- and SeOl- at pH 6 and desorb them at pH 13. Reddy et al.
(l99Sa) successfully used this phenomenon and isolated SeO 4 2- and Se032- from ground-
water samples containing sol- and DOC concentrations greater than 10,000 and SO
mglL, respectively.
EXPERIMENTAL AND MODEL PREDICTIONS
Dissolved Se Speciation with CuO
The procedure for the comparison of experimentally measured Se speciation in soil
water with model predictions is outlined in Figure 7.2. Soil samples were extracted with
distilled-deionized H
2
0 after reacting for 24 hours on a mechanical shaker at 200 rpm
under the laboratory temperature. Each soil H
2
0 suspension filtered and each filtered
sample was divided into two subsamples. One subsample was analyzed for Se (+6), Se
(+4), and DOC-Se with the HG-AA in addition to major cations and anions, as well as
pH (Figure 7.2).
The other subsample was used for extracting SeOl- and SeOl- with the CuO par-
ticles. The experimental procedure to isolate and extract SeOl- and SeOl- species from
soil water involved adding CuO particles to solutions and lowering the pH to 6.0 and
reacting for 4 hours, followed by separating the CuO particles from solution and in-
creasing the pH to 12.S and analyzing solutions for SeOl- and SeOl- by HG-AAS. The
concentrations of metal-SeO/- and metal-SeOl- complexes in soil water were calcu-
lated by subtracting the concentrations of Se042- and SeOl- from Se (+6) and Se (+4)
concentrations, respectively. From SeOl-, SeOl-, metal-SeOl-, metal-Se032-, and
DOC-Se, the speciation of dissolved Se in soil water samples was calculated.
Dissolved Se Speciation with GEOCHEM
As discussed earlier, several geochemical models including GEOCHEM and
MINTEQA2 are available to model the speciation of dissolved trace elements in soil
150 Fate and Transport of Heavy Metals in the Vadose Zone
pH=6.0 pH- 13.0
Aqueous Solution
Figure 7.1. Illustration of seO/- and SeO/- adsorption and desorption mechanism by the (uO
particles in aqueous solution.
water. For this research the program GEOCHEM was used to calculate the speciation
of dissolved Se (free ions and metal-SeOl- and metal-SeO/- complexes) in soil water
because it computes the highest number of solution species. The pH and the concentra-
tions of cations and anions were used as input to the model to calculate the speciation.
For Se input, Se (+6) and Se (+4) concentrations (determined with HG-AAS) were used
without redox potential because SeOl- and SeOl- reduction and oxidation reactions
are very slow (Reddy et al., 1995b).
From the input of pH and concentration of Se (+6), Se (+4), cations, and other anions,
GEOCHEM calculates the concentration of free ionic species (e.g., SeOl- and Se032-)
and metal-SeO/- and metal-SeO/- complexes using the thermodynamic data of solu-
tion species. The thermodynamic data used to calculate metal-Se042- and metal-SeO/-
complexes in soil water are shown in Table 7.1. The concentration of free SeOl- and
SeO/- ionic species and metal-SeO/- and metal-SeO/- complexes predicted by the
GEOCHEM were compared with the CuO extraction method (Figure 7.2).
Comparison
Selected chemical data of soil water, which are used for the discussion, are presented
in Table 7.2. The pH of soil water ranged between 5.8 and 8.4. Total dissolved Se con-
centrations were between 11 and 162 pg L-
1
These concentrations are well below the
limit of quantification of Se analysis with IC (Blaylock and James, 1993). Soil water 1
contained high concentrations of dissolved Mg, Na, and DOC when compared with
other soil water samples. Soil water 1, 2, 5, and 6 contained higher concentrations of
dissolved Ca than soil water 3 and 4. Dissolved SO/- concentrations in soil water ranged
between 15 and 1666 mg L-
1

Dissolved Se analyses with HG-AAS are shown in Figure 7.3. It should be noted that
concentrations of Se (+6) and Se (+4) also include SeO/- and SeO/- species plus metal
complexes. The soil waters examined in this study were dominated by Se (+6) concen-
trations. The Se (+6) concentrations ranged between 5 and 136 pg L-
1
, whereas Se (+4)
concentrations ranged between < 1 and 7 pg L -I. The DOC-Se concentrations were found
between 1 and 19 pg L-
1
Similar distribution for dissolved Se species was observed by
Fio and Fujii (1990) in soil water from California.
Results from the isolation and extraction of SeOl- and SeO/- with CuO are pre-
sented in Figures 7.4 and 7.5. These results suggest that the removal of SeO/- ranged
between 70 and 83%, except soil water 4, when compared with Se (+6) concentrations
Selenium Speciation in Soil Water: Experimental and Model Predictions 151
Analyze pH, CatioDl, AniODS, Se (6+),
Se (+4), and DOC-Se
Isolate and Extract Selenate
and Selenite with CuO
Se Speciation with GEOCHEM
Input pH, Cations, ADions
Se (+6), and Se (+4)
Se Speciation
Selenate and Selenite Ions
Metal Selenate and Selenite
Complexes
Se Speciation with COO
Determine Metal Selenate Complexes
Selenite Complexes
Se Speciation
Selenate and Selenite Ions
Metal Selenate and Selenite
Complexes
Figure 7.2. Procedures for the dissolved Se speciation comparison between the CuO/HG-MS and
geochemical modeling.
Table 7.1. Metal-SeO/- and Metal-SeO/- Complexation Re-
actions used in the Thermodynamic Database of GEOCHEM
a
No. Reaction J(
1
2
3
4
Ca
2
+ + SeD 2- CaSeD 0
4 - 4
Mg2+ + SeD 2- MgSeD 0
4 - 4
Ca
2
+ + SeD 2- CaSeD 0
3 - 3
Mg2+ + SeD 2- MgSeD 0
3 - 3
a Sposito and Mattigod, 1980.
10
2
.
8
10
2
.
4
10
4
.
2
10
5
.
0
(Figure 7.4). The Se analysis of the Cu 0 supernatant solutions suggested that 17 to 30%
of Se042-was left in the solutions. However, for Se032-the removal rate is 100%, except
soil water 2, when compared with Se (+4) concentrations. For soil water 5 and 6, Se032-
concentrations were below the detection limit of 1 pg L-
1
(Figure 7.5). These results also
suggest that other anions such as sOi- and DOC did not interfere in the SeOi- and
Se032- removal process by CuO particles. If S04
2
- and DOC compete with SeOi- and
Se032- for adsorption sites, one would expect no adsorption of these species by the Cu 0
particles, because the ratio of S04
2
- and DOC to Se is very high. The results also suggest
that metal-SeOl- complexes are not significant.
The observed 70 to 83% removal of SeOi- by the CuO particles from the soil water
could be due to the presence of other Se (+6) species (e.g., MgSe040, CaSe040), which
may not be adsorbed by the CuO particles (Reddy and Gloss, 1993). For example,
Giordano et al. (1983) showed that formation of neutral complexes (e.g., CdCI
2
) lowers
152 Fate and Transport of Heavy Metals in the Vadose Zone
""---,,--,------- -"'-""""'---------."--- " ---------- " --------
Table 7.2. Selected Chemical Data of Soil Water (SW)a
Parameter SWl SW2 SW3 SW4 SW5 SW6
pH 7.5 7.9 8.4 8.0 6.1 5.8
Calcium 188 214 58 39 243 275
Magnesium 922 48 26 26 63 76
Sodium 745 31 36 29 18 31
Potassium 47 14 16 7.2 14 9.4
Sulfate 1666 166 16 15 960 1130
DOC 144 60 33 24 19 7.6
Selenium L-
1
162 21 13 11 69 120
a Units are mg L -1. Data for SW2, SW5, and SW6 adapted from Reddy (1998). Reprinted
with permission of John Wiley 8- Sons, Inc.
.i
1

Legend
Se(+6)
-",*-- Se(+4)
............. DOC-Se
o - - - .. ... ==I ................ -... -... -... -... -.. .....
SW1 SW3 SW4
Soil Water Samples
..............
SW5
Figure 7.3. Dissolved Se concentration in soil water as measured by HG .. AAS.
. ...
SW6
the concentration of Cd
2
+, which decreases the adsorption and increases the mobility of
Cd in sewage sludge amended soils (see also Mattigod et aI., 1979; Bowman and O'Connor,
1982; Elrashidi and O'Connor, 1982).
There may be a number of reasons why CuO particles adsorb both SeOi- and SeOl-
in the presence of other ions; however, the most possible reasons include:
On a time scale, the metal-SeO/- and metal-Se032- complexation reactions in
aqueous solutions are much faster than adsorption reactions of these species by
the CuO particles. Also, SeOl- and Se052- adsorption reactions by the CuO
particles are much faster than reduction and oxidation of these species in aque-
Selenlul1l Speciation in Soil Water: Experimental and ModeJ Predictions 153
D Initial
CuOMethod
SoN3 SlN4
Soil Water Samples
Figure 7.4. Extraction of Se (+6) from soil water with CuO method.
D Initial
CuOMethod
SW1 SW2.
Soil Water Samples
Figure 7.5. Extraction of Se (+4) from soil water with CuO method.
154 Fate and Transport of Heavy Metals in the Vadose Zone
- -----,-,,-- " " - - " " " - - - ~ - - . ' " - " - - - " -".,----""'---- - - - ~ - - - - " ' - " " - , ~ - - " " ~ - - , - - ' " - "" ---"'''''"-----,."---
Table 7.3. Speciation of Se in Soil Water with CuO/HG-AAS
a
Species SW1 SW2 SW3 SW4 SW5 SW6
pH 7.5 7.9 8.4 8.0 6.1 5.8
SeO/- 59 43 38 46 58 72
Se0
3
2
-
4 19 38 46 NS NS
Metal-SeO/- 25 19 8 NS
b
15 22
DOC-Se 12 19 16 9 27 6
Total 100 100 100 100 100 100
a Units are %. Data for SW2, SW5, and SW6 adapted from Reddy (1998). Reprinted
with permission of John Wiley a Sons, Inc.
b NS=not significant <1 %.
ous solutions. For example, less than 30 minutes are required for the adsorption
of Se species by the CuO particles in aqueous solutions, whereas Se oxidation
and reduction reactions take much longer times (Reddy et al., 1995b; Reddy,
1998).
The CuO particles are stable in the pH range of 6 and 13 and show low affinity
for SO/-.
The speciation of dissolved Se concentrations in soil water samples by CuO/HG-AAS
is presented in Table 7.3. These results show that total dissolved Se concentrations were
dominated by SeO/- (38 to 72%). The concentration of SeOl- was between d to 46%.
The DOC-Se complexes ranged between 6 and 27% and the concentrations of metal-
SeO/- complexes were between d to 25%. The DOC-Se concentrations can be further
separated into hydrophobic base, acidic, and neutral fractions by following the proce-
dure as outlined by Wang et al. (1994).
Comparisons between the CuO method and GEOCHEM calculations for SeO/- and
metal-SeO/- complexes are shown in Figures 7.6 and 7.7. The agreement between these
two methods for SeO/- and metal-SeO/- complexes is excellent. These results strongly
suggest that the CuO particles extracted only free SeO/- ions from the soil water. The
accurate determination of SeO/- and metal-SeO/- complexes in soil water is more im-
portant than SeOl-, because SeO/- is more toxic than SeOl- (Page and Bingham,
1986). Additionally, studies have shown that SeO/- is more mobile than Se032- in soils
(Ahlrich and Hossner, 1987; Balistrieri and Chao, 1987) because it is not adsorbed by
the soil mineral phases (Neal and Sposito, 1989).
The agreement for Se032- and metal-SeOl- complexes between the two methods was
poor. The CuO procedure predicted metal-SeOl- complexes as insignificant, because
almost 100% of Se (+4) was extracted from soil water (see Figure 7.5). GEOCHEM
predicted free ion Se032- concentration as less than 1 % of the total Se (+4) concentra-
tion and 99% as metal-SeOl- complexes. One possible explanation for this may be due
to the weak strength of metal-SeOl- complexes, which might have dissociated during
the Cu 0 extraction process.
Nevertheless, both experimental and modeling studies on dissolved Se strongly sug-
gest that soil water may consist of not only SeO/- and SeOl- ionic species but also
metal-SeO/- and SeOl- complexes. The results in this study also suggest that SeO,/-
Selenium Speciation in Soil Water: Experimental and Model Predictions 155
0.16 r----7"'"---------"'II::-----,
Legend
0.14
o Model ExperimeDtal
SW1 SW2 SW3 SW4 SW5 SW6
Soil Water Samples
Figure 7.6. Comparison between CuO method and GEOCHEM for seol- concentrations in
soil water.
60
Legend
50
Model ExperimeIIIal
"0'
<F.
"-'
40
..
j
!
~
~
CIl
1
SWI SW2 SW3 SW4 SW5 SW6
Soil Water Samples
Figure 7.7. Comparison between CuO method and GEOCHEM for metal-SeO/- complexes in
soil water.
may form strong complexes with divalent cations. In contrast, Se032- may form weak
metal complexes in soil water. These results have important implications for our under-
standing of Se biogeochemistry in soils. One such implication is that the formation of
strong metal-SeOl- complexes in soil water may increase their potential mobility in the
soil vadose zone. This may be the one of the reasons why dissolved SeOl- concentra-
156 Fate and Transport of Heavy Metals in the Vadose Zone
tions in soil water, surface water, and groundwater often dominate the dissolved Se
concentrations.
FUTURE RESEARCH
Currently, speciation of dissolved Se in soil water is not well understood. This is par-
tially due to the difficulty involved in separation of Se (+6) and Se (+4) species without
disturbing their natural distribution. However, quantitative description of dissolved Se
speciation is required for predicting the fate and transport of Se in the soil vadose zone.
Selective Se adsorption mechanisms by the CuO particles provided an approach for
the speciation of dissolved Se in soil water. Our results suggest that a significant portion
of dissolved Se in soil water is comprised of SeO/- and metal-SeO/- complexes, which
may not be adsorbed by the mineral surfaces, and thus move along with water flux in the
soil vadose zone. Further research to separate the dissolved Se associated with hydro-
philic base, acidic, and neutral fractions is needed. Finally, research to determine the
stability of metal-dissolved Se and DOC-Se complexes in soil water will be invaluable.
REFERENCES
Adriano, D. C. Trace Element" in the TerredtriaL Environment. Springer-Verlag, New York, pp. 329-
361, 1986.
Ahlrich, J.S. and L.R. Hossner. Selenate and selenite mobility in overburden by saturated flow.
J. Environ. QuaL. 16, pp. 95-98, 1987.
Amacher, M.C. Determination of ion activities in soil solutions and suspensions: Principal
limitations. SoiL Sci. Soc. Am. J. 48, pp. 519-524, 1984.
Baham, J. Prediction of ion activities in soil solutions: Computer equilibrium modeling. Soil Sci.
Soc. Am. J. 48, pp. 525-531, 1984.
Balistrieri, L. and T. Chao. Selenium adsorption by goethite. Soil Sci. Soc. Am. J. 51, pp. 1145-
1151, 1987.
Blaylock, M.J. and B.R. James. Selenite and selenate quantification by hydride generation-
atomic absorption spectrometry, ion chromatography, and colorimetry. J. Environ. Qua!. 22, pp.
851-857, 1993.
Bowman, R.S. and G.A. O'Connor. Control of nickel and strontium sorption by free metal ion
activity. Soi! Sci. Soc. Am. J. 46, pp. 933-936, 1982.
Cooke, T.D. and K.W. Bruland. Aquatic chemistry of selenium: Evidence of biomethylation.
Environ. Sci. Techno!. 21, pp. 1214-1219, 1987.
Cutter, G.A. Species determination of selenium in natural waters. Ana!ytica Chemica Acta. 98, pp.
59-66, 1978.
Elrashidi, M.A., D.C. Adriano, S.M. Workman, and W.L. Lindsay. Chemical equilibria of
selenium in soils: A theoretical development. Soil Sci. 144, pp. 141-152, 1987.
Elrashidi, M.A. and G.A. O'Connor. Influence of solution composition on sorption of zinc by
soils. Soi! Sci. Soc. Am. J. 46, pp. 1153-1158, 1982.
Fio, J.L. and R. Fujii. Selenium speciation methods and application to soil saturation extracts
from San Joaquin Valley, California. SoiL Sci. Soc. Am. J. 54, pp. 363-369, 1990.
Giordano, P.M., A.D. Behel, J.E. Lawrence, J.M. Soileau, and B.N. Bradford. Mobility in soil
and plant availability of metals derived from incinerated municipal refuse. Environ. Sci. Techno!.
17, pp. 193-198, 1983.
Selenium Speciation in Soil Water: Experimental and Model Predictions 157
Lakin, H.W. Selenium Accumulation in Soils and Its Absorption by Plants and Animals, in
GeochemicaL Environment in Re&ztion to HeaLth and DifeaJe, H.L. Cannon and H.C. Hopps., Eds.,
Geological Society of America, Boulder, CO, 1972.
Mattigod, S.V., A.S. Gibali, and A.L. Page. Effects of ionic strength and ion-pair formation on
the adsorption of nickel by kaolinite. C&zyd C&zy Mineral!. 27, pp. 411-416, 1979.
NeaL RH. and G. Sposito. Selenate adsorption on alluvial soils. Soil Sci. Soc. Am. J. 53, pp. 70-
74, 1989.
Nriagu, J.O. Global Cycling of Selenium, in Occurrence and Diftribution of Selenium, M. Ihnat, Ed.,
CRC Press, Boca Raton, FL, 1989.
Ohlendorf, H.M. Bioaccumulation and Effects of Selenium in Wildlife, SeLeniwn in AgricuLture and
the Environment, in L.W. Jacobs, Ed., Soil Science Society of America, Madison, WI, 1989.
Page, A.L. and F.T. Bingham. Availability and Phytotoxicity of Selenium to Crops in Relation
to Chemical Form and Concentration, in 1985-86 TechnicaL Progre.J.J Report, KK Tanji, Ed.,
Univ. of California Salinity Drainage Task Force, University of California Davis, California,
1986.
Reddy, KJ. and S.P. Gloss. Geochemical speciation as related to the mobility of F, Mo and Se
in soilleachates. AppL. Geochem. SuppL. I.J.Jue. 2, pp. 159-163, 1993.
Reddy, KJ., Z. Zhang, M.J. Blaylock, and G.F. Vance. Method for detecting selenium specia-
tion in ground water. Environ. Sci. TechnoL. 29, pp. 1754-1759, 1995a.
Reddy, KJ., M.J. Blaylock, G.F. Vance, and RB. See. Effects of Redox Potential on the
Speciation of Selenium in Groundwater and Coal Mine Backfill Materials, in Decade.J Later: A
Time for ReaJde.Jdment: Selenium: Mining, Rec&zmation and EnvironmentaL Impactd, G.E. Schuman
and G.F. Vance, Eds., ASSMR, Princeton, WV, 1995b.
Reddy, KJ. Selenium Speciation in Natural Waters, in Encyclopedia of EnvironmentaL AnaLYdif and
Remediation, RA. Meyers, Ed., John Wiley & Sons, Inc., New York, 1998, pp. 4291-4300.
See, RB., KJ. Reddy, G.F. Vance, A.A. Fadlelmawla, and M.J. Blaylock. Geochemical Pro-
cesses and the Role of Natural Organic Solutes on the Solubility of Selenium in Coal Mine
Backfill Aquifers, Powder River Basin. FinaL Report, Ahandoned CoaL Mine &zndd Re.Jearch Program,
Office of Surface Mining, Denver, CO, 1995.
Siu, K W.M. and S.S. Berman. The Marine Environment, in Occurrence and Diftribution of Selenium,
M. Ihnat, Ed., CRC Press, Boca Raton, FL, 1989.
Sposito, G. and S.V. Mattigod. GEOCHEM: A Computer Program for the Calculation of
Chemical Equilibria in Soil Solutions and other Natural Water Systems. The Kearney Foun-
dation of Soil Science, Univ. of California, Riverside, 1980.
Sposito, G. ChemicaL Equilibria and Kineticd in Soil!: Chapter 2. Oxford University Press, New York,
1994.
Tanzer, D. and KG. Heumann. Determination of dissolved selenium species in environmental
water samples using isotope dilution mass spectrometry. AnaL. Chem. 63, pp. 1984-1989, 1991.
Wang, D., G. Alfthan, and A. Aro. Determination of total selenium and dissolved selenium
species in natural waters by fluorometry. Environ. Sci. TechnoL. 28, pp. 383-387, 1994.
Workman, S.M. and P.N. Soltanpour. Importance of prereducing selenium (VI) to selenium (IV)
and decomposing organic matter in soil extracts prior to determination of selenium using
hydride generation. SoiL Sci. Soc. Am. J. 44, pp. 1331-1333, 1980.
Yang, G., S. Wang, R Zhou, and S. Sun. Endemic selenium intoxication of humans in China.
Am. J. CLin. Nutr. 37, pp. 872-881, 1983.
CHAPTER 8
Influence of Reducing Conditions on the
Mobility of Divalent Trace Metals in Soils
Philippe Cam bier and Rayna Charlatchka
INTRODUCTION
The behavior of trace elements in soils has been the subject of many studies at first
related to plant nutrition, and now often justified by the ecotoxicologic question raised
by their possible accumulation in soils and in the food chain. The importance of their
bioavailability remains, whatever trace elements are considered as nutrients or pollut-
ants. It is also important to consider their mobility, which can be defined as the ability to
change their speciation, and more concern is about mobility toward soluble or gaseous
species. In fact, many trace elements encountered as soil pollutants, particularly heavy
metals, are generally little mobile; then concern is more about the long-term behavior of
elements likely to accumulate in soils, or already concentrated in polluted sites.
The main factor that influences the behavior of trace elements and which can change
with time and soil use is the pH. Other important factors such as soil texture and consti-
tution appear more constant; however, physical properties, organic content, and redox
conditions, can also quickly change, or be changed. With respect to this latter influence,
trace elements must be separated between at least two groups. Some elements can change
their oxidation number in common soil conditions; they are mainly those occurring as
stable oxoanions (As, Mo, Se, V, etc.), Cr and Hg. The influence of redox conditions on
their speciation appears direct, and they will not be considered here.
Other environmentally important trace elements are divalent in common conditions.
For them, reducing conditions induce indirect effects on their speciation and mobility,
through changing pH and soil constituents. For example, Fe and Mn oxides, and or-
ganic compartments are affected by reducing conditions and affect the behavior of all
trace metals. Published studies on this topic often present overall effects and hypotheti-
cal mechanisms, so that little consistency is obtained: the mobility or the bioavailability
of Cd, Zn, etc., may increase, or decrease, with the redox potential. The present contri-
bution is to review and illustrate the redox processes which influence the mobility of
1::0
160 Fate and Transport of Heavy Metals in the Vadose Zone
divalent trace metals in soils, and to propose a few qualitative rules for understanding
and prediction.
CONTROVERSIAL STUDIES ON SOIL-PLANT SYSTEMS
Forno et al. (1975) showed that Zn uptake by rice was generally lowered during the
first weeks of flooding, without relationship to total soluble Zn which can increase at the
same time. Reddy and Patrick's (1977) experiments on soil suspensions showed that
soluble Cd and total uptake of Cd by rice decreased when redox potential Eh decreased,
while the opposite was observed for Pb. Iu et al. (1981) found Zn and Cu less soluble
and exchangeable when Eh decreased in a waterlogged brown soil; however, the varia-
tion of their availability to plants depended on the species (Iu et al., 1982). These authors
concluded that Co, like Mn and Fe, but not Cu and Zn, become more mobile and avail-
able under reducing conditions. Bjerre and Schierup (1985) also reported effects of wa-
terlogging on the chemical extractability and the plant availability of Zn, Cd, Pb, Cu, in
3 different soils: exchangeable Zn increased while the exchangeable fraction of the oth-
ers remained very low; however, the total plant uptake of these trace metals generally
decreased, although roots and leaves showed different patterns. Again from chemical
extractions, applied to flooded rice soils after adding organic matter, Ghanem and
Mikkelsen (1987) found that exchangeable Zn decreased with the redox potential while
Zn bound to oxides increased; curiously, Zn estimated bound to organic substances also
decreased. Using the same approach on an acid red soil loaded with Cd sulfate, Xiong
and Lu (1993) found Cd less exchangeable after flooding. Ahumada and Schalscha (1993)
applied chemical extractions to near neutral soils spiked or not with Cd and Cu, after
incubating soil suspensions in different conditions; they found that Cd extractability
varied with Eh, whereas no marked influence was observed for Cu. Brown et al. (1989),
performed flooding experiments on Luvisol and Cambisol with or without sludge amend-
ment; they concluded that reducing conditions generally increase or have little effect on
the phytoavailability and solubility of Cd and Ni, depending on plant species and metal
source.
Iu et al. (1981) interpreted that changing Fe and Mn oxides by reduction can increase
their adsorptive capacity for Cu and Zn. Ghanem and Mikkelsen also interpreted that
reduction can generate more amorphous solids that were able to sorb more Zn. Mter
Brown et al. (1989), the dissolution of Fe and Mn compounds can induce the release of
divalent trace metals in the aqueous phase but also the competition with them for plant
assimilation. Bjerre and Schierup (1985) added to these mechanisms that root function-
ing is affected when Eh and oxygen decrease. MandaI et al. (1987, 1988) attributed their
observations to the reductive dissolution of specific sorbents of Zn. It is often estimated
that reductive dissolutions of Fe and Mn oxides can release trace metals (Me Bride,
1989; Reddy and Patrick, 1977, for Pb). Alloway (1995) reported after Bingham et al.
(1976), that Cd uptake by rice from contaminated paddy soils is strongly decreased
when conditions are anoxic, due to precipitation with sulfide. He also estimated that
gleying processes could decrease adsorptive capacities of Cd due to lowering contents of
Fe and Mn oxides. In the same book, McGrath (1995) recalled that a good relationship
exists between exchangeable Ni, i.e., extracted by unbuffered salt solutions, and
phytoavailable Ni, and that both generally increase in poorly drained soils. Kiekens (1995)
Influence of Reducing Conditions on Mobility of Divalent Trace Metals in Soils 161
hardly considered the effect of reducing conditions on Zn behavior in soil-plant systems,
except to recall that much available Fe is antagonistic to Zn plant uptake.
The above review suggests that reducing conditions indirectly affect the mobility and
availability of divalent trace metals through several mechanisms: formation of insoluble
sulfides, release from sorption sites of reduced-dissolved Fe and Mn oxides, or increased
fixation by them after alteration, displacement of divalent trace metal from any exchange
site by increasing soluble Mn
2
+ and Fe
2
+, and antagonisms between the later and the
other cations for plant uptake. The last mechanism concerns only bioavailability of met-
als; the two later ones make the concentration of trace cation in the soil solution and its
plant availability vary in opposite directions: high concentrations of dissolved Fe
2
+ or
Mn
2
+ should increase concentrations of, e.g., Cd
2
+, Zn
2
+ in the soil solution but should
compete against their uptake by plants. The first conclusion is that it is difficult to clear
up at the same time the effects of reducing conditions on the solubility of divalent metals,
and on their bioavailability. So from this point, we will focus more on the chemical mo-
bility and processes than on bioavailability.
The attempts to follow the changing solid speciation of trace metals by chemical ex-
tractions encounter the usual limitations of this method (e.g., Tessier and Campbell,
1991) and a few others, as the possible occurrence of two oxidizable compartments, i.e.,
organic substances and sulfides. Another general chemical approach consists in study-
ing the possible solubility and redox equilibria. This approach has been more often ap-
plied and with more details to Fe and Mn. However, some important works concerned
trace divalent metals.
FORMATION OF INSOLUBLE SULFIDES AND
OTHER SOLUBILITY EQUILIBRIA
When S2- activity rises, formation of insoluble sulfide is expected with "heavy met-
als," "soft," and "borderline" cations according to the HSAB theory (Sposito, 1981).
This solubility mechanism, which can take place in strongly reducing conditions, was
often studied with sediments, and also in particular cultivated systems, i.e., paddy soils.
Sulfide formation can occur at redox potential between -0.1 and -0.2 V at near neutral
pH (Ponnamperuma, 1972; Patrick and Jugsujinda, 1992), i.e., pE around -3 (Lindsay,
1979; Sposito, 1989). pH also determines S2- activity since H
2
S dissociates at pH 7.
Bingham et al. (1976) grew rice on soil amended with CdS0
4
enriched sludge, flooded
or not, and found Cd less available and much less soluble at the end of flooded culture; at
the same time, sulfate became undetectable in solution of flooded samples so that the
results were explained by the formation of sulfide. Huaiman (1984) determined the ion
activity products of Cd and S2-, and other solutes, at equilibrium with anaerobic soil sus-
pension and concluded that CdS rather than CdC0
3
controlled the solubility of this metal.
Recently, Brennan and Lindsay (1996) with soils polluted by metallic ores, and Morse
and Arakaki (1993) with synthetic samples, have shown the importance of CdS, CuS,
NiS, PbS, ZnS, with respect to the solubility of heavy metals in highly reducing condi-
tions. They also underline the importance of ferrous sulfides either as sorbents for these
heavy metals or because they control S2- activity. Under intermediate reducing condi-
tions, obtained by the former authors by adding only organic matter to the soil, solubili-
ties of Cd, Pb, Zn, appeared controlled by carbonates or by reactions with iron oxides.
162 Fate and Transport of Heavy Metals in the Vadose Zone
The enhanced mobility of trace metals from oxidizing sulfides is clear, because this
oxidation and the precipitation of Fe (III) induce acidification, and sulfate compounds
are much more soluble than sulfides. This was often observed for sediments (e.g., Forstner,
1987; Tack et al., 1996; Altmann and Bourg, 1997). After Tack et al. (1996), the solubili-
ties of Cd, Cu, Pb, Zn, remain very low when sediments are kept flooded and reduced,
due to the presence of sulfides, but Co, Ni, and Mn seem always associated to carbon-
ates. Mter oxidation and induced acidity, the increasing solubilities of Cd, Cu, Pb, Zn,
become controlled by iron oxides and possibly organic substances.
Other works, without particular attention to reducing conditions, put in evidence the
role of phosphate in limiting the solubility of at least Pb (e.g., Joponyand Young, 1994;
Lindsay, 1979), and the role of Zn silicates at basic pH (BrUmmer et al., 1983). These
ligands are not produced by reduction like S2-; however, indirect effects of reduction
could be through the activity of these ligands which are known to be released by reduc-
tive dissolution of iron oxide (Balzer, 1982; Ponnamperuma, 1972).
So it is difficult to separate the role of trace metal sulfides and the reactivity of Fe and
Mn compounds-sulfides, and also oxides, carbonates (Forstner, 1987). Since Fe and
Mn compounds are involved in trace metal mobility, it is important to recall briefly how
they are ruled by precipitation-dissolution and redox phenomena. The solubility of Fe
can be determined by sulfides at very low Eh. In moderately reducing conditions, the
activity of Fe (II) and Fe (III) are controlled by pH and pE through iron oxide solubility
and the Nerst equation (Balzer, 1982). However, mixed Fe(II)-Fe(III) oxides and car-
bonates can also impose the final "stable" concentrations (Ponnamperuma, 1972). Schwab
and Lindsay (1983) from soil suspension studies found that Fe(I!) activity can be con-
trolled by FeC0
3
(siderite) or Fe3(OH)s. Recently, Trolard et al. (1996) showed that
another mixed solid, known when synthetic as green rust, can control Fe(II) in tempo-
rary hydromorphic soil horizons. The case of Mn appears still more complex with an
important role played by intermediate species involving the less stable valency 3 of Mn
(Bartlett and James, 1993); however, the more common situations deal either with poorly
soluble, but very reactive, Mn(IV) oxides, in oxidizing conditions, or with Mn(II), whose
the solubility is controlled by carbonate (Ponnamperuma, 1972; Balzer, 1982; Sadana
and Takkar, 1988; Sposito, 1989; Bartlett and James, 1993).
The very low activities, as for Fe
3
+ at common pH, and S2- in presence of much Fe
2
+,
often are not measured during the experiments, they are deduced from some chemical
equilibria. However precipitation-dissolution of definite compounds seem well estab-
lished mechanisms regarding the effects of Eh on trace metal mobility in soils. Kinetic
studies can also put in evidence some mechanisms. Charlatchka and Cambier (1996;
1998) recorded the chemical evolutions of suspensions of polluted soil; when pH 6.2 is
maintained, using only HCI without adding any oxidizing or reducing reagent, soluble
Cd, Pb, Zn, seemed more related to Mn and Fe evolutions than to any other solutes;
particularly, sOi- remained practically constant, so that it was unlikely that sulfide
appeared, despite the low Eh (Figure 8.1).
ROLE OF Fe AND Mn OXIDES AS TRACE METAL SORBENTS
The solubilities of Fe and Mn are ruled by redox and precipitation-dissolution pro-
cesses. The same consensus cannot be brought out for divalent trace metals except in the
Influence of Reducing Conditions on Mobility of Divalent Trace Metals in Soils 163
0,20 r--------------,-r:Z=-n-----, 1,6
0,15 Pb 1,2
~
Cl
~
E
Cl
.ri
0,10 0,8
E
D-
c:
"0 Cd
N
c:
as
"0
(,)
0,05 0,4
0,00 0,0
0 4 8 12 16
10 400
b)
8
300
~ ~
Cl
Eh
E
200
..
0 >
en E
"0
..o----oMn
~
c:
100
as
--
w
c:
--
::::?: 4
oj
0
u.
2
-100
0 -200
0 4 8 12 16
Time (days)
Figure 8.1. Evolution of solutes and redox potential in a soil:suspension (1:10) under nitrogen
atmosphere, at pH 6.2 controlled by HCI addition (cultivated layer polluted by atmospheric
industrial fallout).
case of S2- formation. The role of Fe(I!) in this case can be through controlling S2-
activity, or iron sulfides could act as specific sorbents. In less reducing conditions, it is
often assumed that Fe and Mn oxides intervene as specific sorbents of trace elements.
From several studies on soils or on soil-plant systems, it seems that divalent trace
metals can be fixed back by amorphous or altered iron oxides under reducing conditions
(Ghanem and Mikkelsen, 1987; Iu et al., 1981; Reddy and Patrick, 1977, for Cd). Stud-
164 Fate and Transport of Heavy Metals in the Vadose lone
ies carried out on more or less heterogeneous systems, like pot experiments, hardly sus-
tained that reoxidation in part or whole of analyzed samples was impossible. So
reoxidation-precipitation processes could explain partly decreased solubility, and in-
creased "oxide fraction," under, or rather after reducing conditions.
Many authors rather expect that reductive dissolution of Fe and Mn compounds de-
creases the soil sorption capacity of trace metals, but few presented detailed chemical
data for Fe, Mn, and other metals. Working with synthetic systems, Francis and Dodge
(1990) have shown that anaerobic processes can release heavy metals (Cd, Ni, Zn, and
to a lesser extent Cr, Pb) trapped by goethites. Such a process should increase the solu-
bility of divalent trace metals in soils as surely as it is ruled by adsorption on these oxides
rather than by other reactions. Recently, Chuan et al. (1996), using a device similar to
Patrick and Jugsujinda's and a sandy loam paddy soil highly contaminated by industrial
waters, put in evidence that Eh has less influence than pH on the solubility of Cd, Cu,
Pb, Zn, although lower Ehs do increase the solubility of these elements. The ranges of
variation were [3.3, 8] and [0.33, -0.1 V] for pH and Eh, respectively. Chuan et al.
(1996) reported the increasing solubility of Fe at one pH value (pH 3.3). Chuan et al.
concluded that the main mechanisms involved were the reductive dissolution of ferric
and manganic oxides. Charlatchka and Cambier (1998) incubated soil suspensions dur-
ing similar periods and imposed only pH. Different Eh values were reached at equilib-
rium, depending on reagents added (HC!, HN0
3
, or NaOH, and glucose between 0 and
3% by weight of soil). Dissolved Fe, Mn, Pb, Cd, and less systematically Zn, varied like
-Eh (Figure 8.2), whereas Ca was practically constant.
REDUCING PROCESSES CHANGE pH
The influence of pH on trace element solubility is well known. Either through solubil-
ity equilibria, or due to complexation by soluble and surface ligands, increasing pH,
within an ordinary range, decreases solubility and bioavailability of divalent trace met-
als. Most environmental redox reactions involve protons. More precisely, considering
half-redox reactions toward reduction, they consume H+ (e.g., Lindsay, 1979; Sposito
1981; Bartlett and James, 1993). However, only complete redox reactions actually oc-
cur, so that a right idea on the effect of reducing conditions on pH can be obtained only
if main reduced and oxidized substances are identified and the main redox reaction is
written.
In order to write a possible redox process, 2 half-redox reactions can be added with
the following conditions:
electron exchange must be balanced
the reaction must be favored thermodynamically
some physicochemical and biological conditions must be fulfilled.
The first condition is easily fulfilled by adding 2 half-redox reactions where always ex-
actly 1 electron is involved, like in many published lists, one toward reduction, the sec-
ond toward oxidation. For example, nitrate reduction can be theoretically obtained by
oxidation of organic carbon (written CH
2
0). This oxidation can be more or less com-
plete, leading to different end-products, and to consumption or production of protons:
Influence of Reducing Conditions on Mobility of Divalent Trace Metals in Soils 165
-,------------ ." ----- -- . -----
-- -- ---- " " - - " " - - - - - - ~ - --,-- - --"'" ,- -,----" -----------._-" --------
9 -550
-300
6
-
>
~
-50
e
bO
Gi
e
3
200
0 450
U U
U U:::;-
u
:s
~ ~
~
~ ~
~ <'! ~
~
'+
0
+
6 -550
-300
4
--
>
~
-50
e
bO
Gi
e
2
200
0 450
U U
U U:::;-
u
:s
~ ~
~
~ n ~
~
<'! ~ ~
+ ~
..
'+
~
Ca
210 -550
--"-
\
1\
~
Ik
V
r/
f\
-300
140
:e
-50
Gi
70
200
o
1\
450
Figure 8.2. Concentrations of dissolved metals and Eh values at pH 6.2, in soil suspensions after
14 days following different treatments: +%C indicates glucose addition; 4 first treatments under
nitrogen, 5th treatment under decarbonated air; pH controlled by NaOH or HN0
3
; except 4th one
(+HCI). Eh scale is reversed.
l66 Fate and Transport of Heavy Metals in the Vadose Zone
- --- ---,,- -----
0,3 -550
-300
0,2
>
......
~
-50
e
CI)
~ e
0,1
200
0 450
u
u u:::-
u
u
~
~ u
:E
~
~ oX
~ <'!
"'"
+'-'
'+
~
0,2 -550
0,15 -300
......
>
-50
e
~
~
CI) 0,1
e
0,05 200

450
u u:::-
:E
~
~ ~
~
..
3 -550
-300
>
-50
e
~
2
......
~
e
1
200

450
u u:::-
u
~
~ u
~ <; <;e
..
Figure 8.2 (continued). Concentrations of dissolved metals and Eh values at pH 6.2, in soil
suspensions after 14 days following different treatments: +%C indicates glucose addition; 4 first
treatments under nitrogen, 5th treatment under decarbonated air; pH controlled by NaOH or
HN0
3
; except 4th one (+HCI). Eh scale is reversed.
Influence of Reducing Conditions on Mobility of Divalent Trace Metals in Soils 167
or
(reactions written for a moderately acid environment, where CO
2
does not dissociate)
Keeping the rules for writing redox reactions in mind, it appears that reductive disso-
lutions consume more protons than the other half-redox reactions. In other words, pro-
cesses involving reduction-dissolution of Mn and Fe should increase pH, whatever the
electron donor. On the contrary, when solid sulfides oxidize and oxidation overpasses
the level of iron transformation, acidification is observed which is mainly due to oxida-
tion of FeS2 (pyrite) and precipitation of ferric oxides. The last important redox reaction
involving phase changes is the formation of CO
2
which partly sinks into Mn(II), Fe(II),
or Ca carbonate solids; this also produces protons. Formation of CO
2
does not occur
only during aerobic processes since organic carbon can be the electron donor also under
reducing conditions. The oxidation of carbon can be limited to the formation of some
organic acids, which generally also tends to acidification.
The effect of changing Eh on pH depends on the main complete redox reaction in-
volved, on the reagents and the final products, and finally also on the average pH be-
cause of the acid-base reactions. About this last point, acid-base reactions between
carbonate ions, or sulfide, ammonia, or organic species, Ponnamperuma (1972) esti-
mates that the former play the major role in relation to the importance of respiration.
This also appears from global microbial "respiration" equations written by Krumbein
and Swart (1983).
A slight acidification is often observed as a first step after flooding, which can be
attributed to aerobic processes: biological activity, enhanced by moisture and using trapped
O
2
, produces CO
2
. The next step uses nitrate as oxidant and soil organic matter. As long
as N0
3
- is present, no volatile organic acids and few ammonia form (Ponnamperuma),
and no methane (Christensen and Kjelsen, 1989). Whatever the factors that determine
the ratio N
2
1N
2
0 (Firestone et aI., 1979; Smith et aI., 1983), the consumption of H+ for
denitrification is the same (reaction 1 or 3 below; van Breemen et aI., 1983).
(3)
The more reducing reactions involving N -organic N and not nitrate or nitrite-can
lead to ammonia accumulation, the end-products being CO
2
, fatty acids, and possibly
CH
4
(Ponnamperuma, 1972). The exact balance of protons can vary.
After denitrification, when reducing conditions go on, reductive dissolutions of Mn,
then Fe, occur and, as already noticed, consume H+ and tend to increase the pH (Konsten
and van Breemen, 1994). However, particularly Mn(II) can precipitate as MnC0
3
and
this should in turn release H+ and limit the first effect on pH (Sadana and Takkar, 1988).
What can be summarized after Ponnamperuma (1972), Bartlett and James (1993), is
that hydromorphic conditions tend to shift soil pH toward neutrality. pH of acid soils,
particularly sulfate soils, increases when reduced (Konsten and van Breemen, 1994).
Basic soils, calcic or sodic, tend to acidify, because resultant evolution of pH depends
much on reactions with carbonate (Ponnamperuma, 1972).
168 Fate and Transport of Heavy Metals in the Vadose Zone
~ - - - - - - - ~ - ~ ~ - - - ~ ~ ~ - ~ - - - ~ ~ ~ - - -- ~ ~ - - ~ ~ - - - - - - - ~ - ~ - - - - .. ~ - - -
The interpretation of the effects of reducing conditions on trace metal solubility through
changing pH was present in several studies reviewed above (Bjerre and Schierup, 1985;
Francis and Dodge, 1990). This influence is obviously taken into account within the
solubility approach. Other examples are given below.
U sing soil columns water saturated or not, Welch and Lund (1987) found that Ni was
less mobile in saturated conditions, and that this was related to higher pH. Dutta et al.
(1989) reported that flooding samples from 26 alluvial rice fields increased pH and de-
creased DTP A extractable Cu and Zn; decrease in extractable Zn was correlated to pH
increase, among other variables (total carbonate, sulfide, DTPA-extractable Fe and Mn).
Dudley et al. (1986) followed nitrogen transformations, changes in pH, soluble carbon
and Cu, Ni, Zn, during the aerobic incubation of soils amended with anaerobic sludge. At
first, organic C, NH3 and NH4 + releases were accompanied by increasing pH and soluble
Cu organic complexes, then nilrification occurred, which led to decreasing pH and coher-
ently increasing soluble Cu and Zn. Chanmugathas and Bollag (1987) subjected columns
of 4 different soils spiked with Cd(N0
3
)2 to aerobic or anaerobic incubations. The evolu-
tion of pH during anaerobic incubations varied; however, Cd retention was generally con-
sistent with it; i.e., Cd retention was lower at low pH and increased when pH increased.
Flooding an undisturbed block of silt loam soil from a wheat field polluted by indus-
trial atmospheric fallout, Charlatchka et al. (1995) noticed that soluble Cd, Pb, and Zn,
varied more or less according to pH variations (Figure 8.3). Evolution of Mn and soluble
carbonate were also consistent to common observations. Then, the pH was controlled on
suspensions of the same soil, but different reagents favored different reactions involving
H+: anaerobic transformation of glucose can produce fatty acids (Figure 8.4), as in Glinski
et al. (1996), so that pH spontaneously decreased by more than one unit and NaOH was
used to stabilize the pH. Without glucose, different mineral acids were used, but more
nitric acid was needed as compared to HCl to maintain the same pH because denitrifica-
tion strongly occurred with HN0
3
and consumed protons. It is noteworthy that anaero-
bic transformations of organic substances containing much available carbon (about 1 %
by weight of soil) can lead to acidification.
ROLE OF SOLUBLE ORGANIC LIGANDS
The presence of soluble organic ligands (whatever their origin and the redox condi-
tions) should move the chemical equilibria for trace metals toward the aqueous phase and
it has been found that they can actually increase their mobility (e.g., Chanmugathas and
Bollag, 1988; Dudley et al., 1986; Dunnivant et al., 1992; Gerritse, 1996; Lamy et al., 1993;
Reddy et al., 1995). However, high concentrations of fulvic acids can also increase reten-
tion of metals by building mineral-organic complexes (e.g., Chubin and Street, 1981). The
final distribution and effects depend much on pH (Japenga et al., 1992; Lamy et al., 1993).
Under reducing conditions, several low mass organic acids can increase the solubility
of trace metals, the most abundant and persistent being acetic acid (Ponnamperuma,
1972). Dudley et al. (1986) and Francis and Dodge (1990), evoked the role of soluble
organic ligands, metabolic products, to explain increased mobility under anaerobic con-
ditions. One may also recall that complexed Zn or other trace element can be in solution
but not available to plants (Forno et al., 1975). Mter these authors, and Ponnamperuma,
anaerobic conditions in fields induce a concentration peak of acetic acid during the first
Influence of Reducing Conditions on Mobility of Divalent Trace Metals in Soils 169
- - - - - - - ~ - ~ - - ~ - - - - - - - - , , - ' " " - ..- , - ~ - "-" , ~ ~ - - - ~ - - - - - - - - - - -- - -
600 9
a)
8
~
400
Cl
7
E
I
C a.
0
6 .0
...
200 l
0
5
0 4
0 100 200 300 400
150 640
b)
Ca
~
480
Cl
100
~
E
Fe
c
_ ....
Cl
-- 320 E
:::2:
... ---------- "0
cO
c
50 ~ - - -
0
l
Q)
Mn
160
u. I
0 0
0 100 200 300 400
80 800
\Pb
c)
~
60
I
600
I
Cl I
~
:::L
I
.0
I
Cl
a.. 40 400 :::L
"0
C
C
l
N
"0
20 200
0
0 0
0 100 200 300 400
Time (days)
Figure 8.3. Evolution of waters draining from a flooded undisturbed block of soil (from the same
cultivated layer as in Figure 8.1).
weeks. Enhancement of anaerobic incubations of soil by adding 1 % glucose produced
similar phenomena, or, with 3% glucose, stabilized high levels of butyric and acetic acids
(Figure 804) (Charlatchka et aI., 1995). With respect to the formation of soluble com-
plexes with trace metals, acetate appeared important since in the experiment of Figure
804, about 40% of soluble Cd should have been complexed by acetate, and 70% of Pb
(Figure 8.5). Despite the competition with major cations, particularly Ca, at the end of
170 Fate and Transport of Heavy Metals in the Vadose Zone
1400 ~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~
1200
~
~ 1000
c:
0
800
.0
....
C'tl
()
0
600
'c
C'tl
Cl
....
400
0
200
Toe
Sum e (Ac+But)
e (Butyric Acid
e (Acetic Acid
e (Glucose
o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
o 4 8 12 16
Time (days)
Figure 8.4. Evolution of organic solutes during anaerobic incubation of a soil:suspension with 3%
glucose (by weight of soil).
flooding experiments of Figure 8.3, about 20% of Cd was complexed by acetate after
chemical analyses and calculated speciation.
TRANSFORMATION OF INSOLUBLE ORGANICS
We presented with enough accuracy the influence of major soil mineral transforma-
tions under reducing conditions on trace metal behavior. Also, organic substances are
deeply affected when anaerobic processes take place instead of aerobic ones. Unfortu-
nately, there is much less accuracy about these transformations and still less on how they
affect trace metals. The previous part presented only some processes concerning the
aqueous phase.
The retention properties of natural insoluble organic substances for trace elements,
particularly Cu and Pb among divalent trace metals, have been extensively studied. How-
ever, there is no universal modeling of involved reactions, like for solubility equilibria.
Besides, the organic substances are expected to change in nature and reactivity. The chemi-
cal extraction approach provided some information: after adding organic matter and flood-
ing rice soils, Ghanem and Mikkelsen (1987) found that the organic fraction of Zn
progressively decreased. Iu et al. (1981) also estimated that solid organic fractions of Zn
and Cu decreased under hydromorphic conditions, whereas soluble organic Cu increased.
Following a more global approach, and focusing on biological processes, Francis and
Dodge (1990) noticed that Pb was not remobilized by anoxic microbial activity like other
trace metals from a synthetic iron oxide trap. They interpreted that Pb was fIxed back by
biosorption. Chanmugathas and Bollag (1987) attempted to clear up the role of microbial
activities, aerobic or anaerobic, on the retention of Cd (added as nitrate) by soil materials.
Different trends were observed at the beginning of their experiments or later, and de-
pending on soil properties; anyway, by comparison with sterile materials, microbes gen-
erally increased the retention of Cd, through lower pHs and by more direct intervening.
Influence of Reducing Conditions on Mobility of Divalent Trace Metals in Soils 171
100% -,----
80%
60%
40%
20%
0% +----
Cd Pb Zn
free cation inorganic complexes 0 acetate complexes III other organic complexes
Figure 8.5. Calculated speciation of dissolved trace metals in suspension incubated after glucose
amendment (see Figure 8.4).
Charlatchka (1996) used a particular extraction technique to put in evidence the role
of biomass on trace metal fixation, based on fumigation with chloroform and aqueous
extraction. This technique is recommended to quantifY C from the biomass (Vance et aI.,
1987). Applied at the same time on Zn and Cd, it showed that Zn had been partly fixed
back during some anaerobic incubation by favored biomass.
SUMMARY
When reducing conditions occur in soils, several processes start which modifY soil
components and the mobility of trace elements, even trace metals which practically keep
the same valency. The intensity of the various processes depend on starting conditions;
their sequence, if reducing conditions last, is rather constant, except for organic sub-
stance transformations. Obviously, soil organic matter, and organic amendments, are
heterogeneous and intervene in many redox reactions, at different times. The influence
of some redox processes on trace metal behavior can be qualitatively predicted, follow-
ing the sequence below:
A first step after limiting oxygen, or rather after water saturation, is often observed:
pH decreases and soluble metals increase. Then, nitrate and Mn oxides begin to be
reduced, and consume protons. When nitrate is abundant, denitrification producing
different gas generally consumes protons, and metal solubility should decrease. How-
ever, pH increase can be limited by other H+ transfer. If much available C is present,
like in case of recent organic amendment, formation of weak acids, including CO
2
,
dominates redox processes. Dissociated acids decrease the pH, and despite being also
172 Fate and Transport of Heavy Metals in the Vadose Zone
rather weak ligands for trace metals, they increase their solubility by complexation and
induced acidity.
When both nitrate and organic carbon are available, competing effects occur. pH
variation should be determined by their ratio and by final pH, as it appears in reactions
written by Krumbein and Swart (1983). Another process becomes detectable, the reten-
tion by developing anaerobic organisms of a few trace elements. Unless soils contain
much nitrate, Mn reduction quickly intervenes, and often seems to come with more
soluble trace metals. This is certainly explained by the high retention capacity of Mn
oxides for trace cations, and by the limited increase of pH, if any, which occurs. H+
consumption is limited because dissolution concerns a small amount of reactive Mn ox-
ides, which can partly precipitate as carbonate of low reactivity. Again other H+ trans-
fers can intervene, like mineralization of organic matter to reduce Mn, and Mn
2
+/H+
exchange in soil-plant systems. Reduction of iron takes more time and more electron
donor, and certainly can release more trace metals adsorbed or trapped by ferric oxides.
However, the expected increase of pH should favor the precipitation of these metals or
their fixation back on other soil constituents.
Under strong reducing conditions, sulfide formation immobilizes most divalent trace
metals. This process could appear as the more irreversible in some cases of permanent
flooding because it involves stable pure minerals, and the sequential picture given above
can appear as concerning short-term influence of reducing conditions. It is true that
redox processes in soils are generally cyclic, so that the mechanisms described are also
temporary. However, anaerobic-aerobic cycles certainly result in some translocation of
trace elements. The resultant of one cycle is not always null because soils are open sys-
tems, and there are examples of dramatic evolution regarding soil acidification and chem-
istry (van Breemen et aI., 1983). Other contributions seem necessary for improving our
knowledge of long-term effects of redox cycles on trace metal behavior.
REFERENCES
Ahumada, I. T. and E.B. Schalscha. Fractionation of cadmium and copper in soils. Effect of redox
potential. Agrochimica .37, pp. 281-289, 199.3.
Alloway, B.J. Cadmium, in HearyMetau in SOlM, 2nd ed., B.J. Alloway, Ed., Blackie Academic
& Professional, London, 1995, pp. 122-151.
Altmann, RS. and A.C.M. Bourg. Cadmium mobilisation under conditions simulating anaerobic
to aerobic transition in a landfill leachate-polluted aquifer. Water Air Soil Potfut. 94, pp . .385-
.392, 1997.
Balzer, W. On the distribution of iron and manganese at the sediment/water interface: Thermo-
dynamic versus kinetic control. Geochim. Co.Jmochim. Acta 46, pp. 115.3-1161, 1982.
Bartlett, RJ. and B.R James. Redox chemistry of soils. Ad!'. Agron. 50, pp. 151-208, 199.3.
Bingham, F.T., A.L. Page, RJ. Mahler, and T.J. Ganje. Cadmium availability to rice in sludge-
amended soil under "flood" and "nonflood" culture. Soil Sci. Soc. Amer. J. 40, pp. 715-719, 1976.
Bjerre, G.K. and H.H. Schierup. Influence of waterlogging on availability and uptake of heavy
metals by oat grown in different soils. Plant Soil 88, pp. 45-59, 1985.
Brennan, E.W. and W.L. Lindsay. The role of pyrite in controlling metal ion activities in highly
reduced soils. Geochim. Co.Jmochim. Acta 60, pp . .3609-.3618, 1996.
Brown, P.H., L. Dunemann, R Schulz, and H. Marschner. Influence of redox potential and plant
species on the uptake of nickel and cadmium from soils. Z Pjlanzenerniihr. Bodenl.:. 152, pp. 85-
91, 1989.
Influence of Reducing Conditions on Mobility of Divalent Trace Metals in Soils 173
Brummer, G., K.G. Tiller, U. Herms, and P.M. Clayton. Adsorption-desorption and/or precipi-
tation-dissolution processes of zinc in soils. Geooerma 31, pp. 337-354, 1983.
Chanmugathas, P. and J.M. Bollag. Microbial mobilization of cadmium in soil under aerobic and
anaerobic conditions. J. Environ. QuaL. 16, pp. 161-167, 1987.
Chanmugathas, P. and J .M. Bollag. A column study of the biological mobilization and speciation
of cadmium in soil. Arch. Environ. Contam. ToxicoL. 17, pp. 229-237, 1988.
Charlatchka, R Mobilite des metaux dans un sol contamine. Influence des conditions reductrices.
These de l'Universite de Paris 12, 1996.
Charlatchka, R, P. Cambier, and S. Bourgeois. Mobilization of Trace Metals in Contaminated Soils
Under Anaerobic Conditions, in Contaminateo Soi&, Third Intern. ConE. Biogeochem. Trace
Elements, Paris, May 15-19, 1995, R Prost, Ed., INRA Editions, Paris, 1997, pp. 159-174.
Charlatchka, Rand P. Cambier. Influence of Reducing Conditions on Solubility of Trace Metals
in Contaminateo Soi&. Accepted for Water Air Soil Pollut. 1998.
Christensen, T.H. and P. Kjelsen. Basic Biochemical Processes in Landfill, in Sanitary Lano/iLLing:
Proce.Jd, Technology ano EnvironmentaL Impact, 1989, pp. 29-49.
Chuan, M.C., G.y' Shu, and J.C. Liu. Solubility of heavy metals in a contaminated soil: Effects
of redox potential and pH. Water Air Soil PoLLat. 90, pp. 543-556, 1996.
Chubin, RG. and J.J. Street. Adsorption of cadmium on soil constituents in presence of
complexing ligands. J. Environ. Qua!. 10, pp. 225-228, 198!.
Dudley, L.M., B.L. McNeaL and J.E. Baham. Time-dependent changes in soluble organics,
copper, nickeL and zinc from sludge amended soils. J. Environ. Qua!. 15, pp. 188-192, 1986.
Dunnivant F.M., P.M. Jardine, D.L. Taylor, and J.F. McCarthy. Cotransport of cadmium and
hexachlorobiphenyl by dissolved organic carbon through columns containing aquifer material.
Environ. Sci. Techno!. 26, pp. 360-368, 1992.
Dutta, D., B. MandaL and L.N. MandaI. Decrease in availability of zinc and copper in acidic to
near neutral soils on submergence. Soil Sci. 147, pp. 187-195, 1989.
Firestone, M.K., M.S. Smith, RB. Firestone, and J.M. Tiedje. The influence of nitrate, nitrite,
and oxygen on the composition of the gaseous products of denitrification in soil. Soil Sci. Soc.
Am. J. 43, pp. 1140-1144, 1979.
Forno, D.A., S. Yoshida, and C.J. Asher. Zinc deficiency in rice. I. Soil factors associated with
the deficiency. Plant Soil 42, pp. 537-550, 1975.
Forstner, U. Changes in Metal Solubilities in Aquatic and Terrestrial Cycles, in MetaL Speciation,
Separation, ano Recovery, J.W. Patterson and R Passino, Eds., Lewis Publishers, Boca Raton,
FL, 1987, pp. 3-26.
Francis, J.A. and C.J. Dodge. Anaerobic microbial remobilisation of toxic metals coprecipitated
with iron oxide. Environ. Sci. TechnoL. 24, pp. 373-378, 1990.
Gerritse, RG. Dispersion of cadmium in columns of saturated sandy soils. J. Environ. QuaL. 25, pp.
1344-1349, 1996.
Ghanem, S.A. and D.S. Mikkelsen. Effect of organic matter on changes in soil zinc fractions
found in wetland soils. Comman. Soil Sci. Plant Anal. 18, pp. 1217-1234, 1987.
Glinski J., K. Stahr, Z. Stpeniewska, and M. Brzezinska. Changes of redox and pH conditions
in a flooded soil amended with glucose and manganese oxide or iron oxide under laboratory
conditions. Z. P/Lanzenerniihr. Booenlc. 159, pp. 297-304, 1996.
Huaiman, C. Studies on kinetics and equilibria of CdC0
3
and CdS in soil solution. Acta PeooLogica
Sinica 21(3), pp. 258-267, 1984.
Iu, K.L., I.D. Pulford, and H.J. Duncan. Influence of waterlogging and lime or organic matter
additions on the distribution of trace metals in an acid soil. II. Zinc and copper. Plant Soil 59,
pp. 327-333, 198!.
Iu, K.L., I.D. Pulford, and H.J. Duncan. Influence of soil waterlogging on subsequent plant
growth and trace metal content. Plant Soil 66, pp. 423-427, 1982.
174 Fate and Transport of Heavy Metals in the Vadose Zone
Japenga, J., J.W. Dalenberg, D. Wiersma, S.D. Scheltens, D. Hesterberg, and W. Salomons.
Effect of liquid animal manure application on the solubilization of heavy metals from soil.
Intern. J. Environ. AnaL. Chem. 46, pp. 25-39, 1992.
Jopony, M. and S.D. Young. The solid-solution equilibria oElead and cadmium in polluted soils
previously treated with sewage-sludges. Plant Soil 132, pp. 179-196, 1994.
Kiekens, L. Zinc, in Heavy Metau in Soiu, 2nd ed., Blackie Academic & ProfessionaL London,
1995, pp. 284-305.
Konsten, C.J.M., N. van Breemen, S. Supping, LB. Aribawa, and J.E. Groenenberg. Effects of
flooding on pH of rice-producing, acid sulfate soils in Indonesia. Soil Sci. Soc. Am. J. 58, pp.
871--883, 1994.
Krumbein, W.E. and P.K. Swart. The microbial carbon cycle, in Microbial Geochemidtry, W.E.
Krumbein, Ed., Blackwell Scientific, London, 1983, pp. 5-62.
Lamy, I., S. Bourgeois, and A. Bermond. Soil cadmium mobility as a consequence of sewage
sludge disposal. J. Environ. Qual. 22, pp. 731-737, 1993.
Lamy, I., P. Cambier, and S. Bourgeois. Pb and Cd complexation with soluble organic carbon
and speciation in alkaline 'soilleachates. Environ. Geochem. Health 16, pp. 1-16, 1984.
Lindsay, W.L. Chemical Equilibria in Soiu. John Wiley & Sons, New York, 1979.
MandaL L.N. and B. MandaI. Transformation of zinc fractions on rice soils. Soil Sci.. 143, pp. 205-
212, 1987.
MandaL B., G.C. Hazra, and A.K. Pal. Transformation of zinc in soils under submerged condition
and its relation with zinc nutrition of rice. Plant Soil 106, pp. 121-126, 1988.
McBride, M.B. Reactions controlling heavy metal solubility in soils. Adv. Soil Sci.. 10, pp. 1-35, 1989.
McGrath, S.P. Chromium and Nickel, in Heavy Metau in Soiu, 2nd ed., Blackie Academic &
Professional, London, 1995, pp. 152-177.
Morse, J.W. and T. Arakaki. Adsorption and coprecipitation of divalent metals with mackinawite
(FeS). Geochim. COdmochim. Acta 57, pp. 3635-3640, 1993.
Patrick, W.H., Jr. and A. Jugsujinda. Sequential reduction and oxidation of inorganic nitrogen,
manganese, and iron in flooded soil. Soil Sci. Soc. Am. J. 56, pp. 1071-1073, 1992.
Ponnamperuma, F.N. The chemistry of submerged soils. Adv. Agron. 24, pp. 29-96, 1972.
Reddy, C.N. and W.H. Patrick, Jr. Effect of redox potential and pH on the uptake of cadmium
by rice plants. J. Environ. QuaL. 6, pp. 259-262, 1977.
Reddy, K.J., L. Wang, and S.P. Gloss. Solubility and mobility of copper, zinc, and lead in acidic
environments. Plant Soil 171, pp. 53-58, 1995.
Sadana, U.S. and P.N. Takkar. Effect of sodicity and zinc on soil solution chemistry of manga-
nese under submerged conditions. J. Agric. Sci. Camb. Ill, pp. 51-55, 1988.
Sajwan, K.S. and W.L. Lindsay. Effects of redox on zinc deficiency in paddy rice. Soil Sci. Soc.
Am. J. 50, pp. 1264-1269, 1986.
Schwab, A.P. and W.L. Lindsay. Effect of redox on the solubility and availability of iron. Soil Sci.
Soc. Am. J. 47, pp. 201-205, 1983.
Smith, C.J., M.F. Wright, and W.H. Patrick, Jr. The effect of soil redox potential and pH on the
reduction, and production of nitrous oxide. J. Environ. Qual. 12, pp. 186-188, 1983.
Sposito, G. The Thermodynamic of Soil Solutwnd. Oxford, 1981.
Sposito, G. The Chemidtry of Soiu. Oxford, 1989.
Tack, F.M., O.W.J.J. Callewaert, and M.G. Verloo. Metal solubility as a function of pH in a
contaminated, dredged sediment affected by oxidation. Environ. PolLut. 91, pp. 199-208, 1996.
Tessier, A. and P.G.C. Campbell. Comments on "Pitfalls of Sequential Extractions" by P.M.V.
Nirel and F.M.M. Morel, Water &1.24, pp. 1055-1056, Water Red. 25, pp. 115-117, 1991.
Trolard, F., J.M.R. Genin, M. Abdelmoula, G. Bourrie, B. Humbert, and A. Herbillon. Identi-
fication of a green rust mineral in a reductomorphic soil by Mosssbauer and Raman spec-
troscopies. Geochim. COdmochim. Acta 61, pp. 1107-1111, 1996.
Influence of Reducing Conditions on Mobility of Divalent Trace Metals in Soils 175
van Breemen, N., J. Mulder, and C. T. Driscoll. Acidification and alkalinisation of soils. Plant SoiL
75, pp. 283-308, 1983.
Vance E.D., P.C. Brookes, and D.S. Jenkinson. An extraction method for measuring soil
microbial biomass C. SoiL BioL. Biochem. 19, pp. 703-707, 1987.
Welch, J.E. and L.J. Lund. Soil properties, irrigation water quality, and soil moisture level
influences on the movement of nickel in sewage sludge-treated soils. J. Environ. QuaL. 16, pp.
403-410, 1987.
Xiang & Banin. Long term transformations of toxic heary metals in arid-zone soils incubated. 1.
Under saturated conditions. Water Air SoiL PoLLut. 96, p. 367, 1997.
Xiong, L.M. and R.K. Lu. Effect of liming on plant accumulation of cadmium under upland or
flooded conditions. Environ. PolLut. 79, pp. 199-203, 1993.
CHAPTER 9
Lead Mobilization in
Calcareous Agricultural Soils
Carmen Perez-Sirvent, Josefa Martinez-Sanchez, and Carolina Garcia-Rizo
INTRODUCTION
Soil provides the physical support and the nutrients necessary for plant growth al-
though it can also provide undesirable concentrations of certain elements which, as well
as causing phytotoxicity problems, may pass into the food chain and affect human health
(Kabata-Pendias and Pendias, 1992; Thunus and Lejeune, 1994). There have been nu-
merous studies published concerning the mobility of heavy metals and of their adverse
environmental impact (Allen et al., 1995; Salomons and Forstner, 1995; Ure and Davidson,
1995). However, most studies refer to noncalcareous soils and when similar techniques
to those described in the bibliography are used for highly calcareous soils, the results
may not accurately indicate metal availability because of the peculiar physicochemical
and mineralogical characteristics of such soils. Indeed, the results may be ambiguous
and even contradictory, not fully describing the behavior of the metal in the soil, so that
their environmental impact will not be fully understood. While the present situation in
these soils might be reflected, neither potential future adverse effects can be predicted
nor possible measures for reducing such hazards be proposed.
There are large areas of calcareous soils in SE Spain and many are found in places
where Pb and Zn are mined. These soils often present very high levels of these metals
with a strong spatial dispersion of values. At the same time, these areas often have strong
agricultural traditions and may be dedicated to the growth of horticultural crops, since
the local climate is very favorable to such activities.
The area around Mazarron (Murcia, SE Spain) presents ideal characteristics for a
study such as that undertaken in this chapter. It has geomorphological, edaphological
and climatic characteristics similar to other mining areas in the Mediterranean area with
a scant, degraded vegetation. There are many old mining installations; the soil is calcar-
eous and often has a very high Pb content, and much is known about the original mate-
rial of these soils. Agriculture is intensive, both in the open air and in greenhouse, and is
177
178 Fate and Transport of Heavy Metals in the Vadose Zone
mainly dedicated to horticultural crops. In this chapter we describe the origin, geochemical
process, natural mobility of Pb, and the mobilization provoked by reagents in different
acid and Eh conditions.
SOIL FORMATION FACTORS
Environmental Conditions
A good example of calcareous agricultural soils (Figure 9.1), is found in Las Moreras
Rambla (wadi complex), which may occasionally be flooded after heavy rain. The soils
of the area feeding the rambla are calcic and petrocalcic xerosols, and calcareous rego-
sols (Alias et al., 1986). Nearby there is a drainage channel which carries substantial
quantities of materials (urbic antrosols) from the mining area of Cabezo de San Cristobal
and Las Pedreras. Piping type erosion is present in slags and reveals the existence of
surface particulate flows and infiltrations of subterranean waters. The flood area is wide
(500-1000 m) and three levels of terraces have been identified (Figure 9.2). The fre-
quent braids of the present bed indicate continuous modification of the channel. It is
here where there are high drainage flow densities and where the most intensive agricul-
tural activity takes place. The soils are sandy-calcaric fluvisols (coarse texture) and salt-
calcaric (high salt content) (Alias et al., 1986).
Erosion and sedimentation is substantial in the last stretches of the wadi. Evidence of
this is the water erosion observed over a short period of time (1956-78) by Rodriguez-
Estrella (1993) in Sierra de las Moreras, where there was a very clear modification of
the detritic fans at the pediments and in the terraces.
The spatial distribution of Pb minerals is related to the mining history of SE Spain
which stretches over more than 2500 years, starting with the exploitation of these depos-
its by the Phoenicians and Carthaginians in the fifth and fourth centuries B.C. The Ro-
mans opened up shafts of up to 500 m depth. After a relatively quiet period, mining
started again in the twentieth century but ceased in 1962.
Annual rainfall is 185-312 mm and the average temperature is 16.5 to 18.8C. Mean
potential evapotranspiration is 869-935 mm and there is a very pronounced hydric defi-
cit. Rainfall is scarce and may be torrential, resulting in flash floods with occasionally
disastrous consequences. For example, on 7 September 1989 rainfall in the wadi area
exceeded 100 mmlh.
In the climatic conditions of the area, the high temperatures cause the water con-
tained near the surface of the soil to evaporate rapidly. Weathering occurs extremely
slowly in the dry air. The fundamental alteration and mobilization processes occur at
greater depths, in the unsaturated zone or aeration zone, where the water is in the form
of vadose water. The water remains here for a longer time and there is greater root
development.
Nature of the Materials
The materials which participate in the formation of these soils are sediments from
soils developed on carbonated rocks (limestones, dolomites, sandy marls and conglom-
erates), volcanic rocks and slags which are found in the areas around the gully (Figure
9.1) and which have undergone a process of erosion, transport, and sedimentation.
lead Mobilization in Calcareous Agricultural Soils 179
t
N
Moreras
Mediterranean Sea
Legend
1 0 2 3 4 SKm
I I So.;, Pb
2
+, W
Figure 9.1. Studied zone.
Figure 9.2. Schematic diagram of the wadi complex.
180 Fate and Transport of Heavy Metals in the Vadose Zone
The scarce organic material accompanying these soils (particularly in horizon A) un-
dergoes subsequent oxidation and mineralization, both of which are very important bio-
logical degradation processes in the soils of semiarid areas (Tudela and Martinez-Sanchez,
1997). In this way, N0
3
-, PO/-, K+, Na+, Mg2+, Ca
2
+, and Fe
2
+, are made more available
to the plants.
The materials which contribute Pb to the study area come from mine wastes, which
are a mixture of materials including unaltered parent material (dacites and riodacites),
altered parent materials, primary mineralization products (metallic sulfIdes), secondary
mineralization products (hydrothermal alteration and products of the supergenic alter-
ation of sulfIdes), and remains of the mechanical and metallurgical treatment of exploited
ores. These materials therefore are of a very varied nature with a heterogeneous
granulometry. Tables 9.1 and 9.2 show the chemical and mineralogical composition of
the fresh and altered rocks and the predominant mineralogy of the mining waste studied
in the zone (Arana et al., 1993).
The eroded xerosols contribute calcite and dolomite as principal components, accom-
panied by quartz and feldspars, small quantities of clay minerals such as illite and kaolin-
ite together with chlorite and smectite, iron oxides and organic materials. The calcareous
regosols contribute abundant quantities of clay minerals (illite and smectite), calcite,
dolomite, small quantities of quartz and feldspar and organic materials. Figure 9.3 shows
the sequence of the formation processes which take place.
The calcareous soils of the reference area (Table 9.3) had a low organic matter con-
tent and a slightly basic pH, reflecting their high calcium carbonate content. The texture
was sandy loam to clay loam and the cationic exchange capacity values were low, as was
to be expected in soils with a low proportion of organic matter and containing illite type
clay. In some cases, the concentration of soluble salts was very high for cultivated soils.
A mineralogical study of the fIne earth fraction showed that all the samples were of a
similar composition: calcite (30-40%), quartz (20-35%), illite (10-25%), feldspars (10-
20%), and dolomite (0-15%). Representing less than 5% in all samples and therefore not
easily observed by XRD were the oxides and oxyhydroxides of Fe and Mn, which are of
low crystallinity or amorphous. Some samples contained gypsum. Figure 9.4a shows the
X-ray diffractograms of a representative sample. It must also be noted that total Pb
levels were high (3000 ppm) for an agricultural soil (Table 9.6).
Examination of the samples using a scanning electron microscope equipped with EDS
analyzer revealed quartz, calcite, clays, plagioclase, and feldspars of different particle
sizes. Mapping of the chemical element distribution (Figure 9.5) showed that the lead
was distributed widely but with no set morphology.
TRANSPORT
The above materials are carried by flowing water to the study area in the following way:
as dissolved load from sediment-forming minerals in solution
as suspended load from solid sedimentary materials carried along in suspension
as bed load dragged along the bottom of the stream channel.
The degree and rate of movement of suspended sedimentary material in streams de-
pend on the velocity of the water flow and the settling velocity of the particles in suspen-
Table 9.1. Chemical Composition of Volcanic Rocks
Si0
2
AI
2
0
3
Fe
2
0
3
CaO
% % % %
Unaltered Volcanic Rocks
Max 61.46 17.58
Min 52.86 12.91
Mean 56.1 3 1 5.87
Altered Volcanic Rocks
Max 67.81
Min 21.78
Mean 47.26
17.10
7.30
12.80
5.98
3.08
4.70
31.80
4.35
10.62
4.39
2.64
3.37
3.16
0.30
0.84
MgO Ti0
2
K
2
0
% % %
2.67 0.96 5.32
1.23 0.44 3.32
1.63 0.78 3.50
2.05 1.47 7.80
0.20 0.29 2.45
0.60 0.98 6.80
Na
2
0 PbO DlOOOC
% ppm %
' r
'/I)
III
0.
5.57 391 9.43
1 3:
!O
2.82 70 4.38
10-
1=
3.50 180 6.20
N
III
...
c)"
::l
1.44 5365 28.11
5'
0.58 285 5.26
(")
0.89 2358 12.34
III
i=i
QJ
....
/I)
0
c
'"
~
....
;:;.
t:
;:;
t:
....
a!.
VI
9.
Vi
....
0:1
....
182 Fate and Transport of Heavy Metals in the Vadose Zone
Table 9.2. Summary of the Mineralogical Composition in Mine Waste
Mineral Volcanic Rock Altered Rock Material
Silicates
Sulfides
Oxides and
oxyhydroxides
Carbonates
Insoluble sulfates
Soluble sulfates
Elements
Quartz
Cristobalite
Plagioclase
Sanidine
Phyllosilicates
Amphiboles
Calcite
Quartz
Sanidine
Phyllosilicates
Hematites
Goethite
Calcite
Jarosite Group
Alunite Group
Gypsum
Sulfur
Quartz
Cristobalite
Plagioclase
Sanidine
Phyllosilicates
Sphalerite
Galene
Pyrite
Marcasite
Cerussite
Smithsonite
Azurite
Malachite
Jarosite Group
Alunite Group
Gypsum Anglesite
Copiapite Group
Halotrichite Group
Botryogen Group
Romerite
Alunogen
Coquimbite
ButIerite
Bianchite
Melanterite
Goslarite
Epsomite
Ferrohexahydrite Group
Szomolnokite
Kieserite
Goslarite
Sulfur
sion. Under normal conditions, finely divided silt, clay, or sand make up most of the
suspended loam, although larger particles are transported when the water flows rapidly.
The ability of a stream to carry sediment increases with both the overall rate of flow of
the water (mass per unit time) and the velocity of the water. Both of these are higher
under flood conditions, and so floods are particularly important in the transport of sedi-
ment. Thus the sporadic torrential rains of the area transport large quantities of carbon-
ated materials and materials from the slags. Excessive rates of water flow prevent
infiltration, lead to flash floods, and cause soil erosion.
Dissolved load
Slowly flowing water results in dams being formed on the slopes of the mining areas,
leading to new alteration phenomena. This water flows in the vadose zone for several
lead Mobilization in Calcareous Agricultural Soils 183
1 ... Li.m.es.t.on.e_ ... 1 c:::J
l
H20
Particulate Material
Clay, Calcite, Dolomite ...
C02
t
Dissolution
Cr, HCOi, SO:

Na: K+
Weathering Products:
Oxldates
Hydrates
Hydrolyzates
Sulfates
I Dissolution:
H; pn
Precipitation
Neutralization
Hydrolysis
Hydration
Adsorption
Oxidation
Sedimentation
Figure 9.3. Main processes involved in the soil evolution.
Table 9.3. Analytical Data of Soil Samples
meql
%CaC0
3
100g ds/m
%O.M C/N Equivalent T E.C. H
2
O
Min 1.0 3.3 27.9 7.6 2.7 7.7
Max 1.9 6.7 66.7 16.0 9.4 8.0
Mean 1.3 4.4 39.2 11.0 3.4 7.9
Slags
pH
KCI
7.6
7.8
7.7
J
Volcanic rock
Sulfides
Sulfates
Oxides ..
Partie. Material
Clay, Jarosite, ...
Size
Particle
Distribution
%Sand %SiIt %Clay
23.1 11.0 12.1
55.0 49.4 21.5
42.5 24.5 16.4
days after rain has fallen. Table 9.4 shows the mean composition of the soluble products
contained in the waters that leach from the slags.
The behavior of the elements was studied by correlation techniques. There was a
positive correlation between Pb and Fe(II), and As and Sb, but a negative correlation
between Pb and pH, reflecting the decrease in Pb concentration at lower acidity values
(Lopez-Aguayo et al., 1992).
184 Fate and Transport of Heavy Metals in the Vadose Zone
a)
b)
c)
d)
28
o 10 20 30 40
Figure 9.4. X-ray diffractograms obtained from a soil sample (a) and from the remainders after the
first, second, and third extraction steps (b, c, and d, respectively) recommended by BCR.
SE Fe K Mg Ca
Na o Pb S Si
Figure 9.5. Mapping of the chemical element distribution obtained by SEM-EDS (raw sample).
When gramineae covered the soils developed from limestones and dolomite, which
occurred mainly in spring, the carbon dioxide concentration increased in the soil solu-
tion, leading to the solubilization of calcium carbonate:
Table 9.4. Chemical Composition of Acidic Waters
Major Compounds mmol/L
Si Fe(lII) Fe(1I) AI Mg Zn
Max 5 947 258 1001 477 627
Min 0.9 67 6 44 16 26
Mean 3 349 45 256 200 189
Minor Compounds mmol/L
Pb As Sb Ca Ge Sn
Max 0.030 7.34 0.213 1.996 0.964 0.826
Min 0.005 0.44 0.008 0.200 0.027 0.042
Mean 0.001 2.60 0.120 0.500 0.324 0.337
Mean values for 30 samples collected during a year.
Mn Cu
55 6
4 1
27 3
Cr Cd
0.025 1.156
0.002 0.027
0.010 0.780
S04
1200
826
915
Na
0.2610
0.026
0.100
pH
2.2
1.5
1.8
K
0.084
0.002
0.061
r
<'1l
III
0.
s:
o
g;
N'
III
....
o
:::l
:::l
(')
III
i=i
III
.:;
o
= VI
~
C1Q
'""
;:;.
=
;:;:
=
'""
!l:!.
III
9.
v;
-'
00
U1
186 Fate and Transport of Heavy Metals in the Vadose Zone
It is well known that water with a high dissolved carbon dioxide content in contact
with calcium carbonate contains Ca
2
+ and HC0
3
- ions. Flowing water containing cal-
cium may become more basic through the loss of CO
2
to the atmosphere or through
contact with dissolved bases, resulting in the deposition of insoluble CaC0
3
:
This is very important throughout SE Spain (Martinez-Sanchez, 1982). In sloping
limestone areas there was a substantial mobilization of carbonates due to the transport of
water-soluble bicarbonates 15-30 cm below the surface, which precipitate in the lower
zones to form calcic, sometimes strongly cemented, horizons. An example of this are the
calcic and petrocalcic xerosols at the foot of the Sierra de las Moreras (Figure 9.1).
Soluble salts are dissolved rapidly in marls. Given the semiarid climate, scant vegeta-
tion, and low permeability of the marls, the dissolution of alkaline-earth carbonates takes
place very slowly and may even be prevented. There was no sign of carbonate movement
through the profIle of developed soils (calcaric regosols).
Particulate Forms: Suspended and Bed Loads
When the water flow is rapid, soluble and particulated forms are dragged unchanged
from their place of origin as far as the gully. As mentioned above, the mineralized mate-
rials contribute and generate insoluble sulfates (anglesite, jarosite, alunite), oxides and
oxyhydroxides, and clay minerals such as kaolinite and chlorite, which are transported
along with the soluble materials (Table 9.2). In this way, unaltered materials are incor-
porated into this phase as fragments of unaltered rocks, resistant minerals of encasing
rocks and sulfIdes.
The Pb is bound to:
sulfates and sulfIdes (as a main component)
oxides and hydroxides of iron and manganese (adsorbed and guest in the frame)
silicates (included in the crystalline structure and adsorbed).
The most noteworthy contribution of carbonated particulate materials to the soil is
calcite, with dolomite and phyllosilicates being the most common.
GEOCHEMICAL PROCESSES
When the kinetic energy of the transported system diminishes, materials are
sedimented. Both suspended particles and solutes which are being carried along by the
water are sedimented regardless of their origin, with coagulation and flocculation phe-
nomena and heterogeneous chemical reactions occurring.
The important chemical actors in the medium studied are:
acid waters
waters containing bicarbonates
very saline waters
carbonated particles
lead Mobilization in Calcareous Agricultural Soils 187
silicated particles
mineralized particles (Pb)
Chemical weathering, as a chemical phenomenon, can be regarded as the result of the
tendency of a rock/water/mineral system to attain equilibrium. This occurs through the
usual chemical mechanisms of dissolution/precipitation, acid-base reactions, complex-
ation, hydrolysis, and oxidation-reduction, the kinetics of which is conditioned by pH,
Eh, and temperature.
Water increases the rate of weathering for several reasons. Water, itself, is a chemi-
cally active substance in the weathering process. Rainwater is free of mineral solutes but
is slightly acidic due to the dissolved carbon dioxide it contains. This makes it chemically
aggressive, particularly when compounded by the presence of the sulfuric acid which
results from the supergenic alteration of metallic sulfides.
Mobilization-Physical Weathering-Hydration Relations
As regards the way in which physical weathering affects materials, it is known that
rocks tend to weather more rapidly when there are pronounced differences in the physi-
cal conditions, such as alternate freezing and thawing, alternating wet and dry periods,
and roots growing through cracks.
The most important mechanical aspect is the swelling and shrinking of minerals which
accompany hydration and dehydration. Such processes are clearly in evidence in the
studied zone since the marls contain alkaline and alkaline-earth sulfates and chlorides,
which give rise to polyhydrated salts (gypsum, hexahydrite, epsomite, thenardite, etc.):
Similarly, a large number of sulfates of different degrees of hydration are found in the
products which originate in the mineralized zone and which are formed in the final stages
of the supergenic alteration process. In some cases, these do not remain unaltered but
change with the environmental conditions of humidity and temperature (Lopez-Aguayo
and Arana, 1992). All the above favors contact between the solid particles and the alter-
ation agents, and results in faster chemical reactions:
Fibroferrite [FeS04(OH) 5H
2
0] :=:} Hohmannite + 1.5 H
2
0
:=:} Amarantite + 0.5 H
2
0
Soluble Pb-Adsorbent Precursor Ratio
The mineralized materials dissolve to produce acid waters, which contain SO/, H+,
Fe+
3
, Fe+
2
, Al+
3
, and Pb+
2
, whose evaporation products in a confined atmosphere are
shown in Table 9.5.
188 Fate and Transport of Heavy Metals in the Vadose Zone
Table 9.5. Mineralogy of the Precipitates in Acid Waters
Copiapite
Halotrichite
Alunogen
Paracoquimblte
Bianchite
Formula
(Zn,Fe,Mg)(AI,Fe)i50,J6(OHh' 20 H
2
0
FeAI
2
(50
4
)4' 22 H
2
0
AI
2
(50
4
h' 17 H
2
0
Fe
2
(50
4
h' 17 H
2
0
(Zn,Fe, Mg) 50
4
H
2
0
Max
50
35
15
14
40
Mean
14
18
8
10
15
The mineralogy of these compounds is closely related with the overall chemical com-
position of the samples, as Zodrow (1980) found in Nova Scotia and Yushkin (1984) in
Pai-Khoi. The following groups can be differentiated (a) sulfates of AI (alunogene) or
Fe(III) (coquimbite and paracoquimbite), (b) sulfates corresponding to the isomorphous
series of Zn, Fe(II) and Mg (bianchite), and (c) hydrated double sulfates of trivalent
metals (Fe, AI) and divalent metals (Mg, Fe, Zn), with a molar ratio, M(lII)/M(II), of 4
in the copiapite group and of 2 in the halotrichite group.
These compounds also occur in the alteration products of the slags (Table 9.2). The
same phases were found in in vitro experiments (Perez-Sirvent et al., 1993).
S04 = + R3+ + R2+ + H
2
0 + H+ =}R
2
(S04h n H
2
0 + RS0
4
. n H
2
0
R(S04Hh n H
2
0 + R (S04) . n H
2
0
R+
2
R+
3
2
(S04)4' n H
2
0
R+
2
R+
3
4
(S04)6 (OHh. n H
2
0
where R+3 = AI+
3
, Fe+
3
and R+2 = Fe+
2
, Mg+
2
, Zn+
2
, Mn, Cu are the principal elements,
incorporate smaller quantities of Pb, Ge, Sn, Cd, Co, Ni, Cr.
A simplified model reflecting this process should at least take into account the follow-
ing two factors: the solubility of the crystallized phases and the stoichiometric ratios of
the sulfates. The first sulfates to crystallize are those of the divalent elements since their
solubility product is lower, resulting in the formation of Fe (S04)'7 H
2
0 (melanterite),
which is only stable at very acidic pH. The importance of this mineral species is that the
Fe(II) sulfates are precursors of the genesis of Fe (OHh, which is adsorbent and which
transports Pb.
Bicarbonated-Acidic Water Interaction
The water which runs off and leaches from calcareous materials has a high HC0
3
-
content. When this water reacts with the above-mentioned acidic waters there is a
coprecipitation of basic iron carbonates which retain different proportions of Pb. De-
pending on the pH, the precipitation of Fe(OHh, which acts as adsorbent of Pb, can be
gIven as:
lead Mobilization in Calcareolls Agricultural Soils 189
'-- The soluble Pb can react with HC0
3
-:
a reaction which is favored as the acidity of the medium is reduced by bicarbonate water.
If, in addition to HC0
3
-, ligands such as CI-, S04 =, C0
3
=, OH-, humic and fulvic acids,
and CN- (used in sulfide flotation processes) are also present, Pb complexation reac-
tions may occur.
Acid Water-Mineralized Particulate Material-0
2
-C0
2
Interaction
The instability of sulfides in an oxidizing and acidic medium results in their oxidation.
The most important oxidation reactions for the mobilization of lead are those involving
pyrite and galena:
The carbon dioxide acts in the oxidation process by carbonating the metal and supplying
sulfate ions, which catalyze subsequent oxidation reactions, and by increasing the
medium's acidity:
Of special interest is the redox chemistry of the Fe and Mn present in these outcrops
in relation to the mobility of other metals. Their hydroxides and oxyhydroxides play an
important role in Pb adsorption phenomena (Manahan, 1994). The Fe(OH)3 is a link in
the process of jarosite formation:
This H-jarosite includes Na, K, and Al in the general jarosite formula (Na,K,
Xh(Fe,Alh (S04)2 (OH)6 where X is Ca, Pb, Ag ... The Pb enters the sulfate frame
either because it is retained in the Fe (OHh by adsorption or because it is brought into
the reaction as soluble Pb (plumbojarosite).
As regards the genesis of alunite, the most probable formation mechanism is the trans-
formation of potassium micas via kaolinite (Hueso et al., 1981).
Biogeochemical reactions involving thwhaciLLIM !errooxwafM and thwhaciLLIM thwoxwafM
type bacteria may take place in these processes (Lundgren and Silver, 1980; Allan, 1995).
Acid Water-Carbonated Particulate Material-0
2
-CO
z
Interaction
The calcite present acts as pH buffer in the solutions, counteracts the acidification of
the medium brought about by the H+ ions and, by reacting with the sulfate ions present,
enables cerussite to exist in these media.
190 Fate and Transport of Heavy Metals in the Vadose Zone
The acidic waters react with the carbonates to bring about precipitation of the Fe and
AI hydroxides, which can react with sulfate ions to form jarosite and gypsum. For this
reason, the materials that make up the bed of the wadi in this zone have a pH in water of
3, an EC of 24 mS/cm, and are made up of quartz, feldspar, illite, gypsum, and jarosite
(Perez-Sirvent et al., 1995).
In areas of the gully with a pH<3 and slightly oxidizing conditions, the Pb undergoes
a process of Fe and Mn oxyhydroxide desorption, as mentioned above. If, on the other
hand, the pH is slightly alkaline, the Fe and Mn can be precipitated as carbonates, which
are less adsorbent than the corresponding oxyhydroxides.
In unusual situations of anaerobiosis caused by flooding or the rising of the water
table and in acid conditions, Pb will be mobilized more readily because the conditions
will be ideal for dissolving the Pb compound mentioned above.
Pb Sorption-Desorption
In the above interactions, Pb sorption/desorption reactions occur, with those involv-
ing the oxideslhydroxides of AI, Fe, and Si being especially important. The sorption of
Pb on solid surfaces typically shows an "adsorption edge" onto iron hydrous oxide (Liang
and McCarthy, 1995).
The mechanism regulating these processes can be explained by complexation. The
surface of oxides and hydroxides such as Mn0
2
can adsorb Pb+
2
, giving rise to surface
complexation,
or chelation
M-OH
I + Pb
2
+ ~
M-OH
The pH plays an important role in the sorption/desorption reactions. At high pH
(pH> 6), Pb is preferentially sorbed onto iron oxide, but at pH < 3.0, almost all the Pb
remains in solution. Between pH 3 and 6, sharp increases in Pb adsorption accompany
very small increases in pH. This is why, in our case, the process of Pb sorption is impor-
tant, since the increased pH caused by the mixing of acidic waters with carbonated
suspensions leads to the desorption/adsorption oflead (Lumsdon and Evans, 1987).
The phyllosilicates may have similar action mechanisms to those reported above, and
act as possible adsorbents of: (1) the soluble Pb in the leaching waters, (2) the Pb liber-
ated in situ by the oxidation of particulate Pb sulfide, and (3) the Pb liberated in the
desorption of the particulated Fe and Mn oxidelhydroxides produced by a decrease in
pH following contact with acid waters. No effect of the ionic strength on Pb adsorption
on a goethite surface has been observed (Hayes and Leckie, 1995). From this, we can
deduce that the high salt concentration of the waters from the zone containing marls has
no effect on the Pb adsorbed on the oxideslhydroxides.
lead Mobilization in Calcareous Agricultural Soils 191
Although humic substances act as adsorbents of Pb in soil, this was of little impor-
tance in our case, because of the very small quantities concerned.
MOBILITY
Provoked Pb MObility: Speciation Study
The method traditionally used for heavy metal mobilization studies is based on spe-
ciation studies (sequential and selective extraction procedures) (Tessier et al, 1979;
Forstner, 1985).
Chemical extraction sequences were applied to differentiate between the exchange-
able, carbonated, reducible (hydrous Fe/Mn oxides), oxidizable (sulfides and organic
phases), and residual fractions. The undisputed advantage of this approach to estimate
the long-term effects on metal mobilities lies in the fact that the rearrangement of spe-
cific solid "phases" can be evaluated prior to the actual remobilization of certain propor-
tions of an element into the dissolved phase (Forstner, 1985).
Figure 9.6 shows the fractions (extracted Pb"'lOO/total Pb) extracted in the three se-
quential extraction steps applied to eight representative samples.
The first step corresponds to the acid soluble fraction which represents the metal
bound in the exchange positions, water soluble metal and carbonate-bound metal; the
second step corresponds to the reducible fraction (metal bound to iron and manganese
oxides), and the third step to the oxidizable fraction of organic matter and sulfide-bound
metal, although other unidentified phases may also be extracted (Tessier et al., 1979;
Forstner, 1985; Kersten and Forstner, 1986; Ure et al., 1993). There may also be other
reactions involving desorption, coprecipitation, or neutralization during the three ex-
traction steps (Khebonian and Bauer, 1987; Martinez-Sanchez et al., 1996).
Experiments with calcareous soils applying sequential extractions as recommended
by the BCR procedure (Ure et al., 1993) show that the fractions extracted from each
sample during the different steps vary greatly and there seems to be a relation between
the total carbonate content and the mineralogical species in question (calcite, dolomite).
In the first step, acetic acid attack neutralizes the calcite when it represents up to 20% of
the total mineral content. If the calcite and dolomite contents are above this level they
remain in the residue, as can be seen from the diffractogram of Figure 9Ab.
All the above is in agreement with electron microscopy observations. Figure 9.7 shows
the distribution mappings of lead, which may be partly superimposed on other elements
such as S, 0, and Fe, possibly as jarosite crystal, or be widely dispersed.
In the second step, nitric acid neutralizes the residual calcite fraction with a large
particle size and part of the dolomite (diffractogram Figure 9Ac). SEM reveals that the
lead may be associated with dolomite crystals, sulfur compounds, and silicates.
Acetic acid is also involved in the third step where, judging from the mineral compo-
sition of the residue (diffractogram, Figure 9Ad), it neutralizes the rest of the calcite and
dolomite present. Distribution mapping of the lead shows it to be associated with sulfate
and silicate crystals, which belong to the least mobile residual lead phase.
The quartz/calcite and quartz/dolomite ratios, as deduced from the XRD study of the
untreated samples and the residues of the sequential extractions (Figure 9.8), lend weight
to the above affirmation. The first ratio increases slightly during the first steps to reach a
maximum during the third step, while the quartz/dolomite ratio remains constant at
192 Fate and Transport of Heavy Metals in the Vadose Zone
100
80
60
40
20
% Pb extr/Pb total
!m3rd EXTR
1IlIID2nd EXTR
1st EXTR
O L - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~
1 2 3 4 5 6 7 8
sample
Figure 9.6. Pb recovered from eight samples using BCR methodology.
SE Fe K Mg
Na o Ph S
Ca
Si
Figure 9.7. Mapping of the chemical element distribution obtained by SEM-EDS (after the first step
extraction, acetic acid).
first, rises during the second step, and shows the same behavior as the quartz/calcite
during the third step.
The BCR sequential extraction method applied to highly calcareous soils does not
relate lead to the different speciation forms described in the bibliography, since carbon-
ates are present in the different extraction steps.
To summarize, according to the scheme described in Table 9.7, the first step corre-
sponds to the most soluble lead, the exchangeable form, and that bound to part of the
carbonates in the form of calcite. However, it seems that significant and substantial er-
rors exist as regards lead speciation when using BCR methodology in soils with a high
carbonate content. The carbonate-bound lead is not totally extracted during the first
Lead Mobilization in Calcareolls Agricultural Soils 193
ratio ratio
sample step 1 step 2 step 3 sample step 1 step 2 step 3
a b
Figure 9.8. Evolution of the quartz/calcite (a) and quartz/dolomite ratios (b) in soil samples after
each one of the three extraction steps.
step, and remains unquantified in the following extraction stages. Consequently, the
lead bound to reducible forms (which is given theoretically by the result found for the
second extraction stage) also includes a part of the metal still bound to calcite and dolo-
mite. Lastly, the lead bound to organic matter and sulfides (third step) is also higher
than the true value, because it also includes the metal remaining in the carbonates which
have not been neutralized in the previous steps. It is clear, then, that the results for lead
mobilization given by BCR in such soils are erroneous since, in fact, lead is extracted
from carbonate during all three steps.
Lastly, the application of these methods to calcareous soils does not predict the mobi-
lization provoked by different environmental conditions, such as changes in pH, redox
potential, etc. The reagents used in all the steps were acids, which react with the carbon-
ates present in these soils and so mask the effect which is to be evaluated.
Mobility in the Vadose Zone
Due to the above-mentioned difficulties in obtaining a reliable speciation scheme by
means of sequential extractions, different reagents which mimic environmental condi-
tions were used to study mobility in calcareous soils. Table 9.6 depicts the main values
obtained in selective extractions.
The mobile fraction is extremely slight (0.01 to 0.1%) in aqueous medium, which
agrees with the described mineralogy and pH of the soils studied. The exchangeable
fraction as determined by using ammonium acetate (pH 7) (Simard, 1993) is low but
higher than the soluble phase. It varies with the mineralogy and texture of the clays; that
is, with the exchange capacity of these soils, which is a function of these two factors,
since the organic matter content is less and the amorphous components scarce.
The amount of lead bound to the amorphous forms of iron oxides and extracted with
oxalic-oxalate at pH 3 (Chapman, 1965) is very small in spite of the fact that both soluble
and exchangeable forms, and part of the lead arising from the mobilization of carbonates
are included. These results agree with those obtained in the exchange phase. Given the
194 Fate and Transport of Heavy Metals in the Vadose Zone
Table 9.6. Values Obtained in the Selective Extractions (Soluble x 100ITotal)
Total HAc
mg/kg H
2
O NH
4
Ac 0.5 N EOTA OTPA Bicar Oxal Cit-Oit Pyro
Min 254 <0.01 1,3 4,4 12,7 3,1 0.8 1,4 5,3 4,7
Max 6513 0.1 3,2 37,6 70,1 8,4 3.1 8,2 64,7 37,3
Mean 375 0.02 2,1 19,9 30,2 4,3 1.8 3,5 27,4 14,7
Table 9.7. sequential Extraction Schemes
Kersten and FOrstner, EC/BCR; Ure et al., EC/BCR; Martinez et al.,
1986 1993 1996
Exchangeable 1 M NH4 OAc 0.1 M HOAc Soluble +
ions Exchangeable +
Carbonate-Associated
Carbonate- 1 M NaOAc,
associated pH 5 w/ HOAc
metals
Easily reducible 0.01 M NH2 OH HCI 0.1 M NH2 OH HCI Carbonate - Associated
(Mn-oxides) w/ 0.Q1 M HN0
3
HN0
3
pH 2 + Reducible
Moderately 0.1 M oxalate buffer
reducible pH 3
(amorph. Fe-oxides)
Sulfidic/organic 30% H
2
0
2
pH 2/ 8.8 M H
2
0
2
(2x); Carbonate - Associated
0.02 M HN0
3
NH
4
OAc/HOAc + Sulfidic /Organic
Residual Hot HN0
3
cone.
adsorption properties of the amorphous substances and their great contribution to the
soil's exchange capacity, it might be expected that high exchangeable lead values would
result from high levels of amorphous substances, although, in fact, the opposite occurs.
This extractant also acts on active calcium carbonate since, due to its very small par-
ticle size, the compound is partially dissolved, leading to a partial precipitation of the Ca
and Pb as oxalate and to the mobilization of the metals in the crystalline structure of this
carbonate. This is revealed in the XRD diagrams of the treated samples, in which reflec-
tions, which can be assigned to phases of calcium oxalate in different states of hydration
appear. The quartz/calcite and quartz/dolomite ratios also increase in the treated samples.
Citrate-dithionite is a reducing reagent with a pH near to neutrality, which extracts
easily extractable amorphous forms of low crystallinity (Mehra and Jackson, 1960),
such as Fe and Mn oxides and oxyhydroxides and the metals associated with them. The
complexes that citrate forms with the metals studied are very stable. Since the metals are
released from amorphous forms or those oflow crystallinity by reduction, they are easily
extracted, and are, in some cases, the most abundant fraction. Several carbonates are
also solubilized in this extraction, which is seen when calculating the quartz/calcite ratio
although there are no precipitation reactions as in the case with oxalates, and the dolo-
mite is hardly affected. The water soluble fractions are included in this fraction (Perez-
Sirvent et al., 1997).
lead Mobilization in Calcareous Agricultural Soils 195
Lastly, sodium pyrophosphate is a reagent which extracts forms bound to humic sub-
stances. As indicated above, it dissolves part of the calcium carbonate and dolomite,
precipitating the calcium phosphate and extracting other materials contained in the car-
bonate structure by complexation. In our case, because of the low humic matter content
of the soils studied, it acts as a carbonate complexer and extractor.
Pb Assimilation by Plants
It is generally accepted that the lead levels in soil extracts obtained when using acetic
acid M 0.5 (MAFF, 1981), EDTA-ammonium acetate-acetic acid (pH 4.65) (Lakamen
and Ervio, 1971), or with DTPA (Linsay and Norvell, 1978) provide a good assessment
of metal assimilation by plants. To check that these particular extraction stages really
indicate the quantity of lead assimilated by plants growing in calcareous soils, both the
vegetal material and the extracts obtained by using these recommended extractants
were analyzed.
Table 9.6 shows the results obtained for the soil extracts. It is clear that the results
considerably differ from one extractant to another, which means that the above assump-
tion is not valid when dealing with the particular type of soils discussed here.
EDT A-ammonium acetate-acetic acid (pH 4.65) acts by complexing metals which
seem to be associated to inorganic forms of Fe and Mn with a low crystallinity and fine
granulometry; that is to say, plant-available metals. In our experiments this reagent also
acted by neutralizing carbonates, as can be seen from Figure 9.9, which represents the
diffractogram of the residue after the reagent has acted on a soil sample. The diminution
in the calcite reflection and background noise, which can be put down to the presence of
amorphous substances, is clear. For this reason, the values for plant-assimilable lead
which are obtained by this procedure (12-70% extractable) are too high to represent the
real in situ absorption by the plant.
Both the soluble metal and the metal associated to the exchange complex and to finely
divided forms are affected by complexation with DTP A. For this reason, the low values
observed in this fraction are the result of the above-mentioned behavior of Pb. The x-
ray diagrams obtained after treatment show decreased background noise, (as do the
EDT A residues), increased illite crystallinity, and a constant calcite content.
Acetic acid is the most problematic reagent when evaluating plant assimilable Pb in
carbonated soils, since it dissolves calcite totally and dolomite only when it is finely di-
vided. This extraction cannot be considered as suitable for establishing plant availability.
The high solubility of lead in its presence is evident when it is bound to calcite, reflecting
the effects of complexation (high solubility of lead acetate) and of neutralization.
As indicated above, in order to ascertain the reagent that best reflects the plant avail-
ability of Pb, the metal content of plants cultivated in the studied soils was analyzed. The
low assimilation of metal by plants in the cultivation conditions used is reflected in the
levels found in the root, leaves, and fruit. These are illustrated in Table 9.8, in which the
mean values of all the plant samples are depicted. The most suitable reagent in such soils
tends to be DTP A, and the least suitable 0.5 M acetic acid and EDT A.
Pb mobility was assayed in a bicarbonate medium of pH 8.5 (Olsen and Sommers,
1982) since this reagent mobilizes the plant-available phosphorus fraction. The lead in
the above assay also can be easily mobilized in a carbon dioxide rich atmosphere and the
J 96 Fate and Transport of Heavy Metals in the Vadose Zone
Qxal
Cltr
Pyro
DTPA
EDTA
HAc
RS
o
Q
10 20 30
28
40
Figure 9.9. X-ray diffractograms obtained from a soil sample and from the remainders after
selective extractions.
Table 9.8. Mean Lead Concentrations in
Tomato Plants (JIg g-l)a Cultivated in the Soils
Studied
Sample
Root
Leaf
Fruit
Min
1.8
1.2
1.8
Max
4.5
2.8
3.2
a Results for dried material (OW).
Mean
3.0
2.6
2.4
bicarbonate medium would be similar to the bicarbonate environment of the soil solution
in the rhizosphere. The experiment would simulate the mobilization of lead fractions
provoked by the lateral leaching waters from the most superficial zones containing bicar-
bonate ions in solution. Even when the extracted fractions are on the order of 1-3%, these
are mobilization conditions near those existing in carbonated soils. Indeed, the subterra-
nean water extracted from nearby wells for irrigation purposes is rich in bicarbonate (40
meq/L) and the lead that we identifY in this experiment could pass to the soil solution.
CONCLUSION
Because of the high amount of carbonated materials which are supplied by the wadi
complex, an immobilization of the lead lixiviated from the slags takes place.
Lead Mobilization in Calcareous Agricultural Soils 197
The lead speciation methods recommended for other soils and based on the use of
conventional sequential extractions are unreliable. A different approach which takes
into account the particular physicochemical characteristics and mineralogy of these highly
calcareous soils is necessary.
REFERENCES
Alias, L.J., R Ortiz, J. Martinez, and P. Linares. Mapa de Suelod, Mazarrtfn, 976. ICONA,
Universidad de Murcia, 1986.
Allan, RJ. Impact of Mining Activities on the Terrestrial and Aquatic Environment with
Emphasis on Mitigation and Remedial Measures, in Heavy Metau, W. Salomons, U. Forstner,
and P. Mader, Eds., Springer-Verlag. Berlin, 1995, p. 412.
Allen, E.H., C.P. Huang, G.W. Bailey, and A.R. Bowers. MetaL Speciation and Contamination of Soil,
CRC Press, Inc., Boca Raton, FL, 1995, p. 358.
Arana, R., C. Perez-Sirvent, and R. Ortiz-Gonzalez. Explotaciones mineras e impacto ambiental
en el sector de Mazarron (Murcia), in Pr06lemdtica Geoam6ientaL y DedarroLio. Part 2. R Ortiz-
Silla, Ed., Actas V Reunion de Geologia Ambientaly Ordenaci6n del Territorio. Murcia. 1993,
pp. 811-835.
Chapman, H.D. Methodd of SoiLAnaLYdi..l, Part 2. C.A. Black, Ed., American Society of Agronomy,
Inc., Madison, WI, 1965, pp. 891-900.
Forstner, U. Chemical Forms and Reactivities of Metal in Sediments, in ChemicaL Methodd for
AJdedding Bio-Availa6le Metau in SLUJged and Soiu, R Leschber, RD. Davis, and P. L'Hermite,
Eds., Elsevier Applied Science, London, 1985, pp. 1-30.
Hayes, K.F. and J.O. Leckie. Modeling ionic strength effects on cation adsorption at hydrous
oxide/solution interfaces. J. CoLloUJ Inter:/. Sci. 115, pp. 564-572, 1987.
Hueso, R, J. Rodriguez-Gordillo, and F. Lopez-Aguayo. Las jarositas de Sierra de Almagrera
(Almeria). Mineralogia y genesis. BoL. Soc. Elp. Mineralogta, 4, pp. 29-36, 1981.
Kabata-Pendias, A. and H. Pendias. Trace Elementd in Soiu and Plantd. 2nd ed., CRC Press, Inc.,
Boca Raton, FL, 1992, p. 365.
Kersten, M. and U. Forstner. Chemical fractionation of heavy metals in anoxic estuarine and
coastal sediments. Water Sci. Techno!. 18, pp. 121-130, 1986.
Khebonian, C. and C.F. Bauer. Accuracy of selective extraction procedures for metal speciation
in model aquatic sediments. AnaL. Chem., 59, pp. 1417-1423, 1987.
Lakanem, E. and R. Ervio. A comparison of eight extractants for the determination of plant
available micronutrients in soils. Acta Agric. Ferm. 123, pp. 223-232, 1971.
Liang, L. and J.F. McCarthy. Colloidal Transport of Metal Contaminants in Groundwater, in
MetaL Speciation and Contamination of Soil, E.H. Allen, C.P. Huang, G.W. Bailey, and A.R
Bowers, Eds., CRC Press, Inc., Boca Raton, FL, 1995, p. 358.
Linsay, W.L. and W.A. Norvell. Development of a DTPA test for zinc, iron, manganese and
copper. SoiL Sci. Am. J., 42, pp. 421-428, 1978.
Lopez-Aguayo, F. and R Arana. Las etapas finales en la alteraci6n supergenica de sulfuros, in
RecurdOd Mineraw de Elpaiia. J. Garcia-Guinea and J. Martinez-Frias, Eds. Textos U niversitarios,
C.S.I.C., Madrid, 1992.
Lopez-Aguayo, F., C. Perez-Sirvent, R. Ortiz Gonzalez, and R Arana. Composici6n quimica de
las aguas de lixiviaci6n minera en el Cabezo de San Cristobal (Mazarron, Murcia). Rev. Soc.
GeoL. Elpana 5(3-4), pp. 73-79, 1992.
Lumsdon, D.G. and L.J. Evans. Predicting Chemical Speciation and Computer Simulation, in
ChemicaL Speciation in the Environment, A.M. Ure and C.M. Davidson, Blackie Academic &
Professional, Chapman & Hall, New York, 1987, p. 408.
! 98 Fate and Transport of Heavy Metals in the Vadose Zone
Lundgren, D.G. and M. Silver. Ore leaching by bacteria. Ann. Rev. MicrobioL., 34, pp. 263-283,
1980.
Manahan, S.E. Environmental Chemutry, 6th ed. CRC Press, Inc., Boca Raton, FL, 1994, p. 811.
Martinez-Sanchez, J. FAudio Eddjico de la.J SierrM oe Orce y Maria, Tesis Doctoral. 1982.
Martinez-Sanchez, J., C. Perez-Sirvent, and C. Garcia Rizo. Errores de Evaluaci6n de riesgos
en la movilizaci6n de metales pesados en suelos carbonatados. Acta.! III Congruo del Meoio
Ambiente, Madrid, 1996, pp. 1053-1101.
Martinez-Sanchez, J., C. Perez-Sirvent, and C. Garcia Rizo. La problematica del Zn y Pb en el
estudio de la posible contaminaci6n por metales pesados en suelos agricolas de zonas aridas,
in Recur,}o,} Natura0 y Medio Ambiente en el Surute Penin.Jular, L. Garcia-Rossell and A. Navarro,
Eds., Instituto de Estudios Almerienses, 1997, pp. 445-454.
Mehra, O.P. and M.L. Jackson. Clay,} and Clay Minerau, Monograph No.5, Earth Science. Series:
1960, pp. 317-327.
Ministry of Agriculture, Fisheries and Food. The Analy,}u of AgricuLturaL Materiau. Her Majesty's
Stationery Office, London, 1981.
Olsen, S.R. and L.E. Sommers. Phosphorous, in Method,} of Soil Analy,}u, Part 2, 2nd ed., A.L.
Page, Ed., American Society of Agronomy, .Madison, WI, 1982.
Ortiz Gonzalez, R. Quimumo de !O,} producto'} de aLteracwn ,}uperginica en el dutrito minero de Mazarron
(Murcia). Tesis Doctoral. Universidad de Murcia. 1991, p. 466.
Perez-Sirvent, C., M.M. Garrido, R. Arana, and F. Lopez-Aguayo. Cristalizaci6n de Sulfatos de
Fe (III), Zn y Mg: grupo de la Copiapita. Bo!. Soc. Elp. MineraL. 16, pp. 143-153, 1993.
Perez-Sirvent, C., J. Martinez-Sanchez, and C. Garcia-Rizo. Assessment of the Risk of Heavy
Metal Mobilization in Calcareous Agricultural Soils, in Proceeding'} of the Third InternationaL
Conference on the Biogeochemutry of Trace Element,}, 110-PDF. Contaminated Soils. INRA, 1997.
Ritchie, G.S.P. and G. Sposito. Speciation in Soils, in ChemicaL Speciation in the Environment, A.M.
Ure and C.M. Davidson, Blackie Academic & Professional, Chapman & Hall, New York,
1995, p. 408.
Rodriguez-Estrella, T. EI caracter torrencial de la Rambla de las Moreras (Murcia) y su incidencia
en la ordenaci6n del territorio, in Problemdtica Geoambiental y Duarrollo. Part 2. R. Ortiz-Silla,
Ed., Actas V Reunion de Geologia Ambiental y Ordenaci6n del Territorio. Murcia. 1993, pp.
835-853.
Salomons, W. and U. Forstner. Metau in the Hydrocycle, Springer, Berlin, 1981.
Salomons, W., U. Forstner, and P. Mader. Heavy Metau. Springer-Verlag, Berlin, 1995, p. 412.
Simard, H. Ammonium Acetate-Extractable Element, in Soil SampLing and MethOd,} of AnaLy,}u,
M.R. Carter, Ed., for Canadian Society of Soil Science, Lewis Publishers, Boca Raton, FL,
1993, p. 823.
Sposito, G. and S. Mattigod. Geochem, University of California, 1980.
Sposito, G. The Chemutry of Soiu, Oxford University Press, New York, 1989, p. 277.
Tessier, A., P.G.C. Campbell, and M. Bisson. Sequential extraction procedure for the speciation
of particulate trace metals. AnaL. Chem. 51, pp. 844-851, 1979.
Thunus, L. and R. Lejeune. In Handbook on Metau in CLinicaL and AnaLyticaL Chemutry, H.G. Seiler,
A. Sigel, and H. Sigel, Eds., Marcel Dekker, Inc., New York, 1994.
Tudela, M.L. and J. Martinez-Sanchez. Desertizaci6n progresiva en el sureste peninsular y su
relaci6n con la puesta en cultivo, in Recur,}o,} Naturale,} y Medio Ambiente en eL Surute Penin.Jular,
L. Garcia-Rossell and A. Navarro, Eds. Instituto de Estudios Almerienses, 1997, pp. 433-444.
Ure, A.M., Ph. Quevauviller, H. Muntau, and B. Griepink. Speciation of heavy metals in soils
and sediments. An account of improvement and harmonization of extraction techniques
undertaken under the auspices of the BCR of the Commission of the European Communities.
Int. J. Environ. Anal. Chem., 51, pp. 135-151, 1993.
lead Mobilization in Calcareous Agricultural Soils 199
Ure, A.M. and C.M. Davidson. Chemica! Speciation in the Environment, Blackie Academic & Profes-
sional, Chapman & HalL New York, 1995, p. 408.
Yushkin, N.P. Mineralogy of Pai-Khoi copiapite, Akad. Nauk. SSSR Komi Fi!, InA. Ceo!., 45, pp.
79--86, 1984.
Zodrow, E. Hydrated sulfates from Sydney Coalfield, Cape-Breton Island, Nova-Scotia, Canada.
The copiapite group, Am. Mineralogidt, 65, pp. 961-967, 1980.
CHAPTER 10
Metal Retention and Mobility as Influenced
by Some Organic Residues Added to Soils:
A Case Study
Luis Madrid
INTRODUCTION
This chapter provides a review of the present knowledge of how retention and mobil-
ity of metals in soils and other environmental systems are influenced by organic sub-
stances, especially those present in residues added to soils for recycling purposes, and is
further illustrated with examples of recent research carried out with an agricultural resi-
due which represents a major concern in Mediterranean countries.
Soil as a Sink for Trace Metals
As Domergue and Vedy (1992) say, "mobility is a concept frequently used in soil
science to estimate the risk of contamination of other environmental compartments by
toxins, and especially by heavy metals." However, the definition of mobility depends to
a large extent on the scientific field where it is used. In the following text, we shall use
this term in the context of the dissolved metal forms in opposition to those bound to solid
phases, as the former are likely to be those responsible for immediate environmental
risks. Nevertheless, it must be kept in mind that long distance transport of heavy metals
occurs largely as part of fine solid particles of earth materials, causing a slow buildup of
soil pollution (Tiller, 1989).
Soils receive trace metals from various sources, mostly from lithogenic origin, i.e.,
from the parent material, or from anthropogenic origin; e.g., present in fertilizers, as part
of wastes (industrial, agricultural, or otherwise), irrigation waters, etc. (Singh and
Steinnes, 1994). When metals enter the soils, they undergo a number of reactions, namely
dissolution/precipitation, adsorption/desorption, complexation, inclusion in minerals
through formation of solid solutions (Alloway, 1990), which determine their "final" dis-
tribution in various chemical forms in the soil with very different solubilities. This final
201
202 Fate and Transport of Heavy Metals in the Vadose Zone
distribution is frequently described in terms of several fractions, operationally defined
according to a sequential extraction technique. Although different fractionation tech-
niques have been proposed, most of them distribute the soil metals in fractions called
"soluble," "exchangeable," "bound to carbonates," "bound to iron/manganese oxides,"
"bound to organic matter," or similar names, and that part of the metal content which
cannot be extracted but by strong concentrated acids is called "residual." Although these
names refer to a given chemical nature, in fact the fractions cannot usually be identified
with a definite group of chemical compounds. In the past, the term "speciation" was used
for the process of distinguishing different species of an element, and such species could
be defined either operationally (by the procedure or extractant used) or functionally
(bioavailable, exchangeable, etc.). Nowadays, the term "speciation" is preferred to be
applied to the distinction of specific chemical forms of an element (Ure et al., 1993).
Figure 10.1 is an example of the fractionation of some metals in Polish soils, where the
solubility or availability of each fraction increases from bottom to top (data from Kabata-
Pendias, 1995). Regardless of how the fractions are called, those fractions that are more
easily released usually account for a minor proportion of the metal content. Therefore, as
a consequence of the set of reactions mentioned above, soil acts as a geochemical sink for
metals and controls their transport to other media (Kabata-Pendias, 1995), so that their
persistence in soils is high: while the residence time of easily leachable elements like Ca,
Mg, etc., is usually below 3pO years, for many heavy metals residence times of several
millennia have been 1979). Consequently, the budgets for some trace
metals in soils are positive in many cases. Table 10.1 gives some examples for European
soils (from Kabata-Pendias, 1995). Only in the case of some acid forest soils, losses of
Mn or Zn are found.
Summarizing, metals tend to accumulate in soils, either by long-distance transport of
fine solid particles or by immobilization of the soluble forms, so that a sort of "chemical
time bomb" is building up. The environmental risk of such a chemical time bomb de-
pends whether various chemical "fuses" may cause a more or less sudden increase in
solubility, passing to ground or surface waters, or becoming available to plants.
Modeling Approaches for Retention of Metals by Soils
The use of mathematical models for a quantitative description of the various reactions
involved in a given process is a common practice in the study of environmental problems
(Basset and Melchior, 1990). Various models have been specifically developed for un-
derstanding surface reactions involving metals or other ionic species. Some of them,
which can be illustrated by those used by Davis and Leckie (1978) and by Barrow et aI.
(1981), assume some "mechanistic" characteristics of the retention process, e.g., pairwise
association between adsorbing ions and oppositely charged surface groups, or location
of adsorbing ions in planes of a given value of the electric potential. In both models, a
number of parameters that usually cannot be estimated by independent experiments
have to be adjusted by some iterative procedure.
Another group of models (Amacher et aI., 1986) do not assume any particular mecha-
nism for the retention reaction. Some of them use simple equations to describe equilib-
rium situations, e.g., one- or two-surface Langmuir equation, Freundlich equation, etc.
Others try to describe the time dependence of the solution concentration of the sorbing
Metal Retention and Mobility as Influenced by Some Organic Residues 203
100
-
80
!
=
8
-co
...
.s
60
....
=
...
5
Col
""
~
40
~
20
' ~ - - -
Cd
C:=:J Residual
rzza Organic matter
~ Bound to oxides
~ Exchangeable
IIIJJ Soluble
Zn Cu Pb
Figure 10.1. Example of fractionation of some metals in soils (from Kabata-Pendias, 1995, with
permission).
Table 10.1. Metal Budgets of Surface Soils in Europe (g ha-
1
yr-
1
)a
Ecosystem (Method) country Cd Cu Mn Pb Zn
Pine forest (seepage water) W. Germany 3 10 -360 104 134
Spruce forest (Iysimeter) Sweden -1 5 -600 75 -130
Farmland (drainage water) Denmark 3 260 130
Farmland (seepage water) Poland 2 14 90 160 360
a From Kabata-Pendias, A., in Heavy Metals, Problems and Solutions, W. Salomons, U. Forstner, and
P. Mader, Eds., Springer-Verlag, Berlin, 1995, pp. 3-18, with permission.
species, assuming reversibility or irreversibility and a given reaction order. In general,
these models are too simple and are not likely to represent the experimental data of
retention of metals by soils satisfactorily, probably because such retention is due to many
204 Fate and Transport of Heavy Metals in the Vadose Zone
different reactions, as mentioned above. That is why models considering the contribu-
tion of various simultaneous, concurrent reactions are probably more realistic. In an
exhaustive review, Selim (1992) described a number of multireaction kinetic models,
empirical in nature, based on the assumption that a fraction of the total sites are highly
kinetic, and the remaining sites interact slowly, irreversibly, or instantaneously with the
adsorbing species. Some models include concurrent and concurrent-consecutive pro-
cesses. Many metal retention processes are well described by this kind of model (Selim
et aI., 1990).
METAL CONCENTRATIONS IN THE SOIL SOLUTION
The various metal retention processes that may occur in soils cause the solution con-
centration of many metals to be generally below the values corresponding to the solubility
of pure components. A detailed description of the reactions controlling heary metal solu-
bility can be found elsewhere (McBride, 1989). The occurrence of coprecipitation of met-
als with the major metal oxides from soil solutions is accepted, but the nature of the
precipitate is not well defined. It can be easily shown (Driessens, 1986) that the solubility
of a metal can be lowered in a mixed ionic compound relative to the solubility of the pure
compound, and the activity of the metal compound is a function of its mole fraction. That
means that when a metal is incorporated in an iron or aluminum oxide in trace propor-
tions, the solution concentration of the former may be several orders of magnitude below
that expected from the solubility of the trace metal compound alone (McBride, 1989).
Various observations suggest that metals retained by oxides are concentrated at solid
surfaces (McBride, 1978), and little or no difference is found whether they are
"coprecipitated" or adsorbed on preexisting oxide surfaces. Formation of surface solid
solutions has been shown to be the cause for Cu adsorption by calcium carbonate
(Papadopoulos and Rowell, 1989), and by soils rich in lime (Madrid and Diaz-Barrientos,
1992), and in both cases the resulting equilibrium concentrations are several orders of
magnitude below that corresponding to the solubility of Cu carbonate. Figure 10.2 shows
that, while Ca solutions in the presence of CaC0
3
produced equilibria located close to
the solubility line of CaC0
3
in solubility diagrams, Cu solutions showed a definite ten-
dency to be in equilibrium with Cu(OHh, considerably less soluble than the metal car-
bonate. Madrid and Diaz-Barrientos (1992) treated calcareous soils with acetate buffer
at pH 5 to eliminate carbonates and equilibrated the original and treated soils with Cu
solutions and subsequently with Cu-free solutions, and found that all the equilibria with
the original samples corresponded with solubilities considerably below those of the ox-
ide, hydroxide or basic carbonate, while the equilibria with treated soils were close to the
line for CuO (Figure 10.3). Lindsay (1979) estimated that the solubility of Cu in soils
could be described by an equilibrium:
which causes a value of the ratio Cu
2
+/(H+)2 of 10
2
.
8
, giving Cu
2
+ concentrations several
orders of magnitude lower than those corresponding to the solubility of most Soil-Cu
/mmerals. Lindsay (1979) considered it likely that "soil-Cu" could correspond to cupric
fer0te, CuFe204' and suggested that the corresponding Zn compound, Zn-ferrite, could
Metal Retention and Mobility as Influenced by Some Organic Residues 205
Equilibrated with CU
2
+
.. Equilibrated with Ca
2
+

10
8


c.
6
4
2L---....lL.._--+----L __ --l
6 7 8 9
pH
CuO
CU
2
(OH)2COJ
CU(OH)2
CuCO
J
CaC0
3
Figure 10.2. Solubility diagrams for Cu
2
+ or Ca
2
+ solutions equilibrated with A.R. CaC0
3
(from
Papadopoulos and Rowell, 1989, with permission).
also account for the solubility of Zn in soils. However, other authors (Baker, 1990;
McBride, 1989) believe that formation of mixed oxides similar to cupric ferrite is un-
likely in the conditions found in soils. There is general consensus that adsorption on the
various surfaces present in soils is the main process responsible for the low metal solu-
tion concentrations often found in many soils. Only in some cases with high metal load-
ings do some solid phases determine metal solubility; e.g., malachite in soils with very
high eu contents (McBride and Bouldin, 1984; Lund and Fobian, 1991).
FACTORS CAUSING A REVERSAL OF IMMOBILIZATION
a large scale, soils tend to immobilize trace metals, some conditions of
the soil can undergo changes driving to a localized reversal of the process.
Metals with different valences are frequently more soluble in their lower oxidation states,
so that reductive conditions may cause a local increase in solubility, and a decrease in pH
will also cause an enhanced solubilization of heavy metals. A decrease in pH also causes
a decrease in negative electric charge of variable-charge surfaces, so that a decrease in
I
206 Fate and Transport of Heavy Metals in tile Vadose Zone
Adsorption Desorption
o c Untreated soils
'II'
.. Treated soils
0
-1
-2
.CU(OH)l
C
-3
=
U
'-'
~
-4
=
-
0
:t:
% 0
-5
0
0 c
0
0
c
0
0 c
c
-6
CC c
Cc
Cc
CC
C
C
-7
4 5
pH
6 7
Figure 10.3. Solubility diagrams for adsorption and desorption equilibria of Cu with soils rich in
lime, before and after being treated with acetate buffer at pH 5 (from Madrid and Diaz-Barrientos,
1992, with permission).
adsorption of cations also contributes to an increase in the presence of metals in solution
(McBride, 1989). Such effects are usually localized and easily reversed when the cause
disappears. An example of such reversible solubilizations can be found in the case of
some rivers close to mining areas, which may show low pH values and high concentra-
tions of several metals immediately downstream from the mining site, but a short dis-
tance further downstream the pH increases several units, metal concentrations decrease
sharply and the bank sediments show high contents in metal compounds of low solubil-
ity (Arambarri et aI., 1984).
The presence of organic compounds with functional groups with ability to form stable
complexes with transition elements is another factor which may increase the solution
concentration of heavy metals, and is a known fact since the end of the past century. In
contrast with acidification or with reducing conditions, formation of soluble complexes
may enhance the transport of metals in solution to long distances, due to the high stabil-
ity frequently shown by such complexes. Frequently a high selectivity is found in the
interaction of complexants with metals, as shown by Xue et aI. (1995), who found that
the concentration of free Cu
2
+ in a eutrophic lake was about 10
6
times lower than total
Cu concentration, whereas a substantial part of Zn was present as free Zn ions and
~ a k complexes.
Metal Retention and Mobility as Influenced by Some Organic Residues 207
The presence of a complexing agent in waters in contact with soils or sediments may
actively desorb metals from the solids, or hinder the adsorption processes responsible for
metal retention. This was shown to happen in the case of synthetic complexants added to
waters: in a study of nitriloacetic acid (NT A) as a possible alternative to polyphosphate
in detergents, it was found that the presence of NT A in waters resulted in a significant
increase in solubility of metals in river sediments and a decrease in their metal-sorbing
ability (Salomons and Forstner, 1984). Also, Davis and Singh (1995) found that when
soil columns contaminated with Zn were leached with EDT A solutions, the metal re-
tained was almost quantitatively removed. Recently, we have seen that eu added to soil
columns is strongly removed by glyphosate, a herbicide with zwitterion structure.
INTERACTION WITH NATURAL ORGANIC MATTER
The behavior of natural organic polymers as humic or fulvic acids is not as straight-
forward as that of simple complexants like NT A. Their action on heavy metals is two-
fold: on the one hand, soluble components of the humic substances can form complexes
which will be responsible for an enhanced solubility of metals, and, on the other hand,
fine solid particles of humic substances or organic coatings of minerals often contribute
to the retention (and immobilization) of metals. In either case, the interaction with the
metal ions occurs through various kinds of functional groups present in the organic poly-
mers (Senesi, 1992), and whether the interaction results in metal mobilization or immo-
bilization is strongly related with the size and solubility of the organics rather than with
the functional groups involved. Fulvic acids play an important role in the transport of
heavy metals in water, due to their lower molecular mass and much greater solubility as
compared to humic acids (Forstner and Wittman, 1983, and references therein). The
stability sequence of the complexes for some selected cations is considered to be (Klamberg
et aI., 1989)
Therefore, organic matterlheavy metal interactions determine three broad, dynami-
cally interrelated groups of species with different influence on the availability of the
metals for the living organisms (Figure lOA):
(a) Solid organic surfaces can retain metals, which are not immediately available
to plants. From this point of view, metals retained in this way behave similarly
to those retained by inorganic solid surfaces. However, some metals, e.g.,
cadmium, may show a preferential association with aqueous solution compo-
nents as compared to solid surfaces (Neal and Sposito, 1986).
(b) Some components of the natural organic matter can eventually be dissolved
and form metal complexes in solution, which in turn will be affected by
adsorption/desorption equilibria with the solid surfaces. Monomeric organic
molecules (e.g., released from plant roots) or large polymers can contribute to
these equilibria. If such complexes are strong enough, usually the toxic effects
of t ~ metal are ameliorated, but it has been observed that some lipid-soluble
comjexes can rapidly penetrate a biomembrane, so that substances forming
208 Fate and Transport of Heavy Metals in the Vadose Zone
Solid
Increasing bioavailability
Low mol. weight
complexes
:>
surfaces ... E1--------
Free
hydrated
metal
High mol. weight
complexes
Figure 10.4. Schematic distribution of metal species from the point of view of their availability to
biological systems.
this kind of complexes can cause an enhanced toxicity of the metal (Carlson-
Ekvall and Morrison, 1995).
(c) The free or weakly complexed metal ions will usually be the most bioavailable
species. Their concentration will be influenced by the various equilibria with
complexing species in solution and by adsorption/desorption reactions on
organic as well as inorganic solid surfaces.
Another important feature that distinguishes natural organic matter is the heteroge-
neity of its complexing sites. Figure 10.5 shows schematically the dependence of the
cumulative distribution of complexing sites of fulvic acid upon their stability c o n s t a n t ~
(based upon data of Buffle, 1988). Therefore, a single value for the stability constan:
cannot be defined, although different graphical approaches for estimating apparent o ~
"average" constants have been proposed (Fitch and Stevenson, 1984). Another useft:.
parameter in this context is the complexation capacity, defined as the maximum quantity 0:
a given metal that can be bound per gram of the substance (Perdue, 1988). The amoun r
and distribution of complexing sites of natural complexants have been estimated by vari-
0us techniques, especially those able to distinguish free from complexed metals. The use
of ion-selective electrodes (e.g., Buffle et al., 1977, 1980; Gamble et al., 1980; Stevensor.
and Chen, 1991; Stevenson et al., 1993) appeared to give reliable results in terms at
applicability and sensitivity (Sterrit and Lester, 1984). Polarographic techniques han
proved to give information of the lability of the complexes, and have been successfull.,.
applied to the characterization of metal complexes with humic and fulvic substance,
(Greter et al., 1979; Wilson et al., 1980; Filella et al., 1990; Pinheiro et al., 1994). Equili-
bration with ion exchangers has also been used for studying metal complexation with
organic matter. The method proposed by Schubert (1948) could be used only for mono-
nuclear complexes with respect to the metal ion, but Zunino et al. (1972a,b) proposed
some changes in the equations involved that allowed them to be applied to complexes at
the type Ma(ligandh, where a and 6 are integers;::: 1.
Metal Retention and Mobility as Influenced by Some Organic Residues 209
-- . --- - - --- --- ------------ --- --
-
80

:::t.
--
rIl 70

....
...
rIl
Cil
60
=
...
tIj
-
50
=-
a
Q
u
40
""'
Q
a.

30

a
:=
=
20



S
10
-:=
a
:=
0
U
4 5 6 7
10gKc
Figure 10.5. Example of cumulative distribution of metal complexing sites in a solution of fulvic
acid as a function of their stability constants.
EFFECT OF ORGANIC RESIDUES ON METAL SOLUBILITY
Land application is a widely used practice for disposal of many organic wastes. They
frequently contain substances which improve soil fertility, although other components
can have undesirable consequences for crops, so that research on many different aspects
of their addition to soils has increased in the last few years. One of the points that must
be carefully considered when such wastes are added to soils is their possible effect on
metal mobility or solubility: such organic residues are likely to contain polymers with
not very different properties from those of natural humic or fulvic substances, and there-
fore it can be expected that metal complexes similar to those known to form with the
latter will also be formed with such residues. In soils receiving dairy cattle-manure slurry,
Del Castilho et al. (1993) found significant increases in tKe concentrations of Cd, Zn,
and Cu in soil solutions. In the case of Cu, its concentrat4>ns were correlated with the
dissolved organic carbon, while Zn and Cd were also by low pH and high
levels of electric conductivity. They concluded that Cu showed high stability,
and a considerable part were of high molecular weight and nonlabile. Japenga et al.
(1992) also studied the effect of the liquid fraction of animal manure on heavy metal
solubilization in soil, and found a significant relationship between dissolved organic car-
bon and Cu concentrations in aqueous extracts (Figure 10.6). They concluded that, to-
gether with pH, complexation involving dissolved, high molecular weight organic matter
is the most important driving force for heavy metal solubilization. Metal complexation
was also considered to be one of the causes of metal leaching from a soil in a reed bed
210 Fate and Transport of Heavy Metals in the Vadose Zone
1.2 . - - - - - - , I - - - - ~ I----r-
I
---.--
I
----,
1.0 -
-
!l 0.8 f-
bi)
e
--
g 0.6 f-
Q
(J
=
U
0.4 f-
0.2 -
0.0
0
0
o
o
o
f8
~ o
0 I I I
200 400 600
DOC (mgI;l)
0
-
0
00
00
00%
0
0
0
-
0
0
-
-
-
I
800 1000
Figure 10.6. Relationship between dissolved organic carbon and Cu in aqueous extracts of a soil
mixed with liquid animal manure (selected data from Japenga et aI., 1992).
system continuously flooded with sewage (Wenzel et al., 1992). Barrado et al. (1995)
also concluded that extracts from eucalyptus and oak leaf litter showed complexation
ability for metals, and could estimate the complexing constants for various metals.
THE CASE OF SEWAGE SLUDGE
Addition of sewage sludge to soils was found to decrease the sorption of Cd at low
concentrations of this metal (Neal and Sposito, 1986). In soils treated with sewage sludge
and artificially contaminated with high doses of Cu in the form of Cu carbonate, Cheshire
et al. (1994), using electron magnetic resonance, found evidence of Cu solubilization
through complexation. The results for organically bound Cu in the soil solution indicate
oxygen ligand coordination in equatorial arrangement. Keefer et al. (1994) also found
significant metal-organic association in soils amended with sewage sludge, and McGrath
and Cegarra (1992) observed large increases in the most soluble fractions of metals in a
soil with long-term applications of sewage sludge. They found that the fractionation of
metals in the original sewage sludge differed from that observed in the soil treated with
the residue.
Frequently, sewage sludges have relatively high metal contents, so that their effect on
metal mobility in soils has been often attributed to the metals present in the residue itself:
Sposito et al. (1982) concluded that the accumulatienohnetals in soils amended with
sewage sludge was governed by the metal content in the sludge, and Cavallaro et al.
Metal Retention and Mobility as Influenced by Some Organic Residues 211
Table 10.2. Examples of Maximum Permissible Concentrations of Some Metals in
Soil after Application of Sewage Sludge (mg kg-
1
)
Soil pH Country Cd Cu Ni Pb Zn
5.0-5.5 UK
a
80 50 200
5.5-6.0 UK
a
100 60 250
6.0-7.0 UK
a
135 75 300
> 7.0 UK
a
200 11P 450
> 5.0 UK
a
3 300
< 7.0 Spain
b
1 50 30 50 150
> 7.0 Spain
b
3 210 112 300 450
a Data selected from Department of the Environment, Code of Practice for Agricultural
Use of Sewage Sludge, HMSO Publications, London, 1992, p. 6.
b Data selected from Boletin Ofloal del Estado, No. 262, Madrid, 1990, p. 3234.
(1993) found that increases in Mehlich-3 extractable eu and Zn in soils treated with
sewage sludge were similar to the amounts of these metals added in the residue. Most
countries have established regulations concerning the use of sewage sludge on the basis
of the maximum permissible contents of potentially toxic elements in soil after applica-
tion of sewage sludge (Table 10.2) and annual rate of addition of such toxic elements, so
that no legal limit exists if the sewage sludge added to a soil shows a low content in toxic
metals. It is thus forgotten that solubilization of the metals already present in the soil can
be enhanced by complexation, as shown by some of the authors quoted in the previous
paragraph. This lack of attention paid to the effect of soil management practices, espe-
cially the use of sewage sludge, on the solubility of the metals present in amounts below
the legal limits in the soil has been claimed by several authors (McBride, 1994; Evans et
aI., 1995), and has been favored by the conclusions of some authors, who even found a
decrease in metal mobility in some cases (Emmerich et aI., 1982; Saviozzi et aI., 1983;
Hooda and Alloway, 1994).
A MEDITERRANEAN CONCERN: OLIVE Mill WASTEWATER
Setting Up the Problem
In areas with extensive production of olive oil, disposing of the residues from manufac-
turing plants for this agricultural product represent a major concern. The traditional pro-
cedure implies generating large amounts of wastewater (called aLpecbin, from now on
OMW) with extremely high BOD and other undesirable properties which have caused
the existence of regulations prohibiting its disposal in rivers since 1981. Everyyear, about
10 million tons of this waste have to be disposed of in the Mediterranean countries, mainly
by storing them in evaporation ponds, composting the resulting sludge with other plant
refuse or, in countries where the production of this residue is not especially high, dis-
charging them into watercourses. In recent years, olive oil production plants are being
adapted for new techniques using much smaller volumes of water, so that production of
OMW is decreasing sharply, but its disposal must still be considered until total substitu-
tion of the old manufacturing plants, and the existence of small factories which cannot
afford the changes cannot be forgotten, at least during several years in the near future.
212 Fate and Transport of Heavy Metals in the Vadose Zone
While the effect of sewage sludge on heavy metal availability has been extensively
studied, as summarized in the previous section, literature on the relationships between
heavy metals and OMW is scarce, although in the last few years some authors have
found evidence of significant metal solubilizing effects of this residue, both when added
to soils and when present in freshwaters. OMW is a slightly acid (pH 4-5), dark-colored
aqueous phase with highly variable composition, containing 10-15% organic matter and
1-2% of mineral salts. Its contents in heavy metals' is usually negligible, except Fe (10-
20 mg L-
1
), Mn 5 mg L-
1
), and Zn 2 mg L-
1
). Several authors have given detailed
descriptions of the composition of this waste (e.g., Gonzalez-Vila et aI., 1992; Martinez-
Nieto and Garrido-Hoyos, 1994).
Effect of OMW on Metal Retention Properties of Soils
As with other organic wastes, one of the first ideas that emerge when recycling OMW
is considered is its application to soils as fertilizer, and it has been frequently used to
irrigate olive trees. Considering previous knowledge of the nature of this residue, Saiz-
Jimenez et aI. (1987) deemed it of interest to carry out a detailed study of its chemical
composition in order to evaluate its potential value as soil fertilizer. They concluded that
the composition of the humic fraction was different from soil humic acids, but still sug-
gested that the residue had good properties as fertilizer. On a relatively short-term basis,
applications of a composted olive mill sludge to soils have been found to cause no harm-
ful effects on plants, the improvement of soil physical properties is apparent, and signifi-
cant increases are found in soil organic N. Also, increases in available Cu, Zn, Mn, and
Fe determined by DTPA extraction have been observed (Martin-Olmedo et al., 1995).
In a study specifically oriented toward the effect of OMW on metal availability, Perez
and Gallardo-Lara (1993) found that although OMW initially caused a slight decrease
in Zn availability and hardly any effect on Cu availability, a significant residual increase
in Cu availability was observed after growing barley and ryegrass.
A fundamental aspect that must be considered is whether the presence of OMW af-
fects the action of soils as a sink of heavy metals which are added in soluble forms.
Madrid and Diaz-Barrientos (1994) chose three soils (called A, B, and C) with widely
differing contents in organic matter, carbonate, and clay fraction and CECs for that
purpose. Soil A had been manured in the field with 150,000 kg ha-
I
of a compost ob-
tained from OMW and other plant refuse, and soil B had received a similar dose of raw
OMW. Soil C was untreated. Moreover, samples of the three soils were aged in vitro
with freeze-dried OMW in a proportion corresponding to twice the dose received by
soils A and B in the field. The reaction of several metals with the original, manured, and
aged samples was studied. Figure 10.7 shows the results for the adsorption isotherms
obtained for Cu and Zn.
The adsorption of these two metals was strongly decreased by mixing and aging the
soils with OMW, while manuring with OMW or compost obtained from it only caused
a significant decrease in the case of Zn. Manuring even caused a slight increase in Cu
adsorption by soil A. The pH values of the adsorption experiments did not show differ-
ences large enough to explain the differences in adsorption. The authors suggested that
the decrease in adsorption when OMW was added in large doses could be the result of
coating the sorbing surfaces with organic matter. However, in the samples containing
o Original samples - Manured samples \l Aged samples
; / J::
80 )7 \I l 90 150
- 60
'7bt) I II if 60 f-'Jf r:f -I 100
40
1 20 f Cu/A jl 30 fg d " fD -I 50
!. 0 0 t7" j 0
] 200,0, 10 30 40 150 t 4 6 8 10 12 150 0 5 10 15 20
a..

! 150 Ii Zn/A 100 f- / _ -I 100
.f!
100
50 _____________ 50 iI / 50
ZnlB
o I j 0 0 '---'-----"--L-L-L----'----------'---L
o 10 20 30 40 50 60 70 80 0 10 20 30 40 50 0 10 20 30 40 50 60 70
Final metal concentration (J.&mol 1;1)
Figure 10.7. Adsorption isotherms of Cu and Zn by three soils (A, B, and C), before and after different treatments: Soils A and B were manured with
composted and raw OMW, respectively, and the three soils were aged in vitro with a high dose of freeze-dried OMW (from Madrid and Diaz-
Barrientos, 1994, with permission).



;0
I'll
....
I'll
:::l
....
o
:::l
III
:::I
C.

o
g
;:;:
'<
'Ill
VI
a
E"
I'll
:::l
I"'l
I'll
C.
c;
'<
VI
o
3
I'll
o
....
ttQ
III
:::l
n
?i
VI
C.'
C
I'll
VI
N
....
w
214 Fate and Transport of Heavy Metals in the Vadose Zone
OMW some oversaturation was found in eu experiments with two of the soils, both in
the adsorption and the desorption steps, with respect to the eu compounds likely to
form. This result suggests that formation of soluble complexes could also contribute to
the decrease in adsorption of this metal.
Selim et al. (1990) considered the adsorption reactions caused by several kinds of
surfaces, with different kinetic characteristics. Madrid and Diaz-Barrientos (1996) ex-
amined whether the effect of OMW on the metal-sorbing properties of soils manifested
itself when the multireaction model (MRM) of Selim et al. (1990) was applied to kinetic
data of eu adsorption by two soils which had received OMW compost. The MRM model
distinguished several fractions in the amount sorbed: Se' that reacts "instantaneously"
with the sorbate in solution, described by a Freundlich equation with a distribution coef-
ficient ~ and exponent" (Equation 1); s, and S2' which react kinetically and reversibly
with the solution at different rate, described by the forward and backward rate coeffi-
cients k, and k2 for s, and k3 and k4 for S2' and reaction orders nand m respectively
(Equations 2 and 3); and finally Sirr' which reacts kinetically and irreversibly with the
solution by a first order reaction, described by a rate coefficient ks (Equation 4). The
original model considered that some of the fractions could undergo other consecutive
reactions, generating further fractions, but Madrid and Diaz-Barrientos (1996) found
out that this latter assumption did not increase the percentage of variance explained
when the model was applied to their data. The mathematical structure of the model as
used by Madrid and Diaz-Barrientos (1996) is summarized below, where p and e repre-
sent the bulk density and water c o n t ~ n t of the soil in the experiments.
(1)
(2)
(3)
(4)
Table 10.3 shows the parameters estimated by Madrid and Diaz-Barrientos (1996)
for their original samples and those treated with OMW compost. It can be observed that
addition of OMW to the soils causes significant changes on some of the parameters: the
instantaneous distribution coefficient ~ is strongly increased by a factor of 50 in both
soils; the irreversible fraction Sirr becomes irrelevant, and the rate coefficients of the
kinetic fraction s, are significantly decreased. S2 does not show any significant effect.
Madrid and Diaz-Barrientos (1996) concluded that the eu-immobilizing action of the
soils was altered by the presence of composted OMW, probably by the presence of new
solid, organic surfaces, which react instantaneously with the metal in solution. The de-
pendence of the instantaneous fraction upon the solution concentration of eu means that
the metal adsorbed by the soil with OMW must easily come into solution if its concen-
tration decreases, in contrast with the behavior of the untreated soils, which hardly re-
lease eu by dilution. The disappearance of fraction Sirr also means a decrease in the
immobilization of the metal.
Metal Retention and Mobility as Influenced by Some Organic Residues 215
Table 10.3. Average Values of the Model Parameters. k, to k4 and kSl h-'; b, nand m, dimension-
less; Ku mg kg-' a
Parameters of Each Fraction of the Model
se
S,
S2 sirr
Soil Kd b k, k2 n
k3 k4
m
ks
R2
A orig. 9 0.4 24.5 2.79 0.3 0.33 0.13 0.7 0.04 0.998
A compo 440 0.4 1.82 0.79 0.4 0.50 0.04 0.5 0 0.997
C orig. 18 0.4 28.4 1.15 0.4 0.61 0.08 0.7 0.15 0.995
C compo 925 0.9 9.7 0.45 0.4 0.77 0.05 0.4 0 0.994
a From Madrid, L. and E. Diaz-Barrientos, Taxicol. Environ. Chern. 54, pp. 93-98, 1996, with
permission.
OMW in the Aqueous Phase as a Mobilizing Agent of
Insoluble Metal Forms
Although discharges ofOMW in watercourses is prohibited, its accumulation in ponds
may cause a slow migration of its soluble components to groundwaters. Moreover, acci-
dental releases of significant amounts into rivers do occur. Thus, another important point
that must be considered is the solubilization of metals in "immobile" forms when in con-
tact with water containing OMW. Bejarano and Madrid (1992, 1996a) studied the solu-
bilization of heavy metals from river sediments, with high metal contents due to their
location close to mining sites, when treated in vitro with dilute solutions of OMW. These
solutions were prepared from freeze-dried OMW and adjllsted at pH values between 3
and 5, considering the slightly acid pH of the residue. They found that some metals, e.g.,
Mo or Zn, were not solubilized by the residue, and in the case of Mo the sediment even
retained part of the metal originally present in the OMW. On the contrary, Cu, Fe, and
Pb from the sediments were solubilized when in contact with OMW solutions. The pres-
ence of OMW favored Pb solubilization at any pH, while Cu and Fe were dissolved to a
greater extent than in the absence of OMW only at the higher pH tested. Considering
that the solubility of these metals usually increases at lower pH values, this result sug-
gests that the solubilization of these two metals can be related with the formation of
soluble complexes with OMW components. Figure 10.8 shows a summary of the results
obtained by the authors for Pb and Cu. In the second paper mentioned (Bejarano and
Madrid, 1996a) the authors showed that the amounts solubilized by OMW were compa-
rable with the metals originally present in the sediments in forms bound to carbonates
and to oxides, according to a conventional fractionation technique.
The hypothesis of complexation by OMW components had been previously checked
by Cabrera et al. (1986). Using the cation-exchange resin method of Zunino et al. (1972b),
they found that the freeze-dried residue showed a complexing ability of 0.66 mmol of Cu
per gram of OMW. Bejarano and Madrid (1996b) studied the time-dependence of the
release of several metals by solutions with three different OMW concentrations from a
river sediment, and the resulting solutions were filtered through C-18 reverse-phase
cartridges. The metals complexed by polymers present in the OMW, especially those
forming less labile complexes, were supposed to be retained by the cartridges, together
with the uncomplexed organic polymers. Previously they checked that no free metal was
retained in the cartridges. Figure 10.9 summarizes the results for Cu, Mo, and Zn. The
216 Fate and Transport of Heavy Metals in the Vadose Zone
------. ----- -- ------------- ---
0.7 10
0.6
-
8
....
-
,!..;l 0.5 ....
OJ) ,!..;l
!.
OJ)
0.4
a 6
1 -
"C
..
0.3
Q,j
-
Q ..
"-1
- "-1
Q
4
...
~
"C
0.2
...
,Q "C
~
='
0.1
U
2
0.0
o 5 10 15 20 25 30 35 o 5 10 15 20 25 30 35
OMW concentration (g C 1)
Figure 10.S. Pb and Cu dissolved from a river sediment by solutions of various OMW concentra-
tions and pH values (from Bejarano and Madrid, 1992, with permission).
authors concluded that most of the Cu and Zn released into solution by the OMW was
in complexed forms, as nearly 100% of both metals were retained by the cartridges. In
the case of Mn, only about 50% of the metal dissolved by OMW was retained by the
cartridges, suggesting that complexation was less complete or that the complexes formed
were more labile.
Considering that some of the results commented on in this and the previous section
suggest that some components of OMW do form complexes with several metals, Bejarano
and Madrid (1996c) appfied the techniques of Zunino et al. (1972a,b) to determining the
complexation parameters of this residue for several metals. They found that the maxi-
mum complexing ability (MCA) was inversely related with the ionic radii of the metal
ions, and a direct dependence between the logarithm of the conditional stability con-
stants and the metal electronegativity (Figure 10.10). This latter result agrees with the
fact that the stability of complexes formed by a given ligand with a series of metals is
expected to increase with the electronegativity of the metals (Irving and Williams, 1948),
thus showing indirectly that complexation by OMW does occur. The authors concluded
that the "average" complex formed is mono-nuclear, with a bidentate bond for Cu
2
+. For
other larger M2+ ions, a progressive steric hindrance seems to exist.
This simple model of a mono-nuclear, bidentate complex between the "average" com-
ponent of OMW and a metal ion was further developed elsewhere (Bejarano and Madrid,
1996d). The reaction was assumed to be

...

'"
o v
o
x
1.6
1.2
1.2
-
....
r ~ l f
J
'...:l
~
B 0.8

-
a 0.4 f
~
0.4 f- 11\ r
0.0
0.0
o 5 10 15 20 o
t(days)
1,2,3 g L-
I
OMW and blank, total extracted
1,2,3 g L-
I
OMW and blank after filtration
4
3
-
....
'...:l
~
B 2
-
=
N
-! 1
o
5 10 15 20 o 5 10 15
t (days) t (days)
20
Figure 10.9. Time dependence of metals found in solutions of OMW in contact with a river sediment, and effect of filtering the solutions through
(-18 reverse-phase cartridges (from Bejarano and Madrid, 1996, with permission),
3:
rc
....
to;
;0
rc
...
~
...
o
::::l
to;
:::l
C.
3:
o
g
;:;:
'<
to;
VI
S'
.....
i:
rc
::::l
,-.,
rc
C.
0-
'<
III
o
3
t'P
o
I cia
III
::I
n'
;0
1'0
VI
is:
c
1'0
\II
N
....
'1
218 Fate and Transport of Heavy Metals in the Vadose Zone
0.5 r--O---r,-----r-,----,r-"'!
Cu
~ 0.4 I-
-
~
-
I
-
~ 0.3 t-
O
~
:::: 0.2 -
Q
.!
~ 0.1 -
~
Zn
o OMn
-
Pb -
o
1 _1 I
0.0 ' - - - - - ' - - - - - - ' - - - - - ~
0.6 0.8 1.0 1.2
M2+ ionic radius (A)
5 r--'I"--'I--'I--"--"---'
4 I-
3 -
Mn
o
Zn
o
Pb
o
Cu
o
-
2 L - _ ~ ~ __ ~ 1 __ L - 1 _ ~ 1 L - _ ~ 1 __ ~
1.4 1.5 1 ~ 1 ~ 1 ~ 13 2.0
Metal electronegativity
Figure 10.10. Relationship between maximum complexing ability of OMW and the radius of each
metal ion, and dependence of the stability constant on the metal electronegativity (from Bejarano
and Madrid, 1996, with permission).
A double acid dissociation was assumed for the organic ligand, defined by two acid
dissociation constants. From Cu
2
+ and Zn
2
+ data of experiments similar to those men-
tioned above with reverse-phase cartridges, estimates of free metal concentration [M2+]
and, by difference, of complexed metal [ML] were obtained. Total ligand concentra-
tions were considered to be those of the residue (in g L -I). By an iterative, computer
simplex method the authors obtained values for the conditional constant KML and the
two acid dissociation constants. The graphs in Figure 10.11 show the experimental data
and the corresponding calculated solution compositions. As can be seen, the model was
a good approximation to the behavior of OMW as complexant. The values of the stabil-
ity constants obtained by this model (Bejarano and Madrid, 1996d) and by a cation
exchange resin (Bejarano and Madrid, 1996c) were reasonably congruent despite the
different techniques and conditions, and agreed with that previously obtained for Cu
2
+
by Bejarano et al. (1994) using voltammetric techniques.
SUMMARY
During the last decades, the view of land as a sink for any waste has been ruled out as
erroneous, and concern for the long-term environmental hazards of accumulation of
wastes has gained increasing importance. In the preceding pages we have tried to show
that, even though soils can "fix" large quantities of potentially toxic metals, acting as a
barrier against metal pollution of ground and surface waters, organic matter, either natural
or, especially, added to soils, is a very important factor able to change the status of the
metals in the system. Whether such change is in the direction of increasing the fixation of
metals or of mobilizing them depends on several circumstances, but many studies sug-
gest that the presence of soluble organic matter generally increases metal mobility through
formation of soluble complexes. Therefore, those processes for maturing organic wastes
Metal Retention and Mobility as Influenced by Some Organic Residues 219

o
-1
Ig L OMW;
-4.00 r----,,--"--,..- ,--'--,--,'-'
-4.25 - -
-4.50 _ ill ~
-
o i ~
~ -4.75 I- -
-
-5.00 r -
-5.25 L-_...L_I __ -,---l_...ll __ -'- '_...J'LJ
5.0 5.5 6.0 6.5 7.0 7.5
pH
n -1
v 2g L OMW;
I I
-4.0 r
.. ~

-4.5 I- o ~
~

-6.0 -
o 3gC
1
0MW
,
'Y
'\} ~ .
'Y.
o
'\}
o
, ,
-
-
-
-
-
I I I I I
5.25 6.00 6.75 7.50 8.25
pH
Figure 10.11. Comparison of the metal concentrations found in solutions of OMW equilibrated
with a river sediment and subsequently filtered through C-18 reverse-phase cartridges (hollow
symbols) and values calculated by a simple model of mono-nuclear bidentate complexation (filled
symbols) (From Bejarano and Madrid, 1996, with permission).
previous to their use as soil amendments must be aimed, among other purposes, at mini-
mizing the proportion of soluble components. Thus, the mobilizing action will also be
minimized and the resulting highly polymerized, sparingly soluble organic wastes will
contribute to immobilize metals and consequently will help to keep a low bioavailability
of such potentially toxic elements.
REFERENCES
Alloway, B.J. Soil Processes and the Behaviour of Metals, in Heavy Metau in Soiu, B.J. Alloway,
Ed., Blackie and Son, Glasgow, 1990, pp. 261-279.
Amacher, M.C., J. Kotuby-Amacher, H.M. Selim, and 1.K. Iskandar. Retention and release of
metals by soils-Evaluation of several models. Geooerma, 38, pp. 131-154, 1986.
Arambarri, P., F. Cabrera, and C.G. Toea, La ContaminaciOn oet Rio Guaoiamar Y .Ill Zona oe
InfLuencia, MarumaJ oeL GuaoaLqllirir y Coto 1)onana, por ReJiOuoJ Oe InollJtrlaJ MineraJ y AgrfcolM.
CS I C, Madrid, 1984.
Baker, D.E. Copper, in Heavy Metau in Soiu, B.J. Alloway, Ed., Blackie and Son, Glasgow, 1990,
pp. 151-176.
Barrado, E., M.H. Vela, R. Pardo, and F. Rey. Determination of acidity and metal ion complexing
constants for eucalyptus and oak extracts. Commlln. Sotl Sci. Plant AnaL., 26, pp. 2067-2078,
1995.
Barrow, N.J., J.W. Bowden, A.M. Posner, and J.P. Quirk. Describing the adsorption of copper,
zinc and lead on a variable charge mineral surface. AwtraL. J. SoiL Red., 19, pp. 309-321, 1981.
220 Fate and Transport of Heavy Metals in the Vadose Zone
Basset, R.L. and D.C. Melchior, Chemical Modeling of Aqueous Systems: An Overview, in
ChemicaL MOdeling of AquoUJ SY.Jtenu II, D.C. Melchior and R.L. Basset, Eds., ACS Symposium
Series No. 416, American Chemical Society, Washington, DC, 1990, pp. 1-14.
Bejarano, M. and L. Madrid. Solubilization of heavy metals from a river sediment by a residue
from olive oil industry. Environ. TechnoL., 13, pp. 979-985, 1992.
Bejarano, M. and L. Madrid. Solubilization of heavy metals from a river sediment by an olive mill
effluent at different pH values. Environ. Techno!., 17, pp. 427-432, 1996a.
Bejarano, M. and L. Madrid. Release of heavy metals from a river sediment by a synthetic
polymer and an agricultural residue: Variation with time of contact. ToxicoL. Environ. Chem., 55,
pp. 95-102, 1996b.
Bejarano, M. and L. Madrid. Complexation parameters of heavy metals by olive mill wastewater
determined by a cation exchange resin. J. Environ. SCt: HeaLth, B31. pp. 1085-1101, 1996c.
Bejarano, M. and L. Madrid. Solubilization of Zn, Cu, Mn and Fe from a river sediment by olive
mill wastewater: Influence of Cu or Zn. Toa:ico!. Environ. Chem., 55, pp. 83-93, 1996d.
Bejarano, M., A.M. Mota, M.L.S. Gonr;alves, and L. Madrid. Complexation ofPb(II) and Cu(II)
with a residue from the olive-oil industry and a synthetic polymer by DPASV. Sci. TotaL
Environ., 158, pp. 9-19, 1994.
Boldin OficiaL deL &tado, Madrid, No. 262, p. 3234, 1990.
Bowen, H.J.M. EnvironmentaL Chemutry of the ELement.J. Academic Press, London, 1979, p. 333.
Buffle, J. Complexation Reaction,} in Aquatic SY.Jtenu:AnAnaLyticaLApproach. Ellis Horwood, Chichester,
1988, p. 33.
Buffle, J., F.-L. Greter, and W. Haerdi. Measurement of complexation properties of humic and
fulvic acids in natural waters with lead and copper ion-selective electrodes. AnaL. Chem., 49, pp.
216-222, 1977.
Buffle, J., P. Deladoey, F.L. Greter, and W. Haerdi. Study of the complex formation of
copper(II) by humic and fulvic substances. AnaLytica Chimica Acta, 116, pp. 255-274, 1980.
Cabrera, F., M. Soldevilla, F. Osta, and P. Arambarri. Interacci6n de Cobre y Alpechines.
Limnitica, 2, pp. 311-316, 1986.
Carlson-EkvalL C.E.A. and G.M. Morrison. Toxicity of copper in the presence of organic
substances in sewage sludge. Environ. Techno!., 16, pp. 243-251, 1995.
Cavallaro, N., N. Padilla, and J. Villarrubia. Sewage sludge effects on chemical properties of acid
soils. SoiL Sci., 156, pp. 63-70, 1993.
Cheshire, M.V., D.B. McPhaiL and M.L. Berrow. Organic matter-copper complexes in soils
treated with sewage sludge. Sci. TotaL Enriron., 152, pp. 63-72, 1994.
Davis, A.P. and I. Singh. Washing of zinc(II) from contaminated soil column. J. Environ. Eng.,
121, pp. 174-185, 1995.
Davis, J.A. and J.O. Leckie. Surface ionization and complexation at the oxide/water interface.
2. Surface properties of amorphous iron oxyhydroxide and adsorption of metal ions. J. CoLloid
Inteiface Sci., 67, pp. 90-107, 1978.
Del Castilho, P., W.J. Chardon, and W. Salomons. Influence of cattle-manure slurry application
on the solubility of cadmium, copper, and zinc in a manured acidic, loamy-sand soil. J. Environ.
QuaL., 22, pp. 689-697, 1993.
Department of the Environment, COde of Practice for AgricuLturaL U.Je of Sewage SLudge. HMSO
Publications, London, 1992, p. 6.
Domergue, F.-L. and J.-C. Vedy. Mobility of heavy metals in soil profiles. Int. J. Environ. AnaL.
Chem., 46, pp. 13-23, 1992.
Driessens, F.C.M. Ionic Solid Solutions in Contact with Aqueous Solutions, in Geochemical
Proce.J.Je.J at MineraL Suiface.J, J.A. Davis and K.F. Hayes, Eds., ACS Symposium Series No. 323,
American Chemical Society, Washington, DC, 1986, pp. 524-560.
Metal Retention and Mobility as Influenced hy Some Organic Residues 221
Emmerich, W.E., L.J. Lund, A.L. Page, and A.C. Chang. Solid phase forms of heavy metals in
sewage sludge-treated soils. J. Environ. Qual., 11, pp. 178-181, 1982.
Evans, L.J., G.A. Spiers, and G. Zhao. Chemical aspects of heavy metal solubility with reference
to sewage sludge amended soils. Int. J. Environ. Anal. Chem., 59, pp. 291-302, 1995.
Filella, M., J. Buffle, and H.P. van Leeuwen. Effect of physico-chemical heterogeneity of natural
complexants. Part 1. Voltammetry oflabile metal-fulvic complexes. Analytica ChimicaActa, 232,
pp. 209-223, 1990.
Fitch, A. and F.J. Stevenson. Comparison of models for determining stability constants of metal
complexes with humic substances. Soil Sci. Soc. Am. J., 48, pp. 1044-1050, 1984.
Forstner, U. and G.T.W. Wittman. Metal Pollution in the Aquatic Environment, 2nd ed., Springer-
Verlag, Berlin, 1983, p. 223.
Gamble, D.S., A.W. Underdown, and C.H. Langford. Copper(II) titration of fulvic acid ligand
sites with theoretical, potentiometric, and spectrophotometric analysis. Anal. Chem., 52, pp.
1901-1908, 1980.
Gonzalez-Vila, F.J., T. Verdejo, and F. Martin. Characterization of wastes from olive and
sugarbeet processing industries and effects of their application upon the organic fraction of
agricultural soils. Int. J. Environ. Anal. Chem., 46, pp. 213-222, 1992.
Greter, F.-L., J. Buffle, and W. Haerdi. Voltammetric study of humic and fulvic substances. Part
I. Study of the factors influencing the measurement of their complexing properties with lead.
J. Electroanal. Chem., 101, pp. 211-229, 1979.
Hooda, P.S. and B.J. Alloway. Sorption of Cd and Pb by selected temperate and semi-arid soils:
Effects of sludge application and aging of sludged soils. Water Air Soil Pollut., 74, pp. 235-250,
1994.
Irving, H. and R.J.P. Williams. Order of stability of metal complexes. Nature, 162, pp. 746-747,
1948.
Japenga, J., J.W. Dalenberg, D. Wiersma, S.D. Scheltens, D. Hesterberg, and W. Salomons.
Effect of liquid animal manure application on the solubilization of heavy metals from soil. Int.
J. Environ. Anal. Chem., 46, pp. 25-39, 1992.
Kabata-Pendias, A. Agricultural Problems Related to Excessive Trace Metal Contents of Soils,
in Heavy Metau, Problem! and Solutiow, W. Salomons, U. Forstner, and P. Mader, Eds.,
Springer-Verlag, Berlin, 1995, pp. 3-18.
Keefer, R.F., S.M. Mushiri, and R.N. Singh. Metal-organic associations in two extracts from nine
soils amended with three sewage sludges. Agric., &o.JY.Jt. Environ., 50, pp. 151-163, 1994.
Klamberg, H., G. Matthess, and A. Pekdeger. Organo-Metal Complexes as Mobility-Determin-
ing Factors of Inorganic Toxic Elements in Porous Media, in Inorganic Contaminant.J in the
Vado.Je Zone, B. Bar-Yosef, N.J. Barrow, and J. Goldshmid, Eds., Ecological Studies Series,
Vol. 74, Springer-Verlag, Berlin, 1989, pp. 3-17.
Lindsay, W.L. Chemical Equilibria in Soiu. John Wiley & Sons, New York. 1979, p. 222.
Lund, U. and A. Fobian. Pollution of two soils by arsenic, chromium and copper. Geoderma, 49,
pp. 83-103, 1991.
Madrid, L. and E. Diaz-Barrientos. Influence of carbonate on the reaction of heavy metals in
soils. J. Soil Sci., 43, pp. 709-721, 1992.
Madrid, L. and E. Diaz-Barrientos. Retention of heavy metals by soils in the presence of a residue
from the olive-oil industry. Eur. J. Soil Sci., 45, pp. 71-77, 1994.
Madrid, L. and . Diaz-Barrientos. Nature of the action of a compost from olive mill wastewater
on Cu sorpti n by soils. Toxicol. Environ. Chern., 54, pp. 93-98, 1996.
Martin-Olme ,P., F. Cabrera, R. LOpez, and J.M. Murillo. Successive applications of a
compo ed olive oil mill sludge: Effect on some selected soil characteristics. FreJeniw Environ.
Bull., 4, pp. 221-226, 1995.
222 Fate and Transport of Heavy Metals in the Vadose Zone
Martinez-Nieto, L. and S.E. Garrido-Hoyos. El Alpechin, un Problema Medioambiental en Vias
de Solucion (I). Quimica e I,wlMtria, 17-28, November 1994.
McBride, M.B. Retention of Cu
2
+, Ca
2
+, Mg2+, and Mn
2
+ by amorphous alumina. Soil Sci. Society
Am. J., 42, pp. 27-31, 1978.
McBride, M.B. Reactions controlling heavy metal solubility in soils. Aov. Soil Sci., 10, pp. 1-56,
1989.
McBride, M.B. Toxic metal accumulation from agricultural use of sludge: Are USEPA regula-
tions protective? J. Environ. QuaL., 24, pp. 5-18, 1994.
McBride, M.B. and D.R Bouldin. Long-term reactions of copper(II) in a contaminated calcar-
eous soil. SoiL Sci. Soc. Am. J., 48, pp. 56-59, 1984.
McGrath, S.P. and J. Cegarra. Chemical extractability of heavy metals during and after long-
term applications of sewage sludge to soil. J. Soil Sci., 43, pp. 313-321, 1992.
Neal. RH. and G. Sposito. Effects of soluble organic matter and sewage sludge amendments on
cadmium sorption by soils at low cadmium concentrations. SoiL Sci., 142, pp. 164-172, 1986.
Papadopoulos, P. and D.L. Rowell. The reactions of copper and zinc with calcium carbonate
surfaces. J. SoiL Sci., 40, pp. 39-48, 1989.
Perdue, E.M. Measurements of Binding Site Concentrations in Humic Substances, in Meta!
Speciation: Theory, AnaLY.JiI anO AppLication, J.R Kramer and H.E. Allen, Eds., Lewis Publishers,
Boca Raton, FL, 1988, pp. 135-154.
Perez, J.D. and F. Gallardo-Lara. Direct, delayed and residual effects of applied wastewater
from olive processing on zinc and copper availability in the soil-plant system. J. Environ. Sci.
HeaLth, B28, pp. 305-324, 1993.
Pinheiro, J.P., A.M. Mota, and M.L. Simoes Complexation study of humic acids with
cadmium(II) and lead(II). AnaLytica Chimica Acta, 284, pp. 525-537, 1994.
Saiz-Jimenez, C., J. W. De Leeuw, and G. Gomez-Alarcon. Sludge from the waste water of the
olive processing industry: A potential soil fertilizer? Set: TotaL Environ., 62, pp. 445-452, 1987.
Salomons, W. and U. Forstner. Metal.! in the Hyorocycle. Springer-Verlag, Berlin, 1984, pp. 176-
179.
Saviozzi, A., R. Levi-Minzi, and R Riffaldi. How organic matter sources affect cadmium
movement in soil. Bwcycle, pp. 29-31, May/June 1983.
Schubert, J. The use of ion exchangers for the determination of physical chemical properties of
substances particularly radio tracers in solution. J. Phy.J. Colloid Chem., 52, pp. 340-356, 1948.
Selim, H.M. Modeling the transport and retention of inorganics in soils. Aov. Agron., 47, pp. 331-
384, 1992.
Selim, H.M., M.C. Amacher, and 1.K. Iskandar. MooeLing the Tran.Jport 0/ Heavy Metal.! in Soil.!. U.S.
Army Corps of Engineers, CRREL Monograph 90-2, U.S. Government Printing Office,
1990.
Senesi, N. Substance Complexes in the Environment. Molecular and Mechanistic
Aspects by Approach, in Bwgeochemiltry 0/ Trace Metal.!, D. C. Adriano,
Ed., Springer-Verla, New York, 1992, pp. 429-496.
Singh, B.R and E. St . nnes. Soil and Water Contamination by Heavy Metals, in Soil Proce.J.Je.J ano
Water QuaLity, R Lal and B.A. Stewart, Eds., Advances in Soil Science Series, Lewis Publish-
ers, Boca Raton, FL, 1994, pp. 233-271.
Sposito, G., L.J. Lund, and A.C. Chang. Trace metal chemistry in arid-zone field soils amended
with sewage sludge: 1. Fractionation of Ni, Cu, Zn, Cd, and Pb in solid phases. SoiL Sci. Soc.
Am. J., 46, pp. 260-264, 1982.
Sterrit, R.M. and J.N. Lester. Comparison of methods for the determination of conditional
stability constants of heavy metal-fulvic acid complexes. Water R.J., 18, pp. 1149-1153, 1984.
Stevenson, F.J. and Y. Chen. Stability constants of copper(II)-humate complexes determined by
modified potentiometric titration. SoiL Sci. Soc. Am. J., 55, pp. 1586-1591, 1991.
Metal Retention and Mobility as Influenced by Some Organic Residues 223
Stevenson, F,J., A. Fitch, and M.S. Brar. Stability constants of Cu(II)-humate complexes:
Comparison of select models. SoiL Sci., 155, pp. 77-91, 1993.
Tiller, K.G. Heavy metals in soils and their environmental significance. Adv. SoiL Sci., 9, pp. 113-
142, 1989.
Ure, A.M., Ph. Quevauviller, H. Muntau, and B. Griepink. Speciation of heavy metals in soils
and sediments. An account of the improvement and harmonization of extraction techniques
undertaken under the auspices of the BCR of the Commission of the European Communities.
Int. J. Environ. AnaL. Chern., 51 (Special Issue), pp. 135-151, 1993.
WenzeL W.W., M.A. Pollak, and W.E.H. Blum. Dynamics of heavy metals in soils of a reed bed
system. Int. J. Environ. AnaL. Chern., 46, pp. 41-52, 1992.
Wilson, S.A., T.C. Huth, R.E. Arndt, and R.K. Skogerboe. Voltammetric methods for determi-
nation of metal binding by fulvic acid. AnaL. Chern., 52, pp. 1515-1518, 1980.
Xue, H.B., D. Kistler, and L. Sigg. Competition of copper and zinc for strong ligands in a
eutrophic lake. LirnnoL. Oceano gr., 40, pp. 1142-1152, 1995.
Zunino, H., G. Galindo, P. Peirano, and M. Aguilera. Use of the resin exchange method for the
determination of stability constants of metal-soil organic matter complexes. SoiL Sci., 114, pp.
229-233, 1972a.
Zunino, H., P. Peirano, M. Aguilera, and 1. Escobar. Determination of maximum complexing
ability of water-soluble complexants. Soil Sci., 114, pp. 414-416, 1972b.
CHAP"I'ER 11
The Rhizosphere and Trace Element
Acquisition in Soils
George R. Gobran, Stephen Clegg, and Francois Courchesne
INTRODUCTION
Soil chemical analyses indicate the potential nutrient availability under favorable nu-
trient conditions for root growth and activity (Marschner, 1986). Conventional soil tests
give no information about root induced changes in the rhizosphere due to exudation.
Such processes may ameliorate toxic ions such as Al (Gobran and Clegg, 1996; Inskeep
and Comfort, 1986). Therefore, the use of conventional soil and soil solution tests in both
forestry and agriculture can be unsatisfactory for the prediction of plant responses to
fertilization, acidification, or to other external stresses (Mahendrappa et al., 1986;
Marschner, 1986; Hogberg and Jensen, 1994).
Beier and Cummins (1993) indicated "difficult-to-study" areas in ecosystem manipu-
lation experiments that require more attention in order to understand relationships be-
tween biological and chemical parameters. These areas were roots, soil, in-plot variability,
soil microbiology, and ecophysiology. Studies of the soil, rhizosphere, and roots in ma-
nipulation experiments can directly address these areas, but have hitherto been poorly
investigated in forests (Vogt et al., 1993).
The scope of this chapter is to demonstrate the importance of the rhizosphere in the
soil-plant system and to emphasize its role on the cycling of both macro and trace ele-
ments. T iterature will first be reviewed to characterize the rhizosphere and to con-
trast prope . es of the bulk and rhizosphere soil environments. A conceptual model
for nutrient availa. ility in the mineral soil-root system will subsequently be described.
The model on dynamic feedback processes between plant roots and the sur-
rounding soil materials to illustrate the soil response to environmental stresses. The
results from a series of case studies encompassing a range of field observations from
hydrologically and chemically manipulated forest soils also will be presented. Finally,
the implications of the available data on the rhizosphere dynamics and of the concep-

226 Fate and Transport of Heavy Metals in the Vadose Zone
tual model on our understanding of trace element cycling in the soil-plant system will
be assessed.
History
One of the most notable milestones in rhizosphere research was the isolation and
identification of the nitrogen fixing bacteria (genus Rhizobium) in leguminous plants
(Beijerink, 1888; Hellriegel and Wilfarth, 1888). These studies attracted the attention of
L. Hiltner, who later coined the term, rhizosphere (Hiltner, 1904) due to his opinion that
this zone was unique for soil organisms. Other substantial contributions to rhizosphere
research stem from the work of Bowen (1961), Roriva and Bowen (1966), and Roriva
(1969), who increased understanding of the role and nature of root exudates on the
rhizosphere effect. We now see that the rhizosphere has become an area of intense inter-
est to soil scientists, ecologists, and agronomists.
izosphere-Definitions
The term stems from two Greek words rhizo for root and dphere (sphaira), the environ-
ment in which one acts or exists. Although the term rhizosphere seems self-explanatory,
many conceptual and operational definitions exist. Curl and Truelove (1986) described
the rhizosphere as "that narrow zone of soil subject to the influence of living roots, as
manifested by the leakage or exudation of substances that affect microbial activity."
Alternatively, Lynch (1990) stated that "the total rhizosphere environment is determined
by an interacting trinity of the soil, the plant and the organisms associated with the
roots." According to the Glossary of Soil Science Terms (SSSA, 1997), the rhizosphere
is "the zone of soil immediately adjacent to plant roots in which the kinds, numbers, or
activities of microorganisms differ from that of the bulk soil." The term rhizosphere has
also been macroscopically referred to as that portion of the soil profile where most roots
are located (Hedley et al., 1982; Mengel et aI., 1990). Ulrich (1987) stated that "Mor-
phologically, roots and soil, or micro-organisms and soil, usually could be clearly sepa-
rated," but "From a functional point of view this clear boundary does not exist." However,
this has not stopped the formulation of a number of conceptual and operational subdivi-
sions of the rhizosphere based on distance from the root or on the inclusion of root
materials. This is primarily due to the recognition that the root surface is a critical site for
soil-plant-microbe interactions. This surface has been called the rhizoplane (Clark, 1949;
Richards, 1987; Paul and Clark, 1989), or the soil-root interface (Gobran and Clegg,
1996), while the term rhizosphere is more commonly used to describe a region domi-
nated by soil material. The root-free material surrounding the rhizosphere soil is termed
the bulk soil although the extension of the rhizosphere and its boundary with the bulk
soil are difficult to define precisely. Also, the spatial distribution of the rhizosphere soil
changes with time as roots die or colonize new areas of the soil profile.
Methods of Rhizospheric Study
A wide range of methods was used to study the rhizosphere in the laboratory and in
the field. Plant growth experiments in the greenhouse have been extensively employed
(Spyridakis et al., 1967; Berthelin and Leyval, 1982; Kirlew and Bouldin, 1987; Mengel
The Rhizosphere and Trace Element Acquisition in
et aI., 1990; Chung and Zasoski, 1994). Some researchers used homogenized bulk soil as
the growth medium, while others added a reference mineral like biotite, muscovite, or
phlogopite to the soil materials to study the rhizosphere effect on mineral weathering
(Boyle and Vogt, 1973; Mojallali and Weed, 1978). The more elaborate methods de-
mand reconstruction of the soil profile with the subsequent changes to biological and
chemical conditions. Examples include porous plastic envelopes where roots have no
physical contact with the soil but where nutrients and water pass through the pores
(Brown and UI-Haq, 1984). Soil columns were used where homogenized soil materials
were packed in a container and divided by stainless steel screens to create a series of soil-
root zones (Helal and Sauerbeck, 1983; Dormaar, 1988; Mengel et aI., 1990). High root
densities were achieved in the central zone (under the growing plant), while adjacent
zones were root-free but subjected to the influence of the rhizosphere. Hinsinger et al.
(1992) adapted the Kuchenbuch and Jungk (1982) method to produce a two-dimen-
sional root mat on a polyamide net to simulate a macroscopic root surface. The mat was
then laid down on an agar-mica substrate. This method has enabled detailed examina-
tion of the rhizospheric substrate following sectioning of the agar-mica with a micro-
tome (Hinsinger and Jaillard, 1993; Hinsinger et al., 1993).
Comparison of the mineralogy of bulk and rhizosphere soil materials was also achieved
using field samples collected from agricultural (Kodama et aI., 1994) and forested sites
(April and Keller, 1990; Gobran and Clegg, 1996; Courchesne and Gobran, 1997). The
separation of the rhizosphere and bulk fractions involved drying, gentle shaking in a
plastic container (Hendriks and Junk, 1981; Haussling and Marschner, 1989; Kirlew
and Bouldin, 1987), and brushing of roots to free adhering rhizosphere soil (Haussling
and Marschner, 1989; Clemensson-Lindell and Persson, 1992). The soil particles di-
rectly contacting the root surfaces were consider as rhizoplane soil (April and Keller,
1990). Another useful method that has often been overlooked is the rhizocylinder method
(Riley and Barber, 1969, 1971; Hoffmann and Barber, 1971) in which plant root plus
adjacent soil particles were examined as a separate fraction (the rhizocylinder) and com-
pared with rhizosphere and bulk soil samples (Gobran and Clegg 1996). Kosola (1996)
used pressurized-wall minirhizotron tubes equipped with a borescope and a laparoscopic
sampler for collecting roots of known age in the field. This new technique could help
refine the measurement of the effect of specific living roots on the surrounding soil.
The detailed identification and quantification of chemical, physical, and mineralogi-
cal changes occurring in the rhizosphere not only relied on macroscopic approaches
(chemical extraction, titration, ion exchange, X-ray diffraction) but also involved the
observation of thin sections using microscopic techniques such as transmission electron
microscopy (TEM), scanning electron microscopy (SEM) , SEM with energy disper-
sive spectrometry (EDS), and SEM in the backscattered electron mode (BESI) (April
and Keller, 1990; Kodama et aI., 1994; Bruand et aI., 1996). Conkling and Blanchar
(1989) and Conkling et al. (1991) constructed glass microelectrodes with sensing tips
0.02 mm in diameter and 0.05 mm long for in situ pH measurements. The technique was
successfully used in conjunction with minirhizotrons to measure the rhizosphere pH 'of
alfalfa, corn, and soybean. Future developments in the area of microelectrode technol-
ogy will facilitate the direct and continuous measurement of chemical changes in the
rhizospheric environment.
228 Fate and Transport of Heavy Metals in tile Vadose Zone
RHIZODEPOSITION
Root Distribution and Longevity
Despite a long history of study, our understanding of root distribution and processes
is poor (Jackson et al., 1996; Vogt et al., 1993). Yet, together with litterfall, root produc-
tion provides the greatest input of organic carbon to many soils, which can store carbon
at twice the rate of the above-ground biomass (Waring and Schlesinger, 1985). Fine
roots with their high production of exudates and variable life span contribute the major
portion of the total carbon input (Grayston et al., 1996). Root distribution patterns vary
greatly with depth across the major biomes of the world. For example, Jackson et al.
(1996) found that tundra, boreal forest, and temperate grassland have the shallowest
rooting profiles with 80-90% of roots in the top 30 cm of the soil, whereas deserts and
temperate coniferous forests had deep profiles with -50% of roots in the top 30 cm of the
soil. Moreover, many plant species may have a wide root distribution which can cause
difficulties in identifying which zones roots acquire most of their nutrients. For example,
it was shown that spring wheat during the later periods of the growing season took 30-
40% of total P from the subsoil despite a higher P content in the topsoil and a reasonably
uniform rainfall during the growing season (Fleige et al., 1981). Fox and Lipps (1961)
found that 3% of the total root mass in alfalfa took up 60% of total nutrients from the
subsoil during periods of drought.
On a global scale, the average root distribution has been estimated as being 30%,
50%, and 74% in the top 10 cm, 20 cm, and 40 cm, respectively (Jackson et al., 1996).
Yet in many forest ecosystem studies there has been a tendency to emphasize the role of
the upper organic horizons as a major source of nutrients. Despite this, 40 to 50% of the
fine root biomass is found in the first 30 cm of mineral soil at many forest sites
(Clemensson-Lindell and Persson, 1992; Haussling and Marschner, 1989; Persson et al.,
1995; Wood et al., 1984). This root distribution is possibly related to the large pool ofN,
P, and S retained in the mineral horizons, especially in spodosols (Stevenson, 1991).
Clearly, more investigation is needed on the role of roots in the mineral soil and the
processes by which they acquire elements and alter their availability.
Minirhizotron studies of Norway spruce have shown that roots can live for nine months
or more (Majdi, 1994) and frequently reoccupy old root channels (H. Persson, pers.
comm.). Fahey and Hughes (1994) estimated the median fine root (d-mm) longevity
(50% survivorship) in the forest floor under maple and beech at Hubbard Brook, New
Hampshire, to be about 6 months. In a maple-beech forest of south-central New York,
59% of the spring root cohort were still alive after 5 months.
Moreover, Hendrick and Pregitzer (1993) working in northern hardwood forests in-
dicated that roots born in spring live longer than the average fine root. Persson (1983)
suggested that roots growing deeper in the soil profile live longer than roots in surface
horizons. The seasonal decline in total root biomass is less pronounced in mineral than in
organic horizons although rapid root disappearance can occur in both (Hendrick and
Pregitzer, 1992). This indicates that the rhizosphere may not always be an ephemeral
environment in the soil. The apparent longevity of roots in forests would allow more
time for the establishment of rhizospheric processes than is normally considered in ex-
perimental and agricultural systems. The reestablishment of rhizospheric conditions in
unoccupied root channels may also be enhanced due to the priming effects of dead roots.
The Rhizosphere and Trace Element Acquisition in Soils 229
Such effects may include physicochemical properties such as high porosity and readily
decomposed organic matter as well as biological priming with the propagules of
rhizospheric organisms.
Belowground Carbon Flux
Although the rhizosphere constitutes only 1-3% of the soil volume (Coleman et al.,
1978; Gobran and Clegg, 1996), a large proportion of plant-assimilated carbon is re-
leased in the belowground system as dead roots, exudates, and as substrates for mycor-
rhizae. It is also interesting and important to compare this belowground flux to that
aboveground. Studies indicate that loss of fine roots and mycorrhizae returns two to five
times more organic matter to the soil than the aerial biomass (Fogel and Hunt, 1983;
Waring and Schlesinger, 1985). The magnitude of C fluxes to and from soils vary with
latitude. For example, C cycling in litterfall increases by two- to threefold from 65 to
45N (Van Cleve and Powers, 1995). The annual C input derived from the activity of
root and mycorrhizae and from root decay (estimated as the difference between CO
2
fluxes from the soil surface -litterfall) also increases with decreasing latitude but more
rapidly than litterfall fluxes. At Hubbard Brook the aboveground and belowground C
pools were 9900 and 2300 kmol C ha -I y -I, respectively (Johnson et al., 1995). As for the
C fluxes, 124 kmol C ha-
I
y-I was deposited to the soil as litterfall, while 60 kmol C ha-
I
y-I came from root decomposition. Trettin et al. (1995) compiled C flux values from a
range of northern forested wetlands. The aboveground biomass ranged from 15 to 55 ton
C ha-
I
y-I, while belowground C pools totaled 50 to 1300 ton C ha-
I
y-I, of which 7 to 21
ton C ha-
I
y-I came from root biomass. Above- and belowground C fluxes were esti-
mated at 0.4 to 1.6 and 0.1 to 0.4 ton C ha-
I
y-I. Rhizodeposition varies due to other
factors such as climate, the presence of symbiotic organisms and plant age and type. For
instance, annual plants release less carbon than perennials (Grayston et al., 1996).
Exudates in the Rhizosphere
Apart from the cellulose, lignin, and other compounds released to the soil by dead
roots, there are a wide variety of compounds released by live roots which are collectively
called exudates (Tables 11.1 and 11.2). Exudates and their diversity have been the sub-
ject of many reviews (e.g., Marschner, 1986; Uren and Reisenauer, 1988; Bowen and
Roriva, 1991; Grayston et al., 1996).
Organic acids are believed to be quantitatively the most important component of plant
exudates (Table 11.3). They can serve as carbon nutrient sources for the microbial popu-
lation and playa major role in weathering and complexation of micro- and macronutri-
ents. Organic acids of low molecular weight are ubiquitous in soils, yet the type and
quantity vary not only in the bulk and rhizosphere soil (Fox and Comerford, 1990;
Grierson, 1992; Szmigielska et al., 1996), but also within the biosphere compartments
such as plant canopy, forest litter, surface horizons, soil solutions, rhizosphere, rock
surfaces, etc. (Drever and Vance, 1994).
Acid-Base Changes in the Rhizosphere
This has major direct and indirect consequences for the availability of nutrients and
toxic elements and their uptake by roots. Five important factors affecting acid-base condi-
230 Fate and Transport of Heavy Metals in tile Vadose ZOlle
Table 11.1. Root Products Released in the Rhizosphere
a
Product
Root exudates
Diffusates
Excretions
Secretions
Root debris
Release
Leakage from or between
epidermal cells
Active excretion
Active secretion
Cell or root death,
sloughing
Compound
Sugars, inorganic acids, amino acids,
water inorganic ions, oxygen, etc.
Carbon dioxide, bicarbonate, protons,
electrons, ethylene, etc.
Mucilage, protons, electrons, enzymes,
siderophores, alleopathic compounds,
etc.
Root-cap cells, cell contents, etc.
a After Uren and Reisenauer, 1988.
Table 11.2. Examples of Root Products Exuded to the Rhizosphere
a
Some Organic Compounds Exuded by Roots
Carbohydrates Arabinose, fructose, glucose, maltose, ribose, sucrose
Amino acids All 20 amino acids
Aliphatic acids Acetic, citric, fumaric, malic, oxalic, tartariC, valeric
Aromatic acids p-Hydroxybenzoic, p-coumaric, gallic, salycyclic
Fatty acids Lineolic, palmitic, stearic
Sterols Campesterol, cholesterol, sitosterol
Enzymes Amylase, deoxyribnuclease, peroxidase, phosphatase
Miscellaneous Plant and microbial growth regulators, stimulators and inhibitors
a After Uren and Reisenauer, 1988.
Table 11.3. Examples of Organic Acids and Other Complexing Compounds Found in the
Rhizosphere
a
Compound Occurrence
Citric, tartaric, lactic,
and malic acids
Oxalic acid
Siderophores
Phenolic acids
2-Ketogluconic acid
a After Stevenson, 1991.
Produced by roots and bacteria in the rhizosphere. Present
in litter extracts and canopy throughfall.
Produced by fungi, including mycorrhizae.
Abundant in acid soils.
Produced by rhizosphere and ecomycorrhyzal fungi.
Produced under conditions of Fe stress.
Formed through the decay of Iignins. Involved in the
mobilization and transport of Fe in acid soils.
Synthesized by bacteria on rock surfaces and in the rhizosphere.
Abundant in habitats rich in decaying organic matter.
tions in the rhizosphere will be discussed here: (1) the production of CO
2
from respira-
tion, (2) the excretion of organic acids, (3) microbial production of acids following as-
similation of released root carbon, (4) ion uptake, and (5) plant genotype.
The Rhizosphere and Trace Element Acquisition in 231
The reviews ofNye (1986) and Marschner and Romheld (1996) do not emphasize the
role of the first three factors. For instance, the differences in air pressure between the
root surface and the bulk soil in normally aerated soils will rapidly diffuse away CO
2
with a negligible effect on pH (Nye, 1986). It has been suggested that low molecular
weight acids released to the rhizosphere by roots and microorganisms may have an acidi-
fYing effect (Mench and Martin, 1991; Petersen and Bottger, 1991). However, the quan-
tities required to have a significant effect in the rhizosphere have not been found, either
because they are produced in too small amounts by roots or because they are rapidly
metabolized by microorganisms (Nye, 1986). Also, the quantity of acidity produced by
microbes which use carbon deposited in the rhizosphere is probably small, due to the
vast quantity of carbon that would be required. Calculations suggest that one-third of
the total carbon released by roots would have to be converted to acids by microbes to
produce a significant change in pH (Nye, 1986). Finally, Hedley et al. (1982) found that
neither the total number of microbial colonies nor the number of acid-producing colo-
nies were related to rhizosphere pH. The consensus is that the dominant mechanism
responsible for pH changes in the rhizosphere is the net release of H+, HC0
3
-, or OH- in
response to the imbalance between cation and anion uptake by roots (Tinker, 1990).
The most evident factors affecting acid-base conditions of the rhizosphere are ion
uptake and plant genotype. For example, when NH4 + rather than N0
3
- ions were sup-
plied to soils (Riley and Barber, 1971; Soon and Miller, 1977; Rollwagen and Zasoski,
1988), a drop in pH of two units was reported in the soil close to root surface. In acid
soils, such as spodosols, where the rate of nitrification is very low, the form of nitrogen
taken up will mostly be NH/. In this situation cation uptake, will exceed anion uptake
resulting in net H+ excretion and in a pH decrease in the rhizosphere relative to the bulk
soil. Additionally, plants associated with N-fixing organisms (e.g., legumes, ALnlM, and
CaJuarina) also acidify the rhizosphere since the uncharged dinitrogen molecule crosses
the soil-nodule or soil-root, resulting in a higher uptake of cations than anions. In soils
where N0
3
- is the primary N-form, the amounts of anions taken up by plants tends to
exceed cations, thus plants are required to release HC0
3
- or OH- to maintain electrical
neutrality across the soil-root interface. This causes increases in rhizosphere pH com-
pared to the bulk soil. This phenomenon is so well-established that a method to manipu-
late rhizosphere pH has been elaborated by using different N sources (Riley and Barber,
1971; Sarkar and Wyn Jones, 1982).
Given the same soil and form of nitrogen supply, large differences can arise between
differing species or cultivars to acidify their rhizosphere (Marschner and Romheld, 1996).
For example, it has been shown that lime-induced chlorosis (iron deficiency) in different
cultivars of soybean, maize, and peanut supplied with N0
3
- were related to rhizosphere
pH between genotypes. Two cultivars of soybean supplied with N0
3
- in a soil with a pH
of 6.0 showed basal and apical rhizosphere pH values of 6.8 and 5.8 in the iron inefficient
cultivar and 5.6 and 5.3 in the iron efficient cultivar. However, no difference was observed
between three corn hybrids suggesting that they do not differ with respect to the mecha-
nism controlling rhizosphere pH (Kirlew and Bouldin, 1987). Rollwagen and Zasoski
(1988) also observed significant differences in rhizosphere pH values between Douglas
fir, Sitka spru<;e, and western hemlock seedlings planted in the same soil. The ecological
relevance of these differences can be assumed to apply to other plants, genotypes, and
nutrients such as Zn and Mn (Macar and Graham, 1987) but it has been little investigated.
232 Fate and Transport of Heavy Metals in the Vadose Zone
Changing the pH of the rhizosphere has direct implications for forest nutrition and
crop management. For instance, a decreasing pH will favor the dissolution of metals and
their uptake by plants. Sarkar and Wyn Jones (1982) showed that the Fe, Zn, and Mn
content of the shoot and roots of dwarf French beans was inversely proportional to rhizo-
sphere pH. The dissolution of Al solids and KCI-extractable Allevels increased as excess
H + ions were released by the roots of corn seedlings (Kirlew and Bouldin, 1987). Rollwagen
and Zasoski (1988) reported a dramatic increase in foliar Mn content of conifers as their
rhizosphere pH was decreased by ammonium addition to the soil. Cation exchange equi-
libria will also be affected by the pH gradient existing between the root surface and the
bulk soil. Chung and Zasoski (1994) demonstrated that the selectivity for NH4 over Ca
increased as pH decreased but no effect was observed for NH4-K exchange.
RHIZOSPHERIC FEEDBACK lOOPS
Regulating Processes
In spite of the small volume that the rhizosphere occupies in the mineral soil (Gobran
and Clegg, 1996), it plays a central role in the maintenance of the soil-plant system (Clegg
and Gobran, 1997; Clegg et aI., 1997) and influencing the biogeochemistry of forest
ecosystems (Gobran et aI., 1998). These root effects on soils suggest to some investiga-
tors that soil can be considered, in part, as a product of plants and soil biota (Van Breemen,
1993). It has been suggested that interactions between roots, microbial communities,
and the soil under forest conditions are characterized by feedback loops driven by pho-
tosynthate released by roots (Perry et aI., 1989; DeAngelis et aI., 1986; Hobbie, 1992).
Under such circumstances, the rhizospheric community "continually pull themselves up
by their own bootstraps" (Perry et al., 1989) so that nutrient cycling and availability in
the rhizosphere is higher than in the bulk soil, thus buffering the ecosystem against
disturbances. An indication of this mechanism is reflected by the ratio of organisms in
the rhizosphere soil to counts in bulk soil (R/S), which were found to range from 10 to 50
under varying plant species at different stages of development and in different soils and
climates (Paul and Clark, 1989). Moreover, the R/S is typically 10 to 50 for bacteria and
5 to 10 for fungi (Richards, 1987).
Certain tree species actively control nutrient availability by complex feedback loops
(Perry et aI., 1989) between trees, microbial communities, and the soil to maintain a
competitive advantage. This could be achieved, for example, by decreasing the availabil-
ity of essential nutrients such as P in the bulk soil by acidification (Cole, 1995; Van
Breemen, 1993) or by decreasing mineralization rates. This may be achieved by produc-
tion of secondary metabolites which are leached from foliage and litter (Van Breemen,
1995) while simultaneously favoring their own root systems by the establishment of nu-
trient-rich rhizospheric conditions and mycorrhizal association. It is apparent that accu-
mulation of nutrients in the rhizosphere is a natural result of these many feedback
processes occurring under forest conditions.
Due to these regulating processes, the rhizospheric nutrient supply and demand are
more carefully balanced in nutrient-poor forests than under agricultural and short-term
experimental conditions. It is generally believed that in infertile soils (e.g., spodosols),
the roots and microorganisms render the soilless favorable for plant growth (Chapin,
1980; 1993). Although this observation holds when the bulk soil is considered, it is not
The Rhizosphere and Trace Element Acquisition in Soils 233
supported by examination of the rhizospheric chemical properties (Gob ran and Clegg,
1996). At this scale of investigation, it seems that roots and microorganisms do increase
the capacity of soils to support plant in nutrient-poor environments.
Since most mineral nutrients must pass through the rhizosphere before assimilation by
plants, we consider that this small zone has great potential as a regulator of plant nutrient
availability and flux. Yet, in order to have any significant effect on nutrient cycling in
forest ecosystems the rhizosphere must exhibit two features: persistence in time and resil-
ience against perturbation, which are known properties of ecosystem stability (Richards,
1987). We believe the rhizosphere has such features and that they are reflected in the
consistent difference observed between rhizosphere and bulk chemistry and mineralogy
which remain after a wide range of field treatments (e.g., drought, irrigation, and ammo-
nium sulfate) (Clegg and Gobran, 1997; Clegg et al., 1997). In addition to long-term
rhizospheric effects, trees adapted to nutrient-poor sites have low nutrient absorption
rates as well as efficient internal cycling (Van Breemen, 1995). This strategy of trees
under such conditions contrasts with rapidly growing agricultural and ruderal species
growing on intensively managed and fertile soils where mass flow may equal that of diffu-
sion in supplying certain mobile nutrients (Binkley, 1986). Moreover, a relatively large
and long-lived root biomass may benefit from the accumulation of organic matter ob-
served in the rhizosphere (Gobran and Clegg, 1996; Chung and Zasoski, 1994) which
could act as both a source and sink of available nutrients and potentially toxic ions.
Element Supply and Mobility in the Rhizosphere
Three major mechanisms that may be connected with the supply and mobility of ele-
ments from immobile phases in the rhizosphere are (1) disturbance of equilibria be-
tween the solid phases of the soil and the soil solution; for example, under acid conditions
there is a relative abundance of Fe, Mn, Zn, Bo, Mo, and Cu, which could be present in
sufficiently high concentrations in solution to be toxic to common plants (Brady, 1990);
(2) dissolution of sparingly soluble minerals containing Fe and P by the release of hydro-
gen ions (Kochian, 1991); and (3) the proliferation of organic acids in the rhizosphere
could play two roles by either raising or lowering ion mobility. They can, for example, be
synthesized and release micronutrients from the rhizosphere soil or act as chelating agents,
thus reducing the mobility and activity of ions such as Al, Fe, Mn, Cu, and Zn. The type
of biochemical chelator varies with their mobility and life span in the soil. Approximate
concentrations of organic compounds and acids in soil solutions are found in Table liA.
The literature shows that organic acids released to the rhizosphere by roots and my-
corrhizae can play an important part in trace element nutrition to plants (trees/forests).
Thermodynamic calculations from experiments with hypothetical rhizosphere solutions
from PinlM radiata and Hordeum vulgare containing amino and low molecular weight acids
(e.g., oxalic and acetic acids) showed that exudates make a significant contribution to
the total soluble Fe (III), Cu (II), and Zn (II) (Inskeep and Comfort, 1986). Field stud-
ies also give a similar picture. In a study with and without the presence of ectomycorrhizal
mats in Douglas fir (PJeudotJuga menziuil), it was shown that P, S, H, Al, Fe, Cu, Mn,
and Zn were in significantly higher concentrations in the soil solution in the presence
than in the absence of ectomycorrhizal mats. It was concluded that dissolved organic
carbon and particularly oxalate released to the rhizosphere could provide a local weath-
ering environment which raises the availability of P, S, and trace elements.
234 Fate and Transport of Heavy Metals in the Vadose Zone
Table 11.4. Approximate Concentrations of Some Or-
ganic Acids and Compounds in Soil Solutions
a
Compound Concentration
Simple organic acids
Amino acids
Phenolic acids
Siderophores
a After Stevenson, 1991.
1 X 10-
3
to 4 X 10-
3
M
8 X 10-
5
to 6 X 10-
4
M
5 X 10-
5
to 3 X 10-
4
M
1 X 10-
8
to 1 X 10-
7
M
Since the oxidized states of Fe, Mn, and Cu are generally less soluble than the re-
duced states (Brady, 1990), microorganisms in the rhizosphere can alter the availability
or toxicity to plants of Fe, Mn, and Cu by changing their oxidation reduction states and
by mineralization of soil organic matter. Recent work by Bartlett (1996) stated that the
rhizosphere is a poised redox system that can be either a reducing or oxidizing milieu for
Mn. This is important since Mn oxides affect the oxidation mobility and toxicity of trace
elements such as Cr (II), Pu (III), and Co (II). Moreover, dissolved Mn (II) in the
presence of Fe influences Fe crystallization, hence affecting genesis of Fe oxides. Man-
ganese oxides also promote decomposition of organics, thus affecting C and N transfor-
mations (Huang, 1991). Therefore, these processes in the rhizosphere deserve further
attention for the study of soils and environmental quality.
There is a concern about high weathering and mineralization releasing toxic elements
to the rhizosphere/root. However, it appears that organic acids produced in the rhizo-
sphere protect plants from the toxic effects of AI and other metals. This is due to effec-
tive complexation of the free metal species by organic matter, even though total metal
concentrations in the rhizosphere may be higher than in the bulk soil (Gobran and Clegg,
1996). Additionally, high mineralization rates in the rhizosphere may have a positive
effect on degrading toxic organic (Boyle and Shann, 1995; Pidgeon et aI., 1996; Reilley
et aI., 1996). Indeed, the use of rhizosphere systems in site bioremediation is attractive
and a great attention should be focused on the development of plant root systems for
better understanding site remediation technology (Skladany and Metting, 1993).
Microbial Activity and Element Accumulation in the Rhizosphere
Mass flow and diffusion are assumed to be the most important ion transport processes
in supplying roots with nutrients (Nye and Tinker, 1977; Barber, 1984), although the
magnitude of these pathways can vary greatly from one system to another as indicated in
Binkley (1986). Moreover, the magnitudes of these pathways are difficult to quantifY
and there may be considerable nutrient and water transport by biological agents such as
mycorrhizae (Richards, 1987; Finlay and S6derst6m, 1989).
The roots of most soil-grown plants are usually mycorrhizal (Grayston et al., 1996)
and the role of mycorrhizal symbiosis in nutrient uptake has been well documented.
There are several groups of mycorrhizae; those most commonly associated with trees
and shrubs include vesicular-arbuscular mycorrhizae (V AM), ectomycorrhizae (ECM),
and ericoid mycorrhizae (EM). For example, Marschner and Dell (1994) implicated the
three mycorrhizal groups with the ability to take up and deliver nutrients to plant roots
The Rhizosphere and Trace Element Acquisition in 235
(e.g., P, NH
4
, N0
3
, K, Ca, S04' Cu, Zn, and Fe). Depending on tree species, a variable
proportion of the root supply of mineral nutrients pass through the fungal hyphae. For
example, in ECM of plants such as Norway spruce, more than 90% of the root apical
zones are enclosed by a fungal sheath. The interactions between soil acidification, plant
growth, and nutrient uptake in ectomycorrhizal associations of forest trees has recently
been reviewed by Finlay (1995). In this review it was proposed that the ECM mediated
the effects of acidification on forest trees through modification of the soil chemical envi-
ronment, altering patterns of plant nutrient uptake and increasing tolerance to, or detoxi-
fying, the increased levels of heavy metals or aluminum often associated with soil
acidification. Moreover, ECM greatly extend the life span of absorptive roots beyond
the few days that root hairs are active; ECM persist for 6-9 months and even up to 13
years (Fogel, 1983).
Mycorrhizae can also chemically modify the rhizosphere environment by direct and
indirect means to facilitate P uptake. Ectomycorrhizal fungi release large quantities of
oxalic acid which can mobilize P from sparingly soluble Ca phosphates in calcareous
soils and AI and Fe phosphates in acid soils (Marschner and Dell, 1994). It is also pos-
sible that mycorrhizae in the rhizosphere have a large quantity of exchange sites with a
higher affinity for phosphate and other ions (Richards, 1987; Bolan, 1991; Jakobsen et
aI., 1992). These exchange sites may act both as source and sink for nutrients, thus
easing seasonal variations in nutrient supply (Grayston et aI., 1996), and buffer
aboveground parts from the effects of toxic metals, such as AI, Cu, and Zn (Wilkins,
1991; Marschner and Dell, 1994). Mycorrhizae also have several biochemical mecha-
nisms to access and supply nutrients through the production of enzymes. For example,
the excretion of phosphatases may play an important part in mineralizing the large or-
ganic P pool in the rhizosphere (Clegg and Gobran, 1997; Marschner and Dell, 1994;
Haussling and Marschner, 1989).
CASE STUDIES
The case studies in this chapter are based on the most important results presented in
the following papers; Gobran and Clegg (1996), Clegg and Gobran (1997), Clegg et aI.
(1997), and Courchesne and Gobran (1997). Our studies were based on the view that
data from routine soil chemical analyses do not always correlate well to tree health,
nutrient uptake, and leaf chemical composition (Binkley, 1986; Mahendrappa et aI., 1986).
This may be due to routine soil sampling missing the rhizosphere fraction. Since the
rhizosphere soil is altered by many interdependent processes between roots and micro-
organisms, rhizospheric processes have a large potential to change soil fertility and nu-
trient cycling, particularly in undisturbed forest soils.
Additionally, much of the current knowledge about the rhizosphere is based upon
research using agricultural species (often seedlings) grown under controlled conditions
on homogeneous nutrient-rich substrates (Chapin, 1980; Parmelee et al., 1993). This has
enabled description and modeling of ion transport to roots to be based on solid theory
and experiment (Nye and Tinker, 1977; Barber, 1984). However, linking model predic-
tions to nutrient uptake under field conditions may not be straightforward (Wild, 1989;
Hogberg and Jensen, 1994).
236 Fate and Transport of Heavy Metals in the Vadose Zone
The Conceptual Model
Due to the close linkage between plants, microbes, and soil, the study of rhizospheric
processes will be a key to resolving problems related to understanding and predicting
soil/plant interactions in a changing environment. Therefore, we proposed a conceptual
model for nutrient availability in the mineral soil-root system (Cobran and Clegg, 1996).
Our hypothesis was that fine roots and their associated organisms maintain a higher level
of nutrient availability in the rhizosphere (Rhizo) than in the bulk soil (Bulk) (Figure
11.1). This was accomplished by the release, transport, and accumulation of reactive soil
organic matter and inorganic compounds in the soil-root interface (SRI) and rhizosphere.
The interaction between soil, microorganisms, and roots creates a mutually supportive
system that can raise nutrient availability by increasing moisture content, mineralization,
and enriching the pool of cations and anions through increased exchange sites.
We envisioned the soil fractions as a multiple-phase system comprised of a gas phase,
a solution phase, and a surface phase. The surface phase of the bulk soil with the largest
volume of the soil body is represented by a large box, yet has a lower charge per unit
mass, thus a lower cation exchange capacity (CEC) than the rhizosphere and SRI. In
contrast, the rhizosphere and SRI represent a smaller fraction of the soil body but have
a larger CEC due to higher organic matter (OM), clay mineral, and amorphous oxide
content (CM). The organic matter component probably differs from the bulk soil since it
may contain of a higher proportion of easily mineralized and reactive root material, exu-
dates, mycorrhizae, and other associated microorganisms. We hypothesized that organic
matter in the rhizosphere and SRI is the most dynamic part of the system since it acts as
both a source and sink for elements, is involved in weathering, and fuels biological reac-
tions. Accordingly, CEC follows the same trend as organic matter content, increasing
from the bulk soil to the SRI (Figure 11.1). The arrows represent the three major trans-
port mechanism between the soil fractions and phases; mass flow, diffusion, and biologi-
cal transport by mycorrhizae. Our studies focused on mineralogical and chemical changes
in the solution and surface phases which were considered as products of both chemical
and biological interactions.
Field Site and Treatments
In a field investigation, soil samples were taken from a 30-yr Norway spruce [Puea
abie.J (L.) Karst.] stand situated at Skogaby in southwest Sweden (Cobran and Clegg,
1996). The soil was classified as a Haplic podzol (FAO-UNESCO, 1988) with a silty
loam texture throughout the profile. Manipulated field treatments control (C), ammo-
nium sulfate (NS), and irrigation (1) are described in Table 11.5. These treatments caused
rapid increases in growth and are discussed in more detail below.
Soil Fractionation
Separation of the soil fractions was conducted by carefully removing all roots by hand
from the field moist mineral soil, which was then passed through a 2-mm mesh to give
the bulk fraction (Bulk). The remaining fine roots 2-mm) and soil were gently shaken
to separate the soil aggregates (0.5 to 5 mm) from the roots to give the rhizosphere
fraction. This method is similar to that described by Hendriks and Junk (1981). We
The Rhizosphere and Trace Element Acquisition in Soils 237
Rhizosphere Bulk
O
2
(Rhlzo)
O
2
II(

GP
II(

GP
Gas
CO
2
Cft
phase
t+
(GP)
II( ca, Mi' K, AI, H
SOLP
II(
Ca, Mi' K
SOLP Solution

Ca, Mg, K, Na, AI, H Na,AI,H

ca Mg K NaAI Na Ca Mg NaAI Na
I Roots, OM and CM
CEC

caMgKNaAINa caMgKNaAIH

OM and CM Matter (OM)


Cia Minerals (eM
CEC CEC
SOLP)
Surface
phase
Figure 11.1. A conceptual model for nutrient availability in the mineral soil-root system.
Table 11.5. Manipulation Treatments from the Skogaby Site
a
Treatment
Control (e)
Irrigation (I)
Ammonium sulfate (NS)
Description
No treatment
A sprinkler system prevented water storage deficits greater than
20 mm during May to September.
100 kg N ha-
1
and 114 kg S ha-
1
were added as solid
ammonium sulfate annually.
a After Nilsson and Wiklund, 1992.
defined the soil-root interface fraction (SRI) as apparent free space within fine roots and
adhering rhizoplane soil 0.5 mm thick), which is similar to the rhizocylinder fraction
described by Riley and Barber (1969) and Hoffmann and Barber (1971). Before chemi-
cal extraction, the SRI was cut into 5-mm pieces to homogenize the samples and to
facilitate weighing.
Chemical Properties of the Soil Fractions
The three soil fractions generally differed greatly in chemical composition (Figure 11.2)
throughout the upper three mineral horizons. The pH (KCI) was lower in the rhizo-
sphere and SRI than in the bulk soil and tended to increase with depth; however, the differ-
ences were not significant in many cases. Titratable acidity (TA = AI + H, cmol
e
kg-I)
mE oBh _Bs
J
5.0 45,

4.5
=a 4.0
3.5
i
30
1
It
~ 15 b
3.0 o I
Bulk Rhlzo SRI Bulk Rhlzo SRI
45

2.01 b b
1.5 ; b I
- -
30
rn
ID
15
ID
1.0 j

~
c(
(.)

0.5
0 0.0
Bulk Rhlzo SRI Bulk Rhlzo SRI
Figure 11.2. The pH, titratable acidity (TA), base saturation (BS), and calcium-aluminum balance (CAB) in three soil fractions and horizons from the
control plots at Skogaby. At a given horizon, soil fractions with differing letters indicate significant differences at the 5% level. The absence of letters
indicates no significant difference between soil fractions.
L,.:
C!i
:::;
0..
-l
ci
:::l
V1
'"Cl
0
:l
0
....
:r.
It)
w
<::
'<
' ~
~
::.;
V'.
~
-
rp
. <
ru
Q.
0
V1
It)
N
C
:::;
ro
The Rhizosphere and Trace Element Acquisition in Soils 239
increased significantly in the order Bulk<Rhizo<SRI in all horizons. This acidity may be
explained by net H excretion from roots due to cation and ammonium uptake. When
comparing pH and T A we may conclude that pH is less sensitive in detecting chemical
changes in soils than TA or other variables (Gobran and Bosatta, 1988). If acidity alone
was considered, the rhizosphere would seem to be a harsher environment to roots than
the bulk soil since Al and H are considered to have negative effect on roots and plant
growth (Alva et al., 1986; Gobran et al., 1993). Yet, the increased acidity of the rhizo-
sphere was offset by an increase in exchangeable base cations. This can be seen by the
increase in base saturation (BS) which increased in the order Bulk<Rhizo<SRI.
Also, the calcium aluminum balance (CAB = -log(Caexch/Alexch) that reflects the de-
gree to which growth can be supported or hindered (Noble and Sumner, 1988; Gobran
et al., 1993; Cronan and Grigal, 1995) increased in the same order. Moreover, organic
matter was significantly associated with all variables in all fractions and horizons, and it
is known to playa decisive role in reducing the activities of toxic metals in the rhizo-
sphere through complexation (Stevenson, 1991). These observations have led us to con-
clude that rhizospheric chemical conditions may be more favorable for biological activity
and plant uptake than the bulk soil (Gobran and Clegg, 1996), which is in agreement
with the bootstrapping view of Perry et al. (1989). Despite good agreement between the
Conceptual Model and our soil chemical data, the application of the model to other
forest conditions and soil types awaits further investigation.
Weathering in Bulk and Rhizosphere Soil
We compared the mineralogy of the bulk and rhizosphere soil fractions of two profiles
collected from the untreated plots (Courchesne and Gobran, 1997). The working hy-
pothesis was that the mineral assemblage of the two fractions would differ, reflecting the
enhancement of mineral weathering in the immediate vicinity of roots in forest soils. In
each profile, samples were collected from the E (0-5 cm), the Bh (5-15 cm), and the Bs
(15-30 cm). The mineralogy of the clay-sized particles of both fractions was determined
by X-ray diffraction (XRD) of oriented specimens after removal of coatings, saturation
with Mg, Mg-ethylene glycol, or K and heating of the K-saturated specimens to 300 and
550C (Whittig and Allardice, 1986). The integrated intensity of each mineral (I) was
normalized relative to the intensity of the (100) peak (J = 0.426 nm) of quartz (I
Qz
) to
calculate a mineral intensity ratio (1/I
Qz
). Iron and Al were extracted with acid-ammo-
nium oxalate (Ala, Fea) and analyzed by atomic absorption spectrophotometry (AAS).
Oxalate dissolves amorphous organic and inorganic solid phases, most of which are of
pedogenic origin: they either accumulated as in situ weathering products or precipitated
from solution in podzolic soils.
Mineral abundance (l/I
Qz
) in the rhizosphere differed consistently from that in the
bulk soil (Figure 11.3). The magnitude of changes was controlled by the relative stability
of primary minerals in a weathering environment stimulated by root activity, and fol-
lowed the order: amphiboles > plagioclases > K-feldspars. The rhizosphere contained
significantly lower amounts of amphiboles (I < 0.10) relative to the bulk soil. The abun-
dance of plagioclase was also found to decrease in the rhizosphere for five of the six
horizons, but the decrease was not significant. However, XRD showed no rhizosphere
effect for K-feldspars. Our results also indicated that expandable phyllosilicates were
240 fate and Transport of Heavy Metals in the Vadose Zone
OBulk II Rhizo
2.5

2
a
1.5
~
S
1
0.5
a
b
0
Amphibole Int. Plagioclase K-Feldspar
Vermiculite
Figure 11.3. Average mineral composition of bulk and rhizosphere soil for the three horizons. Soil
fractions with differing letters indicate significant differences at the 10% level.
less abundant in the rhizosphere (I < 0.10) and thus less stable than the plagioclase and
K-feldspars. This observation is in agreement with Sarkar et al. (1979) but contrary to
Kodama et al. (1994).
Oxalate extractable Al and Fe were systematically higher in the rhizosphere than in
the bulk soil (Figure 11.4), a fact also noted by Sarkar et al. (1979) and Chung and
Zasoski (1994). These results support the XRD data by pointing toward an accelerated
degradation of mineral structures in the rhizosphere zone. The depletion of weatherable
minerals and the concomitant preferential accumulation of weathering products (Al
o
and Feo) close to root surfaces indicate that the weathering regime was stimulated by
root activity.
Apart from the work of April and Keller (1990) who observed the preferential disso-
lution of biotite compared to muscovite close to root surfaces, we are aware of no other
field study on the impact of roots on mineral weathering in forest soils. All the knowl-
edge on the weathering of minerals in rhizospheric environments is based on experi-
ments conducted in greenhouses, and mostly using agricultural plants and K-bearing
minerals. For example, the accelerated weathering of biotite, phlogopite, or illite and the
subsequent release of nutrients (K, Mg, Ca, Fe) in the rhizosphere of wheat (Mortland
et al., 1956), pine seedlings (Boyle and Vogt, 1973), soybean grown with mycorrhizal
association (Mojallali and Weed, 1978), corn grown with symbiotic and nonsymbiotic
microflora (Berthelin and Leyval, 1982), and clover (Tributh et al., 1987) were demon-
strated. The most resistant K-bearing minerals (muscovite, K-feldspar) were not signifi-
cantly affected by the rhizospheric environment. The micas were generally altered to
vermiculite, although Spyridakis et al. (1967) reported the transformation of biotite to
kaolinite in the rhizosphere of coniferous (cedar, pine, hemlock, spruce) and deciduous
(oak, maple) seedlings. Hinsinger and coworkers (Hinsinger and Jaillard, 1993; Hinsinger
et al., 1991, 1992, 1993) further showed that the vermiculitization of phlogopite in the
rhizosphere of ryegrass and rape was a rapid reaction, on the order of days.
80
_ 60
~ ~
en
- 40
If
CI/:I
<
20
o
AI
E
The Rhizosphere and Trace Element Acquisition in Soils 241
AI
Bh
DBulk
AI
Bs
IIIII Rhlzosphere
Fe
E
Fe
Bh
Fe
Bs
Figure 11.4. Acid ammonium oxalate extractable aluminium and iron in bulk and rhizosphere soil
for the three soil fractions and horizons from the control plots at Skogaby. Soil fractions with
differing letters indicate significant differences at the 10% level.
These mineralogical changes between rhizosphere and bulk soils were accompanied
by physical transformations. Sarkar et al. (1979) and Chung and Zasoski (1994) re-
corded a reduction in particle size in the vicinity of plant roots. Whether the increased
clay content was due to root-induced weathering remains unclear. Bruand et al. (1996)
reported the compression of soil particles close to the soil-root interface and measured
an increase in bulk density of up to 20%. April and Keller (1990) observed the tangential
alignment and bending of phyllosilicate minerals, and the fracturing of grains in contact
with roots. Kodama et al. (1994) demonstrated the existence of specific mineral-root
associations. Clay particles were shown to be attached on root surfaces with binding
preferentially occurring where mucilages accumulated. In some cases, clay-size particles
were found to penetrate into the mucilage layer and even into the root surface.
Our field data from forest soils and the results of greenhouse experiments indicate
that the rhizosphere is more corrosive for weatherable minerals than the bulk soil. It
appears that the preferential accumulation and subsequent transformation of organic
tissues (organic matter of plant or animal origin, microflora, microbes) in the rhizo-
sphere plays a central role in mineral alteration. The production of acidic and/or
complexing substances during the decomposition of organic matter, as a consequence of
nutrient uptake or root exudation and following microbial activity, is well established
and was discussed earlier. These processes generate a series of organic acids that can
efficiently attack mineral structures and complex the dissolution products (e.g., metals).
In some instances, inorganic acids such as nitric or sulfuric acids produced by bacteria
during nitrification or the oxidation of sulfides, can also contribute to create an even
more corrosive environment. However, the dominant acids and complexing agents in
the rhizosphere are organic in nature.
The ability of low molecular weight (e.g., citric, oxalic) and complex (e.g., fulvic)
organic acids to weather soil minerals has long been recognized (Stumm et aI., 1985;
242 Fate and Transport of Heavy Metals in the Vadose Zone
Colman and Dethier, 1986; Tan, 1986; Berthelin, 1988). However, the rhizosphere is of
special interest because the total organic acid concentration is higher than in the bulk
soil. Moreover, the nature and distribution of these acids differ from that observed in the
bulk soil (Mench and Martin, 1991; Petersen and Bottger, 1991; Szmigielska et aI., 1996).
It follows that mineral weathering in the rhizosphere can be enhanced, thus favoring the
release of soluble weathering products like nutrient cations or trace metals and their
uptake by the biota.
Tree Growth and Rhizosphere Chemistry
Soil nutrient availability is one of the most important factors influencing tree nutri-
tion, distribution, and health. However, traditional methods of measuring nutrient avail-
ability in soils are too often poorly correlated with plant nutrient uptake and status. The
following example demonstrates that rhizosphere sampling provides more useful infor-
mation about ecosystem changes following perturbation than bulk soil sampling alone.
For example, we present a case in which we link increased nutrient demands due to
stimulated tree growth and rhizospheric chemistry. Rapid growth could deplete nutri-
ents from the rhizosphere when rhizospheric processes cannot keep pace with high de-
mand, particularly if additional nutrient supply is mediated by weathering. The effect of
rapid growth was studied in three field treatments (Table 11.5). The NS and I caused
significant growth increases of 31 % and 20% within the first three years of application
when compared to the control (Nilsson and Wiklund, 1992). Since soil mineralogy, as
indicated in the previous section, showed that the weatherable K-bearing minerals were
scarce, leaving only feldspars, potassium in the soil is presented here as an example.
Exchangeable K and K saturation (K% = KlCEC) had similar trends within soil frac-
tions and horizons. Figure 11.5 presents only K% data and shows that in the control, K%
increased in the order Bulk<Rhizo<SRI, and that the trend was consistent with soil depth
(E, Bh, and Bs horizons). However, due to the treatments (NS and I), the trend in K%
with soil fraction was reversed, e.g., SRI<Rhizo<Bulk. The reversal in K% and K con-
tent was attributed to rapid tree growth and demand, which depleted K in the rhizo-
sphere more than the bulk soil. Normally, such a large depletion of K would have been
missed if only bulk soil was used to link tree growth to soil chemistry.
IMPLICATIONS AND FUTURE RESEARCH
A number of observations point to significant changes in the rhizosphere of soils com-
pared to the bulk fraction. Root-induced changes include higher organic matter and
dissolved organic carbon levels, enhancement of microbial diversity and population counts,
alteration of soil pH and redox potential, accelerated mineral weathering and neoforma-
tion, increased bulk density, and major and trace elements enrichment in both the dis-
solved and exchangeable phases.
The rhizosphere thus appears as a distinct soil environment, enriched in various or-
ganic substances that favor an intense microbial activity which, in turn, together with
exudates and decomposition products, mobilize nutrients and potentially toxic elements
through acidification, dissolution, reduction, and complexation. From this point of view,
the major reason for conferring a central role to the rhizosphere is the abundance of
organic matter and microorganisms; both having a fundamental influence on the bio-
The Rhizosphere and Trace Element Acquisition in Soils 243
o Bulk lID Rhizo EI SRI
6.0
E Horizon
a
4.0
2.0
0.0 +-'--
c NS
6.0
a
Bh Horizon
4.0
~
~
2.0
0.0
C NS
15.0
8
B8 Horizon
10.0
~
0
~
5.0
b
0.0
C NS
Figure 11.5. The potassium saturation (K%) in three horizons and soil fractions from the control
(e), ammonium sulfate (NS), and irrigated (I) plots at Skogaby. Soil fractions with differing letters
indicate significant differences at the 5% level.
geochemical cycle of most major and trace elements. Although this constitutes a reason-
able assessment of the causes of the rhizosphere effect, it is too static and, thus, has a
limited value. The rhizosphere is also an extremely dynamic environment characterized
by a range of feedback processes linking phenomena in the biosphere with processes in
the pedosphere. For example, reduced nutrient availability in the soil can trigger an
244 Fate and Transport of Heavy Metals in the Vadose Zone
increase in the release of exudates by roots. Under Fe-deficient conditions, roots were
shown to increase the exudation of phenolic compounds (e.g., caffeic acids) or of organic
chelators (phytosiderophore) that are highly effective at solubilizing Fe from the soil
solid phase (Marschner and Romheld, 1996). Moreover, these substances not only form
complexes with Fe but they can also mobilize micronutrients like Zn, Cu, and Mn, thus
possibly preventing other types of deficiencies. Similarly, tree growth, the uptake of
nutrient cations by roots, H+ release, and mineral weathering closely interact to assure a
steady flow of matter in the rhizosphere. These interrelated processes contribute to the
nutritional status and productivity of the forest and allow the soil-plant system to with-
stand a range of disturbances and environmental stresses.
Therefore, rhizosphere studies have great implications for the understanding and
modeling of element acquisition in forest ecosystems. Unfortunately, most forest ecosys-
tem models do not incorporate the functional aspects of the rhizospheric environment of
the soil/plant system they describe. For example, Hogberg and Jensen (1994) ques-
tioned the prediction of future forest ecosystem conditions based on models using bulk
soil chemistry and culture solution input data because they may not represent the buff-
ered nutritional condition of the soil surrounding the roots. In this context, due to the
critical interactions that bind plants, microorganisms, and soils, we believe that the study
of rhizospheric processes will provide a key in addressing emerging issues like the pre-
diction of soil quality, trace element mobility, bioremediation of contaminated soils, and
soil-plant relationships in a changing environment.
Future research needs are varied but we believe that our understanding of the bio-
geochemistry of trace elements in terrestrial ecosystems would strongly benefit from
focusing on:
1. The identification, quantification, and characterization of the organic com-
pounds excreted by roots and the dissolved organic substances present in the
rhizosphere
2. The quantification and speciation of the dissolved trace elements found in the
rhizosphere and their response to chemical changes (pH, organic acids)
3. The characterization of the reactive solid phases (organic and inorganic) pro-
duced in the rhizosphere and the mechanisms responsible for their formation
4. The measurement of changes in redox conditions in the rhizosphere with special
emphasis on their impact on the availability of macro- and micro nutrients
5. The evaluation of the temporal and spatial dynamics of microbial populations in
the rhizosphere
6. The development of small-scale sampling and analytical techniques adapted to
the rhizospheric environment that allow the in situ monitoring of processes
7. The measurement of the rhizosphere effect under a range of field conditions
(disturbed, manipulated, natural) and for a variety of vegetation types (conifer-
ous, deciduous, agricultural) with emphasis on forested ecosystems. These
experiments are required to support and validate the observations obtained
from controlled experiments (laboratory or greenhouse), to reduce the lack of
information on forest systems, and to help constrain the scaling of knowledge
from the laboratory to the field.
The Rhizosphere and Trace Element Acquisition in Soils 245
REFERENCES
Alva, A.A., D.A. Eelwards, C.B. Asher, and P.F.C. Blarney. Effects of phosphorus/aluminum
molar ratio and calcium concentration on plant response to aluminum toxicity. SoiL Sci. Soc. Am.
J. 50, pp. 133-137, 1986.
April, Rand D. Keller. Mineralogy of the rhizosphere in forest soils of the eastern United States.
Biogeochemiltry. 9, pp. 1-18, 1990
Barber, S.A. SoiL Nutrient BioavaiLahiLty. A Mechaniltic Approach. John Wiley & Sons, New York,
1984.
Bartlett, R Manganese Redox in the Rhizosphere. Agronomy Ahdtractd, AnnuaLMeeting, IndianapoLil,
Indiana, USA, Nov. 3-8, 1996, p. 204.
Beier, C. and T. Cummins. The Future, and Current Limitations, of Manipulations and Monitor-
ing of Terrestrial Ecosystems. EcOdYdtemd Re.Jearch Report No.4. ExperimentaLManipuLationd of Biota
and BiogeochemicaL CycLing in EcodYdtemd. Commission of the European Communities, 1993, pp.
338-341.
Beijerink, M.W. Die Bakterien der Papilionaccenknollchen. Bot. Ztg. 46, pp. 725-804, 1888.
Berthelin, J. Microbial Weathering Processes in Natural Environments, in PhYdicaL and ChemicaL
Weathering in GeochemicaL Cyc!e.J, A. Lerman and M. Meybeck, Eels., NATO ASI Series, Kluwer
Academic Publishers, Dordrecht, 1988.
Berthelin, J. and C. Leyval. Ability of symbiotic and non-symbiotic rhizospheric microflora of
maize (Zea mayd) to weather micas and to promote plant growth and plant nutrition. Plant Soil
68, pp. 369-377, 1982.
Binkley, D. Foredt Nutrition Management. John Wiley & Sons, New York, 1986, p. 290.
Bolan, N.S. A critical review on the role of mycorrhizal fungi in the uptake of phosphorus by
plants. Plant SoiL 134, pp. 189-207, 1991.
Bowen, G.D. The toxicity of legume seed diffusates toward rhizobia and other bacteria. Plant SoiL
15, pp. 155-165, 1961.
Bowen, G.D. and A.D. Roriva. The Rhizosphere. The Hidden Half of the Hidden Half, in PLant
Rootd: The Hi2den Half, Y. Waisel, A. Eshel, and T. Kafkafi, Eels., Marcel Dekker, Inc., 1991,
pp. 641-669.
Boyle, J.J. and J.R. Shann. Biodegradation of phenol, 2,4-DCP, 2,4-D, and 2,4,5-T in field-
collected rhizosphere and nonrhizosphere soils. J. Environ. QuaL. 24, pp. 782-785, 1995.
Boyle, J.R. and G.K Vogt. Biological weathering of silicate minerals: Implications for tree
nutrition and soil genesis. Plant SoiL. 38, pp. 191-201, 1973.
Brady, N.C. The Nature and Propertied of Soiu, 10th ed., Macmillan Publishing Co., New York,
1990.
Brown, D.A. and A. Ul-Haq. A porous membrane-root-root culture technique for growing plants
under controlled soil conditions. Soil Sci. Soc. Am. J. 48, pp. 692-695, 1984.
Bruand, A., I. Cousin, B. Nicoullaud, o. Duval, and J.-C. Begon. Backscattcred electron
scanning images of soil porosity for analyzing soil compaction around roots. SoiL Sci. Soc. Am.
J. 60, pp. 895-901, 1996.
Chapin, F.S. The mineral nutrition of wild plants. Ann. Rev. Eco!. SYdt. 11, pp. 233-260, 1980.
Chapin, F.S. The evolutionary basis of biogeochemical soil development. SoiL Sci. 57, pp. 223-227,
1993.
Chung, J.B. and RJ. Zasoski. Ammonium-potassium and ammonium-calcium exchange equilib-
ria in bulk and rhizosphere soil. Soil Sci. Soc. Am. J. 58, pp. 1368-1375, 1994.
Clarholm, M. The Microbial Loop in Soil, in Beyond The BiomaM, K Ritz, I. Dighton, and KE.
Giller, Eels., Wiley-Savce Publications, 1984.
Clark, F.E. Soil microorganisms and plant roots. Adv. Agron. 1, pp. 241-288, 1949.
246 Fate and Transp0l1 of Heavy Metals in the Vadose Zone
Clegg, S. and G.R. Gobran. Rhizospheric P and K in forest soil manipulated with ammonium
sulfate and water. Can. J. Soil Sci., 77, pp. 525-533, 1997.
Clegg, S., G.R. Gobran, and X. Guan. Rhizosphere chemistry in an ammonium sulfate and water
stressed Norway spruce [PiCea abie.:J (L.) Karst.] forest. Can. J. SoiL Sci., 77, pp. 515-523, 1997.
Clemensson-Lindell, A. and H. Persson. Effects of freezing on the rhizosphere and root nutrient
content using two sampling methods. Plant Soil, 139, pp. 39-45, 1992.
Cole, D.W. Soil nutrient supply in natural and managed forests. Plant SoiL, 168-169, pp. 43-53,
1995.
Coleman, D.C., c.v. Cole, H.W. Hunt, and D.A. Klein. Trophic interactions as they affect energy
and nutrient dynamics. 1. Introduction. MiCrob. &01. 4, pp. 345-349, 1978.
Colman, S.M. and D.P. Dethier. Rate.J 0/ ChemiCaL Weathering 0/ Rock.! ano Minerau. Academic Press,
New York, 1986.
Conkling, B.L. and R.W. Blanchar. Glass microelectrode techniques for in.!itu pH measurements.
SoiL Sci. Soc. Am. J. 53, pp. 58-62, 1989.
Conkling, B.L., R.W. Blanchar, and T.L. Niblack. Effects of foliar and soil acidity on the
rhizosphere pH of alfalfa, corn, and soybean. J. Environ. QuaL. 20, pp. 381-386, 1991.
Courchesne, F. and G.R. Gobran. Mineralogical variations of bulk and rhizosphere soils from a
Norway spruce stand, southwestern Sweden. SoiL Sci. Soc. Am. J. 61, pp. 1245-1249, 1997.
Cronan, C.S. and D.F. Grigal. Use of calcium/aluminum ratios as indicators of stress in forest
ecosystems. J. Environ. QuaL. 24, pp. 209-226, 1995.
Curl, E.A. and B. Truelove. The RhizNphere. Springer-Verlag, Berlin, 1986.
DeAngelis, D.L., W.M. Post, and C.C. Travis. POditive Feeoback in NaturaL Sy.!temd. Springer-
Verlag, Berlin, 1986.
Dormaar, J.F. Effect of plant roots on chemical and biochemical properties of surrounding
discrete soil zones. Can. J. Soil Sci. 68, pp. 233-242, 1988.
Drever, J.1. and G.F. Vance. Role of Soil Organic Acids in Mineral Weathering Processes, in
OrganiC ACid.! in GeologiCaL Procedde.J. E.D. Pittmsan and M.D. Lewan, Eds., Springer-Verlag,
1994, pp. 138-161.
Fahey, T.J. and J.W. Hughes. Fine root dynamics in a northern hardwood forest ecosystem,
Hubbard Brook Experimental Forest, NH. J. &oL. 82, pp. 533-548, 1994.
FAa-UNESCO. SoiL Map 0/ the WorLo. Revised legend. FAa, Rome, 1988.
Finlay, R.D. Interactions between soil acidification, plant growth and nutrient uptake in
ectomycorrhizal associations of forest trees. &oL. BuLL. 44, pp. 197-214, 1995.
Finlay, R.D. Uptake and Mycelial Translocation of Nutrients by Ectomycorrhizal Fungi, in
MycorrhizaJ in &o.!ydlemd. D.J. Read, D.H. Lewis, A.H. Fitter, and I.J. Alexander, Eds.,
C.A.B. InternationaL Wallingford, 1992.
Finlay, R.D. and B. Soderstom. Mycorrhizal Mycelia and Their Role in Soil and Plant Commu-
nities, in &ofo.qy 0/ Arable Lano. M. Clarholm and L. Bergstrom, Eds., Kluwer Academic
Publishers, 1989.
Fleige, H., O. StrebeL M. Regner, and H. Grimme. Die potentielle P-Anlieferung durch Diffu-
sion als Funktion von Tiefe, Zeit unt Durchwurzelung bei einer Parabraunerde aus Loss. Mitt.
Dt.!ch. Booenko. Ge.J. 32, pp. 381-386, 1981.
FogeL R. Root turnover and productivity of forests. Plant SoiL 71, pp. 75-85, 1983.
FogeL R. and G. Hunt. Contribution of mycorrhizae and soil fungi to nutrient cycling in a
Douglas-fir ecosystem. Can. J. For. 1W. 12, pp. 219-232, 1983.
Fox, R.L. and R.C. Lipps. Distribution and activities of roots in relation to soil properties. Trnd.
Int. Con gr. SoiL Sci. 7th, 1960, 1961.
Fox, T.R. and N.B. Comerford. Low molecular weight organic acids in selected forest soils of the
southeastern USA. SoiL Sci. Soc. Am. J. 54, pp. 1139-1144, 1990.
The Rhizosphere and Trace Element Acquisition in Soils 247
Gobran. G.R and E. Bosatta. Cation depletion rate as a measure of soil sensitivity to acidic
deposition: Theory. Ecol. MooelL. 40. pp. 25-36. 1988.
Gobran. G.R and S. Clegg. A conceptual model for nutrient availability in the mineral soil-root
system. Can. J. SOlt Set: 76. pp. 125-131, 1996.
Gobran. G.R. S. Clegg. and F. Courchesne. Rhizospheric processes influencing the biogeochem-
istry of forest ecosystems. BiogeochemiJtry, in press. 1998.
Gobran G.R. L.B. Fenn. and I. Al-Windi. Nutrition response of Norway spruce and willow to
varying levels of calcium and aluminum. Fert. Ru. 34. pp. 181-193. 1993.
Grayston S.J . D. Vaughn. and D. Jones. Rhizosphere carbon flow in trees. in comparison with
annual plants: The importance of root exudation and its impact on microbial activity and
nutrient availability. App. SoiL. &ol. 5. pp. 29-56. 1996.
Grierson. P. F. Organic acids in the rhizosphere of Bank.1ia intergrifolia LJ. Plant Soil, 144. pp. 259-
265. 1992.
Haby. P.A. and D.E. Crowley. Biodegradation of 3-chlorobenzoate as affected by rhizodeposition
and selected carbon substrates. J. Environ. QuaL.. 25. pp. 304-310. 1996.
Haussling. M. and H. Marschner. Organic and inorganic soil phosphates and acid phosphatase
activity in the rhizosphere of 80-year-old Norway spruce [PtCea ab0 (L.) Karst.] trees. Biol.
FertiL. Sow 8. pp. 128-133. 1989.
Hedley. MJ .. P.H. Nye. and RE. White. Plant-induced changes in the rhizosphere of rape (BrtMJtCa
napLM var. emerald) seedlings II. Origin of the pH change. New PhytoL. 91, pp. 31-44. 1982.
HeiaL H.M. and D.R Sauerbeck. Method to study turnover processes in soil layers of different
proximity to roots. Soil Biol. Biochem. 15. pp. 223-225. 1983.
HellriegeL H. and H. Wilfarth. Beilageheft zu oer Z Ver RubelkJzucker-Ino DtJch RetChJ. 1888. p. 234.
Hendrick. R L. and K S. Pregitzer. The demography of fine roots in a northern hardwood forest.
&owgy 73. pp. 1094-1104. 1992.
Hendrick. RL. and KS. Pregitzer. Patterns of fine root mortality in two sugar maple forests.
Nature 361. pp. 59-61, 1993.
Hendriks. L. and A. Junk. Erfassung den Mineralstoffverteilung in Wurzeln durch getrennte
Analyse von Rhizo- und Restboden. Z Pflanzenern Booenlco. 144. pp. 195-202. 1981.
Hiltner. L. Uber neuere Erfahrungen und Probleme auf dem Gibiet der Bodenbakeriologie und
unter besonderer Berucksichtigung der Grundungung und Brache. Arb DtJch. Lanowirt GeJ. 98.
pp. 59-78. 1904.
Hinsinger. Ph. and B. Jaillard. Root-induced release of interlayer potassium and vermiculitization
of phlogopite as related to potassium depletion in the rhizosphere of ryegrass. J. Soil Sci. 44.
pp. 525-534. 1993.
Hinsinger. Ph . J.E. Dufey. and B. Jaillard. Biological Weathering of Micas in the Rhizosphere
as Related to Potassium Absorption by Plant Roots. in Plant RootJ ano Their Environment. B.L.
McMichael and H. Persson. Eds .. Elsevier. Amsterdam. 1991.
Hinsinger. Ph . B. Jaillard. and J.E. Duffey. Rapid weathering of trioctahedral mica by the roots
of ryegrass. Soil Sci. Soc. Am J. 56. pp. 977-982. 1992.
Hinsinger. Ph .. F. Elsass. B. Jaillard. and M. Robert. Root-induced irreversible transformation
of a trioctahedral mica in the rhizosphere of rape. J. Soil Sci. 44. pp. 535-545. 1993.
Hobbie. S.E. Effects of plant species on nutrient cycling. TrenoJ &0L. Evol. 6. pp. 336-339. 1992.
Hoffmann. W.F. and S.A. Barber. Phosphorus uptake by wheat (TrittCumaeJitivllm) as influenced
by ion accumulation in the rhizocylinder. Soil Sci. 112. pp. 256-262. 1971.
Hogberg. P. and P. Jensen. Aluminium and uptake of base cations by tree roots: A critique of
the model proposed by Sverdrup et al. Water Air Soil PoLLut. 75. pp. 121-125. 1994.
Huang. P.M. Kinetics of Redox Reactions on Manganese Oxides and Its Impact on Environmen-
tal Quality. in RateJ of Soil ChemtCal ProceJJeJ, D.L. Sparks and D.L. Suarez. Eds .. SSSA Special
Publication No. 27. 1991.
248 Fate and Transport of Heavy Metals in the Vadose Zone
Inskeep, W.P. and S.D. Comfort. Thermodynamic predictions for the effects of root exudates on
metal speciation in the rhizosphere. J. Pi1nt Nutr. 9, pp. 567-586, 1986.
Jackson, R.B., J. Canadell, J.R. Ehleringer, H.A. Mooney, O.E. Sala, and E.D. Schulze. A
global analysis of root distributions for terrestrial biomes. Oecologica 108, pp. 389-411, 1996.
Jakobsen, I., L.K Abbot, and A.D. Robbson. External hyphae of vesicular-arbuscular mycor-
rhizal fungi associated with Tri/oLium,fU6terraneum L. 1. Spread of hyphae in phosphorus inflow
to roots. New PbytoL. 120, pp. 371-380, 1992.
Johnson, C.E., C.T. Driscoll, T.J Fahey, T.G. Siccama, and J.W. Hughes. Carbon Dynamics
Following Clear-Cutting of a Northern Hardwood Forest, in Car60n ForfTkl and Functiofhl in Forut
Sow, W.W. McFee and J.M. Kelly, Eds., Soil Science Society of America, Madison, WI, 1995.
Kirlew, P.W. and D.R Bouldin. Chemical properties of the rhizosphere in acid sub-soil. SoiL Sci.
Soc. Am. J. 51, pp. 128-132, 1987.
Kochian, L.v. Mechanisms of Micronutrient Uptake and Translocation in Plants, inMicronutri-
entd in AgricuLture, 2nd ed., J.J. Mortvedt, Ed., Soil Science Society of America, Madison, WI,
1991.
Kodama, H., S. Nelson, A.F. Yang, and N. Kohyama. Mineralogy of rhizospheric and non-
rhizospheric soils in corn fields. Ci1yd Ci1y Minerau 42, pp. 755-763. 1994.
Kosola, KR Sampling Roots of Known Age in the Field. Agronomy A6dtract, American Society of
Agronomy 88tb AnnuaL Meeting 1996. Indianapolis, IN. 1996.
Kuchenbuch. Rand A. Jungk. A method for determining concentration profiles at the soil-root
interface by thin slicing rhizospheric soil. Pi1nt SoiL 68, pp. 391-394, 1982.
Lyncb, J.M. T6e RiJizodP6ere. Jonn Wiley & Sons. 1990.
Macar. N.E. and RD. Graham. Genotypic variation for manganese efficiency in wheat. J. Pi1nt
Nutr. 10, pp. 2049-2055. 1987.
Mahendrappa. M.K. N.W. Foster. G.F. Weetman. and H.H. Krause. Nutrient cycling and
availability in forest soils. Can. J. Soil Sci. 66. pp. 547-572, 1986.
Majdi, H. E/fectd of Nutrient AppLicationd on Fine-Root Dynamic.! and RootlRbiw.Jpbere Cbemi.Jtry in a
Norway Spruce Stand. Swedish University of Agricultural Sciences. Report 71, 1994.
Marschner, H. MineraL Nutrition of Higber Pi1ntd, Academic Press. London. 1986.
Marschner, H. and B. Dell. Nutrient uptake in mycorrhizal symbiosis. Pi1nt Soil 159. pp. 89-102.
1994.
Marschner. H. and V. Romheld. Root Induced Changes in the Availability of Micronutrients in
the Rhizosphere. inpi1nt Rootd: Tbe HWdenHa!j. 2nd ed . Y. Waisel. E. Amram. and U. Kafkafi.
Eds .. Marcel Dekker. Inc . 1996.
Mench, M. and E. Martin. Mobilization of cadmium and other metals from two soils by root
exudates by Zea maYd L.. Nicotina ta6acum L.. and Nicotina rUdtica L. Pi1nt Soil. 132. pp. 187-196.
1991.
Mengel. K, D. Horn, and H. Tributh. Availability of interlayer ammonium as related to root
vicinity and mineral type. SoiL Sci. 149. pp. 131-137. 1990.
Mojallali. H. and S.B. Weed. Weathering of micas by mycorrhizal soybean plants. SoiL Sci. Soc. Am.
J. 42. pp. 367-372. 1978.
Mortland, M.M . K Lawton, and G. Uehara. AI teration of biotite to vermiculite by plant growth.
Soil Sci. 82, pp. 477-481, 1956.
Nilsson, L.a. and K Wiklund. Influence of nutrient and water stress on Norway spruce
production in south Sweden- The role of air pollutants. Pi1nt SoiL 147. pp. 251-265. 1992.
Noble, A.D and M.E. Sumner. Calcium and aluminum interactions and soybean growth in
nutrient solutions. Commun. SoiL Sci. Pi1nt AnaL. 19. pp. 1119-1131. 1988.
Nye. P.H. Acid-base changes in the rhizosphere. Adv. Pi1nt Nutr. 2. pp. 129-153. 1986.
Nye, P.H. and P.B. Tinker. SoLute Movement in tbe SoiL Root SYdtem. Blackwell Scientific Publica-
tions. Oxford, London. Edinburgh, Melbourne. 1977.
The Rhizosphere and Trace Element Acquisition in Soils 249
Parmelee, RW., J.G. Eherenfeld, and RL. Tate. Effects of pine roots microorganisms and nitrogen
availability in two horizons of a coniferous spodosol. BioL FertiL Soiu. 15, pp. 113-119, 1993.
PauL E.A. and F.E. Clark. SoiL Microbiology and Biochemidtry. Academic Press, Inc., 1989, p. 273.
Perry, D.A., M.P. Amaranthus, J.G. Borchers, S.L. Borchers, and RE. Brainerd. Bootstrapping
in ecosystems. Biodcience 39, pp. 230-237, 1989.
Persson, H. The distribution and productivity of fine roots in boreal forests. Plant SoiL 71, pp. 87-
101, 1983.
Persson, H., Y. Von Fricks, and H. Majdi. Root distribution in a Norway spruce [Picea died (L.)
Karst.] stand subjected to drought and ammonium sulfate application. PLant Soil 168-169, pp.
161-165, 1995.
Petersen, W. and M. Bottger. Contribution of organic acids to the acidification of the rhizosphere
of maize seedlings. Plant SoiL 132, pp. 159-163, 1991.
Pidgeon, C.S., C.M. Reynolds, l.K Iskandar, and F.R Magdoff. Rhizosphere Enhanced
Bioremediation of an Alaska SoiL in Agronomy Abdtractd, AnnuaL Meeting, IndianapoLid, Indiana,
USA, Nov. 3-8, 1996, 1996, p. 229.
Reilley, KA., M.K Banks, and A.P. Schwab. Dissipation of polycyclic aromatic hydrocarbons
in the rhizosphere. J. Environ. QuaL 25, pp. 212-219, 1996.
Richards, B.N. The Microbiology of TerredtriaL &odYJtemd. John Wiley & Sons, New York, 1987.
Riley, D. and S.A. Barber. Bicarbonate accumulation and pH changes at the soybean [GLycine
nuz.x (L.) Merr.] soil-root interface. SoiL Sci. Soc. Amer. Proc. 33, pp. 905-908, 1969.
Riley, D. and S.A. Barber. Effect of ammonium and nitrate fertilization on phosphorus uptake
as related to root-induced pH changes at the root-soil interface. SoiL Sci. Soc. Am. Proc. 35, pp.
301-306, 1971.
Rollwagen, B.A. and R.J. Zasoski. Nitrogen source effects on rhizosphere pH and nutrient
accumulation by Pacific Northwest conifers. Plant SoiL 105, pp. 79--86, 1988.
Roriva, A.D. Plant root exudates. Bot. Rev. 35, pp. 35-57, 1969.
Roriva, A.D. and G.D. Bowen. The effects of microorganisms upon plant growth. II. Detoxifi-
cation of sterilized soils by fungi and bacteria. Plant Soil 14, pp. 199-214, 1966.
Sarkar, A.N., D.A. Jenkins, and R.G. Wyn Jones. Modifications to mechanical and mineralogi-
cal composition of soil within the rhizosphere, in The SoiL-Root Interface, J.L. Harvey and R
Scott-Russell, Eds., Academic Press, London, 1979.
Sarkar A.N. and RG. Wyn Jones. Effect of rhizosphere pH on the availability and uptake of Fe,
Mn, and Zn. Plant SoiL 66, pp. 361-372, 1982.
Skladany G.J. and F.B. Meeting, Jr. Bioremediation of Contaminated SoiL in Soil MicrobiaL
&oLogy. F.B. Meeting, Jr., Ed., Marcel Dekker, Inc. 1993, pp. 483-513.
Soil Science Society of America. GLoddary of SoiL Science Termd. Soil Science Society of America,
Madison, WI, 1997, p. 138.
Soon, Y.K and M.H. Miller. Changes in the rhizosphere due to NH4 + and N0
3
- fertilization and
phosphorus uptake by corn seedlings (Zea maYd L.). SoiL Sci. Soc. Am. J. 41, pp. 77--80, 1977.
Spyridakis, D.E., G. Chesters, and S.A. Wilde. Kaolinitization of biotite as a result of coniferous
and deciduous seedling growth. SoiL Sci. Soc. Am. Proc. 31, pp. 203-210, 1967.
Stevenson, F.J. Organic Matter-Micronutrient Reactions in SoiL in Micronutrientd in AgricuLture.
SSSA Book Series, No.4. 1991, pp. 145-186.
Stumm, W., G. Furrer, E. Wieland, and B. Zinder. In The Chemutry of Weathering. NATO ASI
Seried, J.A. Drever, Ed., D. Reidel Publishing Company, Dordrecht, 1985, pp. 55-74.
Szmigielska, A.M., KC.J. Van Rees, G. Cieslinski, and P.M. Huang. Low molecular weight dicar-
boxylic acids in rhizosphere soil of durum wheat. J. Agric. FOOd Chem. 44, pp. 1036-1040, 1996.
Tan, KK Degradation of Soil Minerals by Organic Acids, in Interactiond of SoiL Minerau with
NaturaL Organic.J and Microbu, P.M. Huang and M. Schnitzer, Eds., Soil Science Society of
America, Madison, WI, 1986.
250 Fate and Transport of Heavy Metals in the Vadose Zone
Tinker. P.B. Nutrient Uptake by Plant Roots in Natural Systems. in Nutrient CycLing in Terr&ftriaL
Eco.!y.!tenu. A.F. Harrison. P. Ineson. and O.W. Heal. Eds . Elsevier Applied Science. London.
1990.
Trettin. C.C . M.F. Jurgensen. M.R Gale. and J.W. McLaughlin. Soil Carbon in Northern
Forested Wetlands: Impacts of Silvicultural Practices. in Carbon Fornu and Function.! in For&ft
Soiu. W. W. McFee and J .M. Kelly. Eds .. Soil Science Society of America. Madison. WI. 1995.
Tributh. H . E. von Bogulawski. A. von Lieres. D. Steffens. and K Mengel. Effect of potassium
removal by crops on transformation of illitic clay minerals. SoiL Sci. 143. pp. 404-409. 1987.
Ulrich. B. Stability. Elasticity. and Resilience of Terrestrial Ecosystems with Respect to Matter
Balance. in EcologicaL StUdied 61, E.-D. Shultze and H. Zolfer. Eds . Springer-Verlag. Berlin.
Heidelberg. 1987.
Uren. N. C. and H.M. Reisenauer. The role of nutrient exudates in nutrient acquisition. Adv. Plant
Nutr. 3. pp. 79-114. 1988.
Van Breemen. N. Soils as biotic constructs favoring net primary production. Geoderma 57. pp.
183-211. 1993.
Van Breemen. N. Nutrient cycling strategies. Plant SoiL 168-169. pp. 321-326. 1995.
Van Cleve. K and RF. Powers. Soil Carbon. Soil Formation and Ecosystem Development. in
Carbon Fornu and Function.! in For&ft Sow. W.W. McFee and J.M. Kelly. Eds . Soil Science
Society of America. Madison. WI. 1995.
Vogt. KA . D.A. Publicover. J. Bloomfield. J.M. Prerez. D.J. Vogt. and W.L. Silver. Belowground
responses as indicators of environmental change. Environ. Exp. Botany 33, pp. 189-205. 1993.
Waring, RH. and W.H. Schlesinger. For&ft Eco.!y.!tenu, Academic Press, 1985.
Whittig, L.D. and W.R Allardice. X-ray diffraction techniques. in MethOd.! of SoiLAnaLy.!iJ. Part 1.
Phy.!icaL and MineralogicaL MethOd.!, 2nd ed., A. Klute, Ed., Agron. Monogr. 9. ASA and SSSA,
Madison, WI. 1986.
Wild, A. Soil scientists as members of the scientific community. J. SoiL Sci. 40, pp. 209-221, 1989.
Wilkins, D.A. The influence of sheathing (ecto) mycorrhizas of trees on the uptake and toxicity
of metals. Agric. EcO.l. Environ. 35, pp. 145-160, 1991.
Wood, T., F.H. Bormann, and G.J. Vogt. Phosphorus cycling in a northern hardwood forest:
Biological and chemical control. Science. 223, pp. 391-393. 1984.
CHAPTER 12
Distribution of Ecologically Significant
Fractions of Selected Heavy Metals in the
Soil Profile
T. Nemeth, K. Bujtas, J. Csillag, G. Partay, A. Lukacs, and M.Th. van Genuchten
INTRODUCTION
The amount of wastes, wastewaters, and sewage sludges produced by agricultural,
industrial, and municipal activities is rapidly increasing worldwide. In the developing
regions of the world this may be simply the result of the improving supply of clean tap
water and canalization. Because of increasing environmental awareness, dumping of
sewage into surface waters is subject to more strict regulations, thus the amount of waste-
waters subjected to treatments is also increasing in nonindustrialized countries. In con-
sequence, a large growth in sewage sludge production may be expected, especially when
taking into account the higher requirements and standards for wastewater treatment.
In Hungary approximately 1000 million cubic meters (mS) of wastewater were pro-
duced per year in the middle of the '80s, of which only 187 million m
3
were sufficiently
treated, the majority only partially treated, and 173 million m
3
not treated at all. At that
time, there was an increasing gap between the development of municipal water supply
and of sewage systems, with the latter lagging behind the substantial improvements in
the water supply. To develop the collection and proper treatment of liquid wastes is still
a problem for many smaller municipalities. According to recent data, the amount of
sewage sludges in Hungary was above 1 million m
3
per year. About 40% of these sludges
were deposited on agricultural fields and on forest plantations.
One reasonable and economic way to dispose of wastewaters and sludges is to apply
them to agricultural fields, thereby exploiting their water and nutrient content. Cur-
rently, this practice is becoming increasingly important in many countries. In the early
nineties about 30-50% of the sewage sludges were disposed by land application in the
majority of the industrialized European countries, which compares to 33% of the annual
sludge production in the United States (McGrath et aI., 1994).
251
252 Fate and Transport of Heavy Metals in the Vadose Zone
Sludge Application
Excessive application of wastewater and sewage sludge to agricultural land may cause
soil and groundwater pollution problems, since heavy metals and other, potentially toxic
elements may be regarded as characteristic contaminants of sewage sludges and waste-
waters. These elements often occur in large amounts in these wastes, limiting their appli-
cability in agriculture (Chang et aI., 1984; Juste and Mench, 1992).
Sludge application in Hungary is officially regulated in order to prevent or limit the
pollution of soils, surface waters and groundwater (Hung. Techn. Dir., 1990; Molnar et
aI., 1995). The regulations use a 3-step system in which water and nitrogen content of
the sludge, and the concentration of any toxic element in the sludge and in the sludge-
treated soil are the limiting factors. Also, limiting conditions have been defined for each
soil factor related to the application of wastewaters and sewage sludge on a specific
agricultural field. Taking into consideration the hazard of contamination of surface and
subsurface waters, a permanent monitoring is required on fields where these materials
are used, in order to control the load effect. The upper limits for the toxic elements used
in the case study discussed in later sections are 15 mg Cd, 1000 mg Cr, 200 mg Ni, 1000
mg Pb, and 3000 mg Zn per kg sludge dry matter (Hung. Techn. Dir., 1990). Assuming
average sludge application practices (i.e., incorporating 500 tlha sewage sludge contain-
ing 5% dry matter into a 20-cm surface soil layer), these limits are equal to loadings of
0.375 kg Cd, 25 kg Cr, 5 kg Ni, 25 kg Pb, and 75 kg Zn per hectare. The cumulative
amounts of the toxic elements after prolonged application of sludge or of any other con-
taminant sources must not exceed their maximum acceptable concentrations in the soil,
which for the metals selected for the present study are the following: 6.4 kg Cd, 320 kg
Cr, 64 kg Ni, 320 kg Pb, and 800 kg Zn per ha for soils with an adsorption capacity of
15-25 cmol/kg soil as calculated from maximum allowed limits of soil concentrations.
Although the limit values relating to sludge application practice are generally lower in
Hungary than in the European Community or in the United States (Hung. Techn. Dir.,
1990; USEP A, 1989; Chang et aI., 1992; McGrath et al., 1994), improper or illegal depo-
sition of sludges and industrial wastes may lead to serious contamination of the Hungar-
ian environment with toxic substances.
Adsorption and Mobility
The potential risks for heavy metal contamination in agricultural fields depend on the
relative amounts of metals being adsorbed versus amounts in the soil solution. Poten-
tially toxic elements may become especially hazardous to the environment, when they
enter the liquid phase of the soil. Soil solution concentrations are regarded to be indica-
tors of the mobile pool of metals in soils (Kabata-Pendias and Adriano, 1995). Water-
soluble forms of an element may move through the vadose zone (Kabata-Pendias and
Pendias, 1992). Although the removal of metals from sludge-treated soils by leaching to
the subsurface is normally very small (McGrath et aI., 1994; Kabata- Pendias and Adriano,
1995), some literature data indicate that sludge-born metals may get further from the
site of contamination under specific circumstances (McBride, 1995).
There are contradictory views in the literature regarding the extent to which sludge-
born metals are vulnerable to leaching. Differences among experimental conditions may
be one reason for the differences observed in the movement of various metals in the soil.
Distribution of Ecologically Significant Fractions of Selected Heavy Metals 253
For instance, Legret et al. (1988) found significant migration of Ni and especially of Cd,
but little or no movement of Pb and Cr during a field experiment on a sludge-treated,
coarse-textured soil. Dudka and Chlopecka (1990) observed leaching of Cd, Cr, Ni, Pb,
and Zn in a lysimeter experiment on a sandy loam contaminated with sewage sludge
containing large amounts of these elements. In contrast, Dowdy and Yolk (1983), Chang
et al. (1984), and Alloway (1990) reported retention of these elements within the zone of
sludge incorporation during both column and field experiments.
Metals originating from anthropogenic sources usually occur in the soil in forms dif-
ferent from the original, native metal content of the soil, thus their mobility and avail-
ability is also different. This may change when the freshly added metals enter the dynamic
equilibria among the various forms of the elements in the soil chemical processes. When
evaluating the biological and ecological impacts of soil contamination, it is necessary to
estimate the total amounts of the toxic elements which may become available even after
longer periods. Treatment of soil samples with 2 mollL HN0
3
at 100C as proposed by
Andersson (1976) was shown to be a suitable method to determine the totaL potentiaLLy
available /ractwn of the metals in the soil.
Extractions and Bioavailability
The plant-available concentratwtU of heavy metals occurring in the soil are typically esti-
mated by different chemical extraction methods. Several extractants and extraction pro-
cedures have been proposed to determine the availability of essential elements, as well as
of some elements potentially toxic to plants. Much research has been carried out to
develop a universal extractant characterizing plant availability, and applicable to several
elements in soils having widely different properties. One of the most generally useful
extractants of this type is acid ammonium acetate + EDT A (AAAc- EDT A). This extrac-
tant has been found suitable for the simultaneous extraction of both macro- and micro-
nutrients (Lakanen and Ervio, 1971), and is now included in the procedures adopted by
the Soil Advisory Service in Hungary. Extensive F AO studies involving 30 countries
proved that this extraction methodology is appropriate for the micronutrients Zn, Fe,
Cu, and Mn (Sillanpaa, 1982), and also for some toxic elements such as Cd, Pb, Co, and
Se (Sillanpaa and Jansson, 1992).
Theoretically, the best measure of the availability of elements in a specific soil-plant
system should be their actual amounts taken up by the plants. These amounts are often
poorly correlated with extractable amounts of the elements in the soil (Marschner, 1991;
Chang et aI., 1992; McBride, 1995), since they depend not only on soil factors but also
on several plant properties and are governed by physiological and biochemical processes.
In contrast, soil extraction methods depend on the laws of soil chemistry. Thus, the
fractions of the element content utilized by the plants and assessed by plant analyses are
not the same as those measured by the soil chemist. However, plant analyses have many
practical limitations, and plant uptake processes are susceptible also to environmental
factors (Sillanpaa and Jansson, 1992). Hence, chemical extraction methods remain use-
ful tools to assess phytoavailability of the elements in the soil.
Recent evidence suggests that elemental concentrations in the soil solution itself may
serve as a useful diagnostic tool for plant uptake of the various elements. It is reasonable
to expect that plant uptake will be a function of the soil solution concentrations since
254 Fate and Transport of Heavy Metals in the Vadose Zone
water-soluble forms of an element are generally most easily and immediately available
for plant uptake (Petruzzelli, 1989). However, there is only a limited amount of data in
the literature to support this idea for the toxic elements (Alloway, 1990).
Several laboratory methods have been developed for obtaining the liquid phase from
undisturbed bulk soils or from air-dried, ground and rewetted soil samples. Many inves-
tigators attempted to characterize the energy status of the liquid phase to be separated
from soils. Some of these studies focused mainly on macro-elements (Zabowski and
Ugolini, 1990; Jones and Edwards, 1993; Csillag et al., 1995). Others have determined
the heavy metal content of the soil solution (e.g., Mullins and Sommers, 1986; cit. in
Campbell and Beckett, 1988; cit. in Kabata-Pendias and Pendias, 1992).
CASE STUDY
The initial hypothesis of this case study presented here is as follows. Following the
application of heavy metals to the soils, quasi-equilibrium concentration in the soil solu-
tion fraction which is held in the soil by forces corresponding to less than -1500 kPa (pF
4.2 = the conventional wilting point of plants), i.e., which are directly accessible to the
roots, may be regarded as the most important variable characterizing plant availability
of the elements. To test this hypothesis, sewage sludge spiked with Cd, Cr, Ni, Pb, and
Zn nitrates was applied to the top layer of large undisturbed soil monoliths. The distri-
butions of the total potentially available and the plant-available fractions (characterized
by the AAAc-EDTA extractable amounts and also by the directly plant-accessible soil
solution concentrations) of the applied metals in the soil profile of the monoliths are
discussed here. Other aspects of the study were presented elsewhere (Nemeth et al.,
1994; Bujtas et al., 1995).
A brown forest soil (Ochrept, from Godollo) and a slightly acidic sandy soil (Psamment,
from Somogysard) were included in the experiments, each one with four monoliths. The
major physical and chemical properties of these soil types are shown in Table 12.1. The
undisturbed, 40-cm diameter, lOO-cm long soil monoliths were prepared following the
methods proposed by Homeyer et al. (1973) and modified by Nemeth et al. (1991). The
monoliths were excavated at the selected field sites and their cylindrical surfaces were
coated with fiberglass cloth impregnated with a synthetic resin. After the coatings solidi-
fied, the monoliths were lifted, and the bottoms were similarly coated. The monoliths
were subsequently transported by truck to the laboratory. The coatings made extremely
close contact with the soil surface by imbibing the outer micropores, thus eliminating
possible "wall effects" during the experiments. The excellent insulating properties of the
coating also eliminated leakage from the monoliths.
Weights of the columns at the beginning and at the end of the experimental procedure
were recorded on a movable scale. Changes in temperature, soil moisture content, and gas
composition along the soil profile were followed by sensors inserted into the monoliths at
various depths through holes drilled in the coatings. Temperature was recorded daily.
Ambient temperature during the experiments was about 25C, with 4.1 C difference be-
tween the two extreme values. Variation in temperatures of the monoliths was less than
2.0C. Soil water content was followed by time-domain reflectometry (TDR), and was
regulated along the soil profile by saturating the columns from the bottom through a spe-
cial built-in valve connected to a hanging water-column or by sprinkler irrigation at the
Distribution of Ecologically Significant Fractions of Selected Heavy Metals 255
Table 12.1. Some Chemical and Physical Properties of the Soils Used in This Study
Depth pH Organic Matter CEC <0.02 mm
(em) Horizon H
2
O KCI (g/kg) (cmol/kg soil) SP (%)
Brown forest soil, Oehrept (Godollo)
0-8 Ap 5.90 5.02 10 9.0 25 19.5
8-16 A1 6.26 5.29 12 8.5 28 18.9
16-43 B 5.85 5.01 11 8.5 28 20.5
43-66 BC 6.51 5.24
a
10.1 28 23.3
>66 C 7.15 5.75
a
58.0 74 54.4
Slightly acidic sandy soil, Psamment (Somogysard)
5-15 Ap 5.63 4.48 13 6.7 27 20.2
40-50 B 6.07 4.84 3 4.7 30 12.2
80-90 C 6.38 5.56
a
2.3 24 5.0
a No measurements.
soil surface. Irrigation water was added on the basis ofTDR measurements and/or weight-
ing. Soil gas phase was sampled by special capillary microsensors, and concentrations of
water vapor, N
2
, O
2
, and CO
2
were measured by quadrupole mass spectrometry (QMS)
(Partay et aI., 1994). Supplemental light was provided in 12 hours day/night cycles.
The monoliths were air-dried to constant weight, then via a bottom-valve gradually
filled up to the surface with deionized water. This step was followed immediately by
gravitational drainage through the bottom-valve. The aim of these procedures was to
bring the soils to a fairly uniform physical status with a moisture content of maximum
water capacity (pF = 0). Also, the procedure compensated the temporary shrinkage of
monoliths caused by the long initial air-drying period of the field moist soils.
When gas composition and soil moisture data remained constant for several days after
completion of drainage, communal sewage sludge spiked with Cd-, Cr-, Ni-, Pb-, and
Zn-nitrates was mixed into the upper 10 cm of the soil. Dry matter content of the com-
pressed sludge was 20.6%, and the inorganic matter content 48.2%. Original concentra-
tions of the selected metals in the sludge were: 12.3 mg Cd, 217 mg Cr, 109 mg Ni, 210
mg Pb, and 3026 mg Zn per kg d.m. These values are comparable to or less than the
limits specified in the Hung. Techn. Dir. (1990). The metal nitrates were added to the
sludge so that the final metal loading rates in the soil were equivalent to 10, 30, and 100
times the permitted loading limits (L-values) specified in the Hungarian Technical Di-
rective assuming average sludge application practices in Hungary (i.e., 500 tlha sewage
sludge containing 5% dry matter incorporated into a 20-cm surface soil layer ). Loadings
corresponding to 1L are 0.125 mg Cd, 8.33 mg Cr, 1.67 mg Ni, 8.33 mg Pb, and 25 mg
Zn per kg soil. Original, unspiked sludge was used as control treatment at rates of 400 g
sludge dry matter per column, resulting in loadings of 0.25 mg Cd, 4.48 mg Cr, 2.25 mg
Ni, 4.34 mg Pb, and 62.5 mg Zn per kg soil. Identical amounts of the same sludge were
used for each treatment in order to obtain as uniform conditions as possible in terms of
such additional factors in the sludge as organic matter content, nutrient levels, and con-
centrations of other elements.
One week after sludge application, nine corn (Zea may.:! L.) seeds were sown per mono-
lith. When the plants reached a suitable developmental stage, microsensors of the QMS
256 Fate and TranspOIi of Heavy Metals in the Vadose Zone
system were implanted into the stems (one per monolith) to study the gas metabolism in
vivo inside the plants. By the end of the experiment the plants had already finished their
vegetation period. Mter harvesting the mature corn plants, the upper parts of the mono-
liths were cut into four consecutive soil layers at the 0-10, 10-15, 15-20, and 20-30 cm
depth intervals. Water potentials of the soil samples were determined, as were relative
root distributions.
Soil pH and total potentially available metal concentrations in the soil (after 2 mol/L
HN0
3
extraction as described by Andersson, 1976) were measured in air-dried soil
samples, and plant-availability of the metals was estimated by: (a) metal concentrations
in acid ammonium acetate-EDTA soil extracts (Lakanen and Ervi6, 1971) of air-dried
soil samples, and by (b) directly plant-available concentrations in the soil solution ob-
tained by centrifugation of the moist soil samples immediately after the cutting proce-
dure, from each layer of the monoliths in triplicates (Csillag et al., 1995).
Elemental concentrations in the 2 mol/L HN0
3
and AAAc-EDTA soil extracts, and
in the soil solution, were measured by inductively coupled plasma atomic emission spec-
trometry (ICP-AES). Concentration values in the soil solution were related to the mass
of the dry soil instead of the volume of the liquid phase, in order to eliminate the differ-
ences caused by the slightly different moisture content of the soil samples (water poten-
tial == pF 0). Recovery of the elements in the potentially available forms was related to
added metal amounts. Metal budgets, based on the total potentially available amounts
were calculated for each depth increments and for the whole columns. The so-called
"soil available factor," SA, expressing the "percentage of the total content of an element
in the soil which is available for uptake to plants" (Coughtrey et aI., 1985), is often used
when describing the distribution and transport of radionuclides in terrestrial ecosys-
tems, but it can be applied similarly to the stable forms of the elements. A similar ap-
proach was adopted to evaluate the availability of the applied metals in our experiments.
Metal concentrations measured either in the plant-available soil solution fractions (cs)
or in the AAAc-EDTA extracts (CAE) were related to the total potentially available metal
concentrations measured in the 2 mol/L HN0
3
soil extract (c
m
):
The calculated percentages (SA
s
and SAAE) characterize the proportion of the directly
plant-available and of the supposedly plant-available metal forms, respectively, in the
total potentially available metal pool of the soil. Also, the total amounts of the metals
found in the liquid phase of the soil layers were calculated from the metal concentrations
of the soil solution samples, for each depth increments and for the whole columns. These
values were compared to the total potentially available metal contents of the columns.
NITRIC ACID EXTRACTION
The 2 moLIL nitric acW-extractahLe concentrations of all the five metals increased propor-
tionally to the initial loading rates in the upper 10 cm layer, i.e., in the initially contami-
nated zone containing the metal-spiked sludge. Figure 12.1 shows linear relationships
between the extracted total metal concentrations and the metal loadings (L) for the top
layer in the acidic sandy soil. Similar linear relationships were obtained also for the
Distribution of Ecologically Significant Fractions of Selected Heavy Metals 257
2000
1500
=5 1000
II)

Cl
E 500
-
.........
........
..
..
Zn ..........
.'
........
. '
.'
..... Pb
..
.
.... ....
....
....
ID
O+--.......
0 10 20 30 40 50 60 70 80 90 100
I-

o 10 20 30 40 50 60 70 80 90 100
LOADING RATE (1 OL, 30L, 1 OOL)
Figure 12.1. Correlations between 2 maUL HN0
3
-extractable soil concentrations and loading rates
of the metals in the 0-10 cm layer of the slightly acidic sandy soil.
brown forest soil. The measured total metal concentrations in the top 10 cm soil layer
accurately reflected at each loading level the differences among the application rates of
the five metals, being the highest for Zn and the lowest for Cd, and very similar for Cr
and Pb which were applied in identical amounts (Figures 12.1, 12.2a.). The differences
are shown by the values of the slopes of the linear regression equations, calculated for
the relationship shown in Figure 12.1, where Y is the total extracted metal concentration
and X the applied metal loading (L):
Cd: Y = 0.076 X + 1.807 R2 = 0.9945
.
Cr: Y = 5.17 X + 13.87 R2 = 0.9984
Ni: Y = 1.03 X + 11.75 R2 = 0.9977
Pb: Y = 5.63 X + 8.75 R2 = 0.9984
Zn: Y = 16.3 X + 102.5 R2 = 0.9983
The slopes correspond closely to the ratios among the loading rates of the metals, i.e.,
Cd : Cr: Ni : Pb : Zn = 0.075 : 5 : 1 : 5 : 15. The brown forest soil showed no substantial
increases in metal concentrations in the deeper layers, even at the higher metal applica-
tion rates, with the exception ofNi and Zn which had slightly elevated concentrations in
258 Fate and Transport of Heavy Metals in the Vadose Zone
W
--l
en
;::
~
l-
X
W
2000
a. 0-10 em depth
1500
1000
500
Cr Ni Pb
200
150
b. 10-15 em depth
100
50
Cr Ni Pb
50 ]
0- CTJ:O
Cr
c. 15-20 em depth
,rTUJ
Ni Pb
Cd
.1
1
=
a. 0-10cm b. 10-15 cm c. 15-20 cm
~ ~ ~ ~ I brown forest soil
slightly acidic sandy soil
Zn
Zn
Zn
Each set of bars represents the applied loadings from left to right:
control(L), 10L, 30L and 100L
Figure 12.2. Metals extracted with 2 maUL HN0
3
.
the 10-15 cm and 15-20 cm depth interval at the highest contamination level. For the
more acidic sandy soil we observed a somewhat more pronounced downward movement
of the elements below 10 cm at the 1 OOL loading rate (Figure 12.2b).
Distribution of the total metal contents among the sampled layers of the upper 20 cm
also shows this feature (Table 12.2). Although most of the metals (generally more than
95% of the total potentially available contents of the upper 20 cm) were found between
o and 10 cm, small but significant amounts of the applied metals from the metal nitrate
enrichments of the sludge were recovered in the layers between 10 and 15 cm, especially
Distribution of Ecologically Significant Fractions of Selected Metals
--
Table 12.2. Recovery of the 2 mol/L-Extractable Metal Forms in the Top 20 cm from the Metal
Nitrates Added as Enrichment of the Sewage Sludge, in the Slightly Acidic Sandy Soil
Recovered Amount of the Metals
Added
0-10 cm 10-15 cm 15-20 cm Total Metals
Loadings mg/column mg/column % mg/column
Cd 18.8 0.1 -0.1 18.8 77 24.4
Cr 1190 25.7 0 1220 75 1630
10 L Ni 252 7.2 2.4 261 80 326
Pb 1240 35.0 2.8 1280 79 1630
Zn 4080 36.6 4.5 4120 126 3260
Cd 38.2 -1.0 -1.6 35.6 49 73.3
Cr 2770 31.6 -4.5 2800 57 4890
30 L Ni 548 5.9 -4.7 549 56 978
Pb 2980 48.6 0.2 3020 62 4890
Zn 9030 76.6 -14.9 9090 93 9780
Cd 145 3.5 -0.4 148 60 244
Cr 9850 266 4.2 10100 62 16300
100 L Ni 1960 105 6.2 2070 64 3260
Pb 10600 291 3.8 10900 67 16300
Zn 31300 1160 65.2 32500 100 32600
at the highest loading rate. Metal budgets for the brown forest soil showed similar fea-
tures, with somewhat higher total recovery but less downward movement of the metals.
Metals originating from the sludge might not have reached the deeper layers as it is
indicated by the distribution of the total potentially available metal contents in the con-
trol treatments of both soils. Although in these treatments the total amounts of the met-
als, reflecting both the native metal content of the soils and the metal amounts added in
the original sewage sludge, were different in the two soils; metal contents in percentage
of the total content of the upper 20 cm were nearly identical, with about 60-70% of the
metals found in the sludge-containing zone, and about 15-20% in each of the two deeper
5-cm-Iayers (Table 12.3). The similarity of the metal contents in the two deeper layers
indicates that no metal may have moved downward from the sludge.
Andersson (1976) showed that the 2 mol/L HN0
3
extraction procedure gave a good
estimate of the total pollution potential of the soil from heavy metals: the method re-
leased 57% (Cr)-86% (Cd) of the total content of various heavy metals from normal,
unpolluted soils, and between 65 and 92% of the total content from sewage sludge-treated
soils. The differences between the extracted amounts accounted for 82-96% of the total
amounts of Cd, Cr, Cu, Ni, Pb, and Zn accumulated during several years of sewage
sludge application. Similar results were obtained in our laboratory (A. Lukacs, personal
communication), where after 2 mol/L HN0
3
extraction the recoveries of Cr, Cu, Ni, Pb,
and Zn were between 69 and 99% of the amounts measured in aqua regia extracts in a
sewage sludge amended soil (BCR No. 143, 1983), but only between 21 and 77% when
compared to the total metal content in an unpolluted reference soil sample (CCRMP,
1979) that contained the metals mostly in various minerals. The results suggest that
nearly the total elemental content, with the exception of the most strongly fixed, residual
forms are extracted from the soil by this method, which thus may be regarded as a good
estimation of the total potentially available amount of the elements in the soil.
260 Fate anci Transport of Heavy Metals in the Vadose Zone
Table 12.3. Distribution of the 2 mol/l Nitric Acid-Extractable Metal Contents in the Upper 20
em of the Control Columns
0-10 em 10-15 em 15-20 em
Total
mgleolumn % mg/eolumn % mgleolumn % mg/eolumn
Brown forest soil
Cd 25 63.4 8 19.5 7 17.1 39
Cr 256 62.2 80 19.5 75 18.3 412
Ni 292 57.6 113 22.3 102 20.1 507
Pb 300 58.1 111 21.6 105 20.3 516
Zn 1880 74.3 327 12.9 326 12.8 2540
Slightly acidic sandy soil
Cd 34 54.2 15 23.7 14 22.1 63
Cr 200 58.6 73 21.4 68 20.0 341
Ni 203 54.9 86 23.3 80 21.8 369
Pb 148 63.6 42 18.2 42 18.2 233
Zn 1440 72.2 316 15.8 240 12.0 2000
In our experiments where the soil was contaminated by a low-metal sewage sludge
spiked with metal nitrates and the extraction procedure took place about 3 months after
the metal contamination, relatively less of the metals was extracted by this method at the
extremely high contamination levels, but recovery of the metals was remarkably uni-
form for Cd, Cr, Ni, and Pb. When calculating the metal budgets, it was found that at
10L loading rate in the slightly acidic sandy soil, about 80% of the applied metal-nitrate
enrichment of the sludge was recovered from these four metals by the 2 mol/L nitric
acid-extraction in the upper 20 cm of the monoliths (Table 12.2). Total recovery was
somewhat smaller, about 50-60% at the two higher loading rates. The added Zn was
practically totally recovered in the upper 20 cm soil layer at all application rates.
AAAc-EDT A EXTRACTION
This extraction provided similar results to the nitric acid extraction. In the slightly
acidic sandy soil the concentrations of the metals increased in the layer between 0-10 cm
when the initial loading rates were higher (Figure 12.3). Very small increases (detect-
able only at the highest load) were found also in the layer below the application zone
(Figure 12.4), that is between 10-15 cm for the more mobile Cd, Ni, and Zn, and for Pb
as well. Similar results were obtained for the brown forest soil (Figure 12.3), but the
AAAc-EDTA extractable amounts did not increase in the deeper layer even at the high-
est application rate as it is shown on the example of Zn (Figure 12.4).
The ratios among the metal concentrations in the AAAc-EDTA extracts were some-
what different from the application ratios. The AAAc- EDT A extractable concentrations
of Cd, Ni, and Zn reflected fairly well the original application rates. In contrast, rela-
tively more Pb was extracted than Cr (Figure 12.3) although the initial application rates
of these elements were identical. The concentration of Pb in the extracts followed well
the loading rates, while the concentration of Cr was relatively smaller and decreased
with increasing loading rates as compared to the applied amounts. The difference be-
tween these two elements at the lower application rates was about twofold, while at the
Distribution of Ecologically Significant Fractions of Selected Heavy Metals 261
:::::-
0
III
Cl
~
-
Cl
E
-
UJ
...J
co
<C
r-
()
~
r-
X
UJ
--------
1400
1200
1000
800
600
400
200
0
Cd Cr Ni Pb
brown forest soil
slightly acidic sandy soil
Each set of bars represents the applied loadings from left to right:
control(L), lOL, 30L and 100L
Figure 12.3. Metals extracted with AAAc-EDTA at 0-10 em depth.
~ ------- .,.
Zn
highest application rate the extractability of Cr was only 1/6 that of Pb in both soils. In
contrast, 2 mol/L HN0
3
-extracted amounts of Cr and Pb were similar, in agreement
with the application rates as was shown on Figures 12.1 and 12.2a.
The similar behavior of Cd, Ni, Pb, and Zn with respect to extractability by the acid
ammonium acetate + EDT A is in agreement with the stability constants of the reactions
of these metals with EDTA (Me + L +--+ MeL), given in Lindsay (1979) as 16.36 for Cd,
18.52 for Ni, 17.88 for Pb, and 16.44 for Zn. Chromium forms more stable complexes
with EDTA; value of the stability constant for Cr(Ill) EDTA is 24.0. However, com-
plexes of Cr(Ill) are described as "inert" since they attain the equilibrium very slowly
(Dwyer and Mellor, 1964). Such behavior might have resulted in a smaller extractabil-
ity of Cr during the limited period of the extraction procedure, despite its stronger affin-
ity to EDTA.
Among the components of the AAAc-EDTA extractant, the acid and neutral ammo-
nium acetate extract the readily soluble and exchangeable fractions of the trace ele-
ments. The addition of EDT A as a suitable chelating agent makes possible the extraction
of trace elements bound by soil organic matter (Lakanen and Ervi6, 1971). However,
only about 5% of the total Cr in soil was shown to be available to common extracting
solutions such as acetic acid or EDT A, and the majority of Cr entering the soil is rapidly
immobilized (Coughtrey and Thorne, 1983). Neutral and acidic ammonium acetate were
shown to dissolve only very small fractions of Cr (Andersson, 1976). In sludge-treated
soils, a great proportion of Cr was bound to Fe-Mn-oxides (Dudka and Chlopecka,
262 Fate and Transport of Heavy Metals in the Vadose Zone
1500
1200
..-..
900
0
!/)
Cl
600
~
-
Cl
E
300
'-"'
W
--l
a:l 0
~ 0
0

0::
l-
X
W
150
100
50
o
a. 0-10 cm depth
brown forest soil
-.
-.
- --
-13
slightly acidic sandy soil
50
b. 10-15 cm depth
..
slightly acidic sandy soil ..
brown forest soil
50
LOADING RATE (L)
100
100
Figure 12.4. AAAc-EDTA extractable amounts of Zn in the contaminated zone and in the layer
below it at different loading rates.
1990) and/or was found in residual forms (Legret et al., 1988). Thus, the AAAc-EDTA
mixture may not be expected to extract most of the Cr.
CONCENTRATIONS IN SOIL SOLUTION
Metal concentrations in the soil's liquid phase were determined using a centrifuga-
tion sampling technique to obtain soil solution fractions available for plant uptake. The
centrifugal speed was calculated by applying an equation used by Cassel and Nielsen
(1986), to correspond to -1500 kPa water potential, which is the conventional value of
the wilting point of plants. Plants generally cannot take up those fractions of soil water
which are held in the soil more strongly than -1500 kPa; thus the soil solution fractions
separated with -1500 kPa are considered to represent the liquid phase available for the
plants at natural soil water contents because they are retained in the soil with suctions
less than the suction corresponding to the wilting point. From water-saturated soil, for
example, solutions at soil water potentials between -1500 and -0.1 kPa are separated
with this technique, while from samples at field capacity, solutions between -1500 and-
20 kPa can be obtained. By analyzing these solution fractions, the quantity and chemical
Distribution of Ecologically Significant Fractions of Selected Heavy Metals 263
composition of the soil solution utilizable for the plant (in the sense of energy condi-
tions) is modeled.
Other methods applied widely for the separation of the soil solution do not represent
soil moisture available to plants. In the case of suction methods, for example, the maxi-
mum suction that can be exerted is -100 kPa, whereas plants can exert much higher
values. Displacement methods also do not give information about the quantity and en-
ergy status of the solution remaining in the soil after the extraction. It is also probable
that the extreme high pressures applied to the moist soil samples in a hydraulic pressure
apparatus during the extraction of the soil solution, disturb the prevailing equilibrium
between the soil phases.
Metal concentrations in the soil solution were generally several orders of magnitude
lower than the 2 mollL HN0
3
extractable total concentrations, with the exceptions of
Cd, Ni, and Zn in the sandy soil and of Ni and Zn in the brown forest soil at the highest
loading rate (Table 12.4). Notice that in Table 12.4 we used the more convenient unit of
Jlg/kg for the soil solution concentration, rather than mg/kg for the total concentrations
in Figures 12.1 and 12.2.
Chromium entered the liquid phase of the soils in negligible amounts, even at the
highest metal loading. Lead showed similar low soil solution concentrations as did Cr in
the brown forest soil, but in the sandy soil Pb concentrations were significantly higher
than Cr concentrations, for all loading rates. Compared to loading rates, release of Cd,
Ni, and Zn into the soil solution was much higher than of Cr and Pb, and increased
substantially at the higher metal application rates in the top 10 cm. Similar increases also
occurred in the originally uncontaminated soil layers directly below the application zone.
Concentrations of Cd, Ni and Zn in the soil solution were several orders of magnitude
higher at the 30 Land 100 L loading rates than in the control treatment, in the top 10 cm.
Of these three metals, Cd was found in much lower concentrations in the soil solution.
The observed low Cd concentrations are consistent with the relatively low application
rate of this metal as compared to those for the other elements. Release of the metals into
the soil solution at all metal application rates was significantly higher in the slightly
acidic sandy soil than in the brown forest soil (Table 12.4).
It is difficult to compare our data, which hold for soil solution concentrations in heavily
contaminated soils, with the highly variable literature data obtained under different ex-
perimental conditions and using a wide array of methods. Still, our results show similar
tendencies to those cited by Kabata-Pendias and Pendias (1992) or by Kabata-Pendias
and Adriano (1995); i.e., the relatively mobile metals, Cd, Ni, and Zn occur in a rela-
tively larger proportion in the solution phase than the less mobile Cr and Pb. These
findings cited were obtained for natural soil solutions separated by centrifugation from
different soils.
Soil solution concentrations expressed as percentage of the total potentially available
concentrations (SAs) indicated very low availability of Pb and especially of Cr in those
fractions of the soil's liquid phase which are directly accessible for plant uptake, since
only negligible amounts of these elements were released into the soil solution (Table
12.5). In contrast, Cd, Ni, and Zn were more readily available for uptake in both soils,
and the proportion of the directly plant-available amounts of these metals increased sharply
at the highest application rate (Figure 12.6); they were more than one order of magni-
tude higher than in the control treatment. Such increases were observed not only for the
264 Fate and Transport of Heavy Metals in the Vadose Zone
Table 12.4a. Concentrations of the Metals in the Soil Solution in Brown
Forest Soil a
( ~ g / k g dry soil)
Loading
Depth
Element (cm) Control 10 L 30L 100 L
0-10
b
0.4 1.2 90.2
Cd 10-15
b b b
11.6
15-20
b b b b
0-10 1.9 1.4 0.7 7.0
Cr 10-15 0.6 3.0
b
2.3
15-20
b
2.4 3.7
b
0-10 4.3 12.9 39.2 2600
Ni 10-15 2.1 4.9 6.8 225
15-20 5.6 3.5 11.3 13.7
0-10
b
5.9
b
7.8
Pb 10-15
b
3.3 6.1 11.5
15-20
b
6.5 2.8
b
0-10 50.8 120 219 28600
Zn 10-15 81.0 82.6 127 761
15-20 130 140 125 232
a Detection limits in mg/L: Cd - 0.005, Cr - 0.005, Pb - 0.05.
b Concentrations below detection limit.
Table 12.4b. Concentrations of the Metals in the Soil Solution in Slightly
Acidic Sandy Soil
( ~ g / k g dry soil)
Loading
Depth
Element (cm) Control 10 L 30L 100L
0-10 2.1 2.0 16.5 357
Cd 10-15 1.7 1.4 2.2 13.2
15-20 0.7 1.0 1.1 1.2
0-10 0.4 2.9 7.2 10.7
Cr 10-15 3.0 17.4 4.8 10.8
15-20 3.2 2.9 1.7 6.0
0-10 24.5 49.3 375 9197
Ni 10-15 27.0 23.6 39.0 623
15-20 17.0 12.8 22.5 29.4
0-10 18.2 10.4 14.1 154
Pb 10-15 20.1 30.1 16.9 33.6
15-20 34.3 14.0 9.7 28.1
0-10 283 500 3492 106200
Zn 10-15 395 273 342 6249
15-20 184 78 168 266
application zone of the metal-enriched sludge, but also for the originally uncontami-
nated 10-15 cmdepth interval (Figure 12.6). The breakthrough-like increases in the
relative amounts of the directly phytoavailable metal forms indicate a decrease in the
metal-buffering capacity of the soil at this extreme metal application rate. The calcula-
Distribution of Ecologically Significant Fractions of Selected Heavy Metals 265
- -- - ""-,-""-'-'"--'---. - -" --- - ~ - - - -------_ ... "-""----"
Table 12.5. Recovery of the Metals in the Liquid Phase of the Slightly Acidic Sandy Soil per
Thousand of the Total Potentially Available Amounts
Total
Amount of the Metals Recovered in the Liquid Phase
Potentially
Available
0-lOcm 10-15 cm 15-20 cm Total Amounts
Loadings mg/column mg/column %0 mg/column
Brown forest soil
Cd
a a a a
0 40
Cr 0.04
0.Q1
a
0.05 0.1 411
Control Ni 0.08 0.02 0.05 0.15 0.3 507
Pb
a a a a
0 516
Zn 0.96 0.76 1.23 2.95 1.2 2530
Cd 1.70 0.11 1.81 8.1 224
Cr 0.13 0.02 0.15 0.01 13100
100 L Ni 49.0 2.12 0.13 51.3 16.3 3140
Pb 0.15 0.11 0.26 0.02 14100
Zn 539 7.17 2.19 548 13.6 40200
Slightly acidic sandy soil
Cd 0.04 0.02 0.Q1 0.07 1.1 63
Cr 0.01 0.03 0.03 0.07 0.2 341
Control Ni 0.46 0.25 0.16 0.87 2.4 369
Pb 0.34 0.19 0.32 0.85 3.7 232
Zn 5.33 3.73 1.73 10.8 5.4 2000
Cd 6.74 0.12 0.01 6.87 33 211
Cr
a
0.10 0.06 0.16 0.02 10400
100 L Ni 173 5.88 0.28 179 73 2440
Pb 2.90 0.32 0.26 3.48 0.3 11200
Zn 2000 58.9 2.51 2060 60 34500
a Values below detection limit (see Table 12.4a).
tions revealed a higher availability and mobility of the elements in the liquid phase of the
slightly acidic sandy soil as compared to the brown forest soil (Table 12.4, Figure 12.6).
Such differences are shown not only by the metal concentrations but also by the dis-
tribution of the total amounts of the metals in the soil's liquid phase. Data for the control
treatment and for the highest contamination level are shown in Table 12.5. While distri-
bution of the metals among the liquid phases of the three sampled soil layers was fairly
uniform in the control treatments, at the highest contamination level most of the metals
were found in the top 10 cm, with slight increases of the Ni and Zn amounts between 10
and 15 cm.
The soil pH in our experiments decreased somewhat as the metal loading rate in-
creased (Figure 12.5). This decrease is likely caused by the acidity of the nitrate salts
being used in our study. The lower pH may have contributed to the increased mobility
and availability of the metals at the higher loadings. Also, the somewhat more acidic
character of the sandy soil may have been responsible for the higher availability of the
metals in this soil. This explanation is in agreement with literature data about the effect
of acidification on the availability and mobility of heavy metals (LJzsbersli et aI., 1991;
Marschner, 1991). The influence of soil pH on the mobility of trace metals depends also
266 Fate and Transport of Heavy Metals in the Vadose Zone
7.5
7
brown forest soil
~
6.5
I
I
c.
slightly acidic sandy soil
6
............ ..... - ... - ........................ ...
"-"'--" .. e
5.5
0 20 40 60 80 100
LOADING RATE (L)
Figure 12.5. Soil pH in the 0-1 0 em layer at the end of the experiment.
upon the geochemical properties of the metal: for example, mobility of Cd has been
classified as medium up to pH 6, and that of Ni and Zn up to pH 5, while Cr and Pb have
only weak mobility above pH 4.5 and 4, respectively (Kabata-Pendias and Pendias, 1992).
Thus, the differences among mobilities of various metals may have been influenced also
by the pH of the soil liquid phase.
Amounts of the elements entering the soil solution seemed to be rather low in com-
parison to literature data. For instance, the highest value of Zn in our experiments was
only 5.8% of the total potentially available amounts, while values up to 50% have been
reported in the literature (Coughtrey et aI., 1985). One likely reason for this disparity is
that literature data are generally based on extraction procedures that use chemical de-
sorption and wide soil:extractant ratios. Such methods should give much higher elemen-
tal concentrations than those which occur in the soil liquid phase at natural field soil
water contents.
The proportion of the phytoavailable metal fractions was by about one order of mag-
nitude higher when the calculations were based on the AAAc- EDT A extractable metal
concentrations (SAAE) as compared to the phytoavailabilities calculated from the soil
solution concentrations (SAs). In an extensive study comparing the availabilities of the
elements in the Hungarian soils (1013 samples), using different extractants, the
AAAc+ EDT A-available amounts of the metals included in our study were between 39
and 91 % of their 0.5 M HN0
3
-extractable amounts (Marth, 1990).
MOVEMENT
While SAs of Cd, Ni, and Zn showed a breakthrough-like phenomenon at the highest
loading rate (Figure 12.6), especially in the slightly acidic sandy soil, values of SAAE
increased linearly or to a smaller extent when the metal loading rates increased. A fur-
ther difference between the results of the two methods used to estimate the
phytoavailability of the metals was that while the directly available amounts of Cr and
Pb in the soil solution were similar to each other (in agreement with the similar applica-
tion rates), the AAAc-EDTA-extractable amounts of these elements were much differ-
Distribution of Ecologically Significant Fractions of Selected HP;nfU Metals 267
%
8
a. 0-10 em depth
6
4
2
0
Cd Ni Zn
% ~ 1
0
b. 10-15 em depth
1,-1]1
o
Cd Ni Zn
%
0:]
Cd Ni Zn
brown forest soil
slightly acidic sandy soil
Each set of bars represents the applied loadings from left to right:
control(L), lOL, 30L and 100L
Y axis: % of total potentially available concentration
(SA
s
= 100 C
s
I Cm , where Cs is the metal concentration measured in the plant-
available soil solution fraction, Cm is the total potentially available metal
concentration measured in the 2 mollL HND3 soil extract)
Figure 12.6. Soil availability factors (SA
s
) of the metals.
ent. The availability of added Pb, as estimated by the acidic ammonium acetate-EDT A
extraction, was comparable to the availability of Zn and Cd, and somewhat exceeded
that ofNi in both soils (Figure 12.7). In the brown forest soil it increased only slightly at
the higher loading rates, while in the slightly acidic sandy soil it followed more closely
the increasing metal pollution levels. Again, in this respect its behavior was similar to
those of Cd, Ni, and Zn.
Metal budgets calculated for the liquid phase of the upper 20 cm also show an in-
creased proportion of the directly plant-available forms of Cd, Ni, and Zn in per thou-
sand of their total potentially available amounts in these layers, at highly elevated metal
application rates (Table 12.5). Literature data on the possible leaching of the elements
show only small downward movement of the heavy metals, and mostly in coarse-tex-
tured, sandy, or gravelly soils. However, in many cases the applied metal concentrations
268 Fate and Transport of Heavy Metals in the Vadose Zone
%
80
70
60
50
40
30
20
10
0
Cd Cr Ni Pb
brown forest soil
slightly acidic sandy soil
Each set of bars represents the applied loadings from left to right:
control(L), IOL, 30L and IOOL
y axis: % of total potentially available concentration
Zn
(SAAE = 100 cAE/ Cm, where CAE is the metal concentration measured in in the
AAAc-EDTA extracts, Cm is the total potentially available metal concentration
measured in the 2 moVL HN0
3
soil extract)
Figure 12.7. Soil availability factors for the AAAe-EDTA extractions (SAAEl at 0-10 em depth.
were much smaller than the provocative overloading of 100 L in our experiments. Careful
evaluation of some data interpreted as being proof of no metal movement reveals such a
slight movement of the metals, which is in the range expectable from the relatively low
metal application rates, and which are sometimes interpreted as resulting from inadvert-
ent mixing of the soil layers (e.g., in the papers summarized in Dowdy and Yolk, 1983).
Downward movement of the metals involves the leaching of the water-soluble forms.
The movement of toxic metals was studied most often in the percolate water of lysimeter
experiments. Our experiments, conducted under natural soil moisture conditions, with
no addition of extra water that might have caused leaching (as shown by the TDR mea-
surements along the soil profile), showed that only a very small percentage of the pollut-
ing metals may enter the soil solution at field soil moisture contents, even at provocative
overloadings. Thus, the expectable concentrations of the metals in the liquid phase of
the deeper layers are small. Such small values are regarded often as proofs of no sub-
stantial movement.
Although in many experiments no metal movement was found in the deeper layers
below 60 cm, in some instances significant increases were reported. In a sludge-amended
clay loam soil under a forest vegetation and with a surface pH of 5, about 3% of the
applied Zn and 4-7% of the applied Cd was leached beyond 120 cm (cit. in Dowdy and
Yolk, 1983). In a lysimeter experiment on a sandy loam contaminated with sewage sludge
containing large amounts of Cd, Cr, Ni, Pb, and Zn, the mobility of these elements,
measured as their concentration in the percolate water, increased two times in sludge-
amended soil (Dudka and Chlopecka, 1990). Since in this chapter various metal inputs
Distribution of Ecologically Significant Fractions of Selected Heavy Metals 269
and outputs (g/ha) are also presented, the amount of the leached, water-soluble forms in
percentage of the total input can be calculated. These values are 5.4% for Cd, 0.003% for
Cr, and 0.17% for Ni and Zn. The Cd and Cr data are comparable to the relative avail-
abilities of these metals in the soil solution in our experiments (Figure 12.6).
SUMMARY
The biological and ecological effects of heavy metal pollution of soils depend not only
on the total amount of the contaminating metals, but to a great extent on their biologi-
cally and ecologically active, easily soluble and mobile fractions. The approach discussed
in this chapter focused on determining those fractions of the soil's metal content which
are directly and easily available for plant uptake (using two different methods to charac-
terize the phytoavailability), or which are potentially becoming available for the plants
during longer periods. Ratios of the directly available and potentially available metal
concentrations were used to estimate the bioavailability of the elements in soil columns
contaminated with sewage sludge spiked with (Cd + Cr + Ni + Pb + Zn)-nitrates.
The total potentially available amounts of the metals reflected well the application
rates, and did not increase or only slightly increased in the soil layers lying directly
below the application zone. Proportion of the directly plant-available amounts of Cd, Ni,
and Zn (measured in the soil solution considered as directly available for the plants)
increased sharply at the highest application rate, and such increases were observed not
only in the contaminated layer but also in the originally uncontaminated 10-15 cm depth
interval. We want to stress the ecological significance of the movement of these water-
soluble forms. Since the solubility of the heavy metals in soils has great significance in
their bioavailability and their migration (Kabata-Pendias and Pendias, 1992), relatively
small increases in solution metal concentrations may have an impact on the environ-
ment. The breakthrough-like increases in the relative amounts of the directly
phytoavailable metal forms might also indicate a decrease in the metal-buffering capac-
ity of the soil at this extreme metal application rate. In contrast, Cr and Pb entered the
liquid phase (i.e., were directly available for the plants) in negligible amounts, even at
the provocative overloadings. This is in agreement with the majority of the literature
data about the transport of these elements into the plants.
The AAAc + EDTA extraction which is a method used to estimate the plant-avail-
abilities of both macro- and micronutrients simultaneously, after a single extraction pro-
cedure, and which was shown to be appropriate also for several toxic elements, indicated
much greater availability of the metals than was estimated from the metal concentrations
measured in the soil's liquid phase. Estimated availability (i.e., extractability) of Cr was
less and that of Pb was higher than expected on the basis of the application rates. In the
AAAc + EDT A extraction procedure Pb behaved similarly to Cd, Ni, and Zn, which are
regarded as mobile elements in the soil.
Thus, the two methods used to assess the plant availability of the selected metals gave
different estimations of the phytoavailable proportions. Comparison of the estimated
phytoavailabilities with the actual metal uptake by plants should give support in favor of
one of these methods. However, further studies have to consider that plant concentra-
tions measured at a specific point during the vegetation period reflect the cumulative
uptake of the elements until that time, while soil solution concentrations pertain only to
270 Fate and Transport of Heavy Metals in the Vadose Zone
a specific situation (soil water content, temperature, etc.) at a specific moment and soil
extraction techniques cannot take into account the plant physiological processes.
ACKNOWLEDGMENTS
Supported by HNSRF, grant numbers T23221 and T23360.
REFERENCES
Alloway, B.J. Heavy Metal! in Soil!. Blackie and J. Wiley and Sons, Glasgow, 1990.
Andersson, A. On the determination of ecologically significant fractions of some heavy metals in
soils. SweoiJh J. Agric. &1. 6, pp. 19-25, 1976.
BCR No. 143. The certification of the contents of cadmium, copper, mercury, nickel, lead and
zinc in a sewage sludge amended soil. Report. Community Bureau of Reference - BCR.
Brussels, 1983.
Bujtas, K, J. Csillag, G. Partay, and A. Lukacs. Distribution of Selected Metals in a Soil-Plant
Experimental System Mter Application of Metal-Spiked Sewage Sludge, in Proc. XXVth AnnuaL
Meeting of ESNA, M.H. Gerzabek, Ed., Castelnuovo Fogliani (PiacenzaiItaly), Seibersdorf:
Osterreichisches Forschungszentrum Ges. m.b.H., Austria, 1995, pp. 99-105.
Campbell, D.J. and P.H.T. Beckett. The soil solution in a soil treated with digested sewage
sludge. J. SoiL Sci. 39, pp. 283-298, 1988.
Cassel, D.K and D.R. Nielsen. Field Capacity and Available Water Capacity, in Methood of Soil
AnaLYdiJ, Part I, 2nd ed., A. Klute, Ed., American Society of Agronomy, Madison, WI, 1986,
pp. 913-915.
CCRMP (Canadian Certified Reference Materials Project). Certificate of Analysis. Reference
Soil Sample SO-4. CANMET Report 79-3. Ottawa, Canada, 1979.
Chang, A.C., J.E. Warneke, A.L. Page, and L.J. Lund. Accumulation of heavy metals in sewage
sludge-treated soils. J. Environ. QuaL. 13, pp. 87-91, 1984.
Chang, A.C., T.C. Granato, and A.L. Page. A methodology for establishing phytotoxicity criteria
for chromium, copper, nickeL and zinc in agricultural land application of municipal sewage
sludge. J. Environ. QuaL. 21, pp. 521-536, 1992.
Coughtrey, P.J. and M.C. Thorne. RaownucLwe DiJtrilllltwn ano Trandport in TerrutriaL ano Aquatic
&odYdtefnd. A CriticaL Review of Data. Vol. 2., A.A. Balkema, Rotterdam, 1983.
Coughtrey, P.J., D. Jackson, and M.C. Thorne. RaownucLweDiJtrwutwn anO Trandport in TerredtriaL
ano Aquatic &odYdtefnd. A Compenoium 0/ Data. Vol. 6. A.A. Balkema, Rotterdam, Boston, 1985.
Csillag, J., T. T6th, and M. Redly. Relationships between soil solution composition and soil water
content of Hungarian salt-affected soils. Arw SoiL &1. RehabiLitatwn 9, pp. 245-260, 1995.
Dowdy, R. H. and V. V. Volk. Movement of Heavy Metals, in Proc. Symp. on ChemicaL MObility anO
Reactivity in SoiL SYdtefnd, SSSA Spec. PuM. No. n., Atlanta, 1981, D.W. Nelson et al., Eds., pp.
229-240, 1983.
Dudka,S. and A. Chlopecka. Effect of solid-phase speciation on metal mobility and phytoavailability
in sludge-amended soil. Water Air SoiL PoLLut. 51, pp. 153-160, 1990.
Dwyer, F. P. and D.P. Mellor, Eds. Chelating Agentd anO MetaL Chelated. Academic Press, N ew York
and London, 1964.
Homeyer, B., KO. Labenski, B. Meyer, and A. Thormann. Herstellung von Lysimetern mit
Boden in naturlicher Lagerung (Monolith-Lysimeter) als Durchlauf-, Unterdruck-oder
Grundwasserlysimeter. Z Pjlanzenernahr. Booenko. 136, pp. 242-245, 1973.
Hungarian Technical Directive. Land and Forest Applications of Waste Waters and Sewage
Sludges. MI-08-1735-1990 (in Hungarian).
Distribution of Ecologically Significant Fractions of Selected Heavy Metals 271
Jones, D.L. and A.C. Edwards. Effect of moisture content and preparation technique on the
composition of soil solution obtained by centrifugation. Commun. SoiL Sci. Plant AnaL. 24, pp.
171-186, 1993.
Juste, C. and M. Mench. Long-Term Application of Sewage Sludge and Its Effects on Metal
Uptake by Crops, in BiogeochemutryofTraceMetau, D.C. Adriano, Ed., Lewis Publishers, Boca
Raton, FL, 1992, pp. 159-193.
Kabata- Pendias, A. and D.C. Adriano. Trace Metals, in SoiLAmendment,} and EnvironmentaL QuaLity,
J.E. Rechcigl, Ed., CRC Press/Lewis Publishers, Boca Raton, FL, 1995, pp. 139-167.
Kabata-Pendias, A. and H. Pendias. Trace Element,} in SoiU and Plant,}, 2nd ed. CRC Press, Boca
Raton, FL, 1992.
Lakanen, E. and R. Ervi6. A comparison of eight extractants for the determination of plant
available micronutrients in soils. Acta Agr. Fenn. 123, pp. 223-232, 1971.
Legret, M., L. Divet, and C. Juste. Migration et speciation des metaux lourd dans un sol soumis
a des epandages de boues de station d'epuration a tres forte charge en Cd et Ni. Wat. RiAf. 22,
pp. 953-959, 1988.
Lindsay, W.L. ChemicaL EquiLibria in Soiu. John Wiley & Sons, New York, 1979.
E., E. Gjengedal, and E. Steinnes. Impact of Soil Acidification on the Mobility of Metals
in the Soil-Plant System, in Heavy Metau in the Environment. Trace Metau in the Environment. i.,
J.P. Vernet, Ed., Elsevier, Amsterdam, 1991, pp. 37-53.
Marschner, H. Plant-Soil Relationships: Acquisition of Mineral Nutrients by Roots from Soils,
in Plant Growth: Interactiond with Nutrition and Environment, J.R. Porter and D.W. Lawlor, Eds.,
Cambridge University Press, 1991, pp. 125-155.
Marth, P. Comparative Study on Soil Extractants. Postgraduate thesis, Agricultural University
of G6d6ll6, 1990, (manuscript in Hungarian).
McBride, M.B. Toxic metal accumulation from agricultural use of sludge: Are USEPA regula-
tions protective? J. Environ. QuaL. 24, pp. 5-18, 1995.
McGrath, S.P., A.C. Chang, A.L. Page, and E. Witter. Land application of sewage sludge:
Scientific perspectives of heavy metal loading limits in Europe and the United States. Env.
Review,} 2, pp. 1-11, 1994.
Molnar, E., T. Nemeth, and o. Palmai. Problems of Heavy Metal Pollution in Hungary- "State-
of-the-Art," in Heavy Metau. Pr06LemJ and SoLutiond, W. Salomons et al., Eds., Springer, Berlin,
1995, pp. 323--344.
Mullins, G.L. and L.E. Sommers. Characterization of cadmium and zinc in four soils treated with
sewage sludge, J. Environ. QuaL. 15, pp. 382-387, 1986.
Nemeth, T., G. Partay, I. Buzas, and H. Gy. Mihalyne. Preparation of undisturbed soil monoliths.
Agrolcemia iJ Talajtan 40, pp. 236-242, 1991, (in Hungarian).
Nemeth, T., G. Partay, K. Bujtas, and A. Lukacs. Application of Quadrupole Mass Spectrometry
to Assess Effects of Sewage Sludge on Gas Composition in Undisturbed Soil Columns, in
Biogeochemutry of Trace Element,}, D.C. Adriano, Ch. Zueng-Sang, and Y. Shang-Shyng, Eds.,
Science and Technology Letters, Environ. Geochem. HeaLth 16, pp. 141-151, 1994.
Partay G., A. Lukacs, and T. Nemeth. Soil monolith studies with heavy-metal containing sewage
sludge. Agrolcemia iJ Talajtan 43, pp. 211-221, 1994.
Petruzzelli, G. Recycling wastes in agriculture: Heavy metal bioavailability. Agric. Eco,}Y,}temJ
Environ. 27, pp. 493-503, 1989.
Sillanpaa, M. Micronutrients and the Nutrient Status of Soils. FAO Soiu BuLletin 48, 1982.
Sillanpaa, M. and H. Jansson. Status of cadmium, lead, cobalt and selenium in soils and plants
of thirty countries. FAO SoiU BuLletin 65, 1992.
USEPA (U.S. Environmental Protection Agency). Standards for the disposal of sewage sludge;
Proposed Rules 40 CFR Parts 257 and 503. Fed. Regut. 54, pp. 5746-5902, 1989.
Zabowski, D. and F.C. Ugolini. Lysimeter and centrifuge soil solutions: Seasonal differences
between methods. SoiL Sci. Soc. Am. J. 54, pp. 1130-1135, 1990.
CHAPTER 13
Heavy Metal Contamination in Industrial
Areas and Old Deserted Sites:
Investigation, Monitoring, Evaluation,
and Remedial Concepts
Irena Twardowska, S. Schulte-Hostede, and Antonius A.F. Kettrup
INTRODUCTION
The historical development of industrialization in Germany and Middle European
countries based on mining and metallurgy resulted in the creation of thickly populated
"hot spots" impacted by high and long-term emission of heavy metals to the environ-
ment. One of the major sources of non point contamination of soils and water by heavy
metals is dry and wet deposition from industrial stacks. To date, despite the attention of
governmental agencies, industry, and environmentalists, as well as of public concern
focused on this source, the understanding of the dimension of the human risk potential
from soil contamination originating from long-term stack emission is still limited.
Due to the long-term effect and frequently occurring unfavorable transformations of
anthropogenic metal-bearing waste properties in time, old abandoned industrial sites
may display high contamination potential in a postclosure period. Because of the lack of
adequate knowledge, environmental protection was not considered when these facilities
were sited and operated.
The characteristics and utilization of contaminated sites, kind of emission, character
of environmental impact and target-oriented protection objectives determine different
criteria and methods of site investigation, evaluation, and remedial concepts. The objec-
tive of this chapter is to present a different, site-specific approach to the evaluation of
heavy metal contamination in areas oflong-term anthropopression, with a special regard
to evaluation of mobile/mobilizable forms of heavy metals in soil and the vadose zone
matrix. Methods of evaluating human risk potential and selection of appropriate reme-
dial options for the site-specific cases, comprising technical measures and the use of
adequate extractants or adsorbents were also discussed.
274 Fate and Transport of Heavy Metals in the Vadose Zone
The presented approach was exemplified in case studies on three sites impacted by
emissions of different character and potential: long-term stack emission, and metal emis-
sion from a large-area deserted industrial site. Localization of the surveyed sites is pre-
sented in Figure 13.l.
The scope of applied investigation procedures was highly site-specific, depending on
the kind of heavy metal emission, extent of risk, and endangered objectives to be pro-
tecyd (e.g., population, agricultural area, groundwater resources). To evaluate actual
eent of anthropogenic impact and pathways of investigated heavy metals in the envi-
rdnment, a broad range of preliminary studies and investigations had to be undertaken
in each case. The general scope of the preliminary studies consisted of collecting avail-
able archival data required for the standard procedure of environmental impact assess-
ment (EIA), and next filling the information gaps and uncertainties with new
investigations, with particular regard to the aim of the studies to be conducted in each
site. As a part of a comprehensive contamination assessment and risk evaluation in the
investigated sites, the following parameters were considered: site localization, area, popu-
lation, land use, natural conditions (morphology, meteorology, hydrology, hydrogeology),
characteristics of soils and the vadose zone matrices as well as the source of heavy metal
emission (type, scope, extent, duration).
IMPACT OF LONG-TERM STACK EMISSION
The investigation and assessment procedures for the areas impacted by the long-term
metal emission from stacks were illustrated in case studies on soil contamination in the
areas of Sendzimir steelworks in Poland (Site I) and Irena Glasswork (Site II). In both
sites, surface soil and the upper part of the vadose zone profile (subsoil) enrichment with
heavy metals from dry and wet deposition of heavy metals from long-term industrial
stack emissions was the main objective of the presented studies.
Site Characteristics
Site I: Nowa Huta nlCracow, Area Adjacent to the Sendzimir Steelwork
Complex, Po/and (Figures 13.1, 13.2A)
One of the 27 "hot spot" areas in Poland, classified as ecological emergency areas
(Chief Statistical Office, 1995) is the Cracow district, mainly due to the extent of atmo-
spheric pollution. Some industrial plants are subject to extreme controversy due to their
unfortunate location. One of these plants, the 40-year-old Sendzimir Steelworks, is lo-
cated in the highly developed agricultural area near Cracow, a city of great architectural
and historical value. The emission of particulate and gases from Sendzimir steelworks
have been considered to be one of the major sources of contamination by heavy metals
and other pollutants in Cracow. The objective of the study was to assess the steelworks
as a source of soil pollution by means of evaluating the concentrations of heavy metals in
the soil in its vicinity.
Morphology Vadose Zone Matrix and Soil Characteristics
The site of the total area 6219.3 ha (Figure 13.2A) represents an area of long-term
anthropogenic impact. Relief is typical for loess areas and consists of a number of table-
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 275
.----.... - ~ \
\
POLAND \
\
I
(""
-.,
I
I
\
\
Site I \
Nowa Huta n/Cracow )
Sendzimir Steelwork;'
9-/ j
..r (
v .... ~ - , \
Figure 13.1. Location of surveyed sites I, II (Poland) and III (FRG):
Site I: Nowa Huta n/Cracow, area adjacent to Sendzimir steelworks, Poland;
Site II: area adjacent to Irena Glasswork n/lnowroclaw, Poland;
Site III: Marktredwitz urban area with chemical plant site, North Bavaria, FRG.
lands, crosscut by valleys of small streams. The soil and the vadose zone matrix comprise
Quaternary loess sediments from Cracow glaciation, 8-12 m thick. Black earth, mostly
degraded, in a minor rate diluvial, accounts for 65.7% of soils; 27.7% of the area are
brown soils. The residual 6.6% are mainly fen soils (5.1 %). These are almost entirely
rich heavy soils (96.2%), composed predominantly of loess, with some clay-like loess
formations, of high or medium humus content. The pH value is mainly neutral, except
for soils located SE, S, and NW of the site, which are acidic (Table 13.1).
The utilization of the land shows its agricultural characteristics and is comprised of:
arable land, mainly horticultural, 83.9%; grassland, 10.2%; forests, 0.5%; disused lands,
3.2%; railroad area, 1.9%; compact settlements, 0.3% of the total area.
SOUfce of Anthropogenic Contamination of Soil by Metals
Since the late fifties, the area has been affected by the emission from Sendzimir Steel-
works complex sited at its SW border, in the direction of dominating winds, blowing
from W, NW, and SW (Figure 13.2B-I). The Sendzimir Steelworks is the major source
of particulate and gaseous emission in the area of the surveyed site and Nowa Huta
satellite district of Cracow. It is considered to be also one of the major sources of air
contamination with dust and heavy metals in the historic downtown of Cracow. Accord-
ing to air monitoring data (Chief Statistical Office, 1993; 1995) average mean annual
wet and dry particulate deposition in 1994 in Nowa Huta area ranged from 68 to 159 mg
km -2 a-I, mean 117 mg km -2 a-I. In Cracow downtown the particulate deposition ranged
from 41 to 105 mg km-
2
a-I, mean 70 mg km-
2
a-I. According to the same source, the
average annual concentration of trace metals in air in both districts represented an or-
276 Fate and Transport of Heavy Metals in the Vadose Zone
Figure 13.2A-C. Site I: Nowa Huta n/Cracow, area adjacent to the Sendzimir Steelworks, Poland.
(A) General map. Soil sampling paints {op}, soil profiles (* P) and spatial distribution of heavy metals
(mg kg-') in surface soil: (B) Fe, (C) Mn. Kriging estimates: linear model, angle 15.
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 277
Zn
P60 1'61

P70 P69

P42 P41 P23 P18
. . . .
...........
P10
. '"
P8 '"

o

Pb
4 6 8 10 12 14 km
10 12 14 km
Figure 13.2D-F. Site I: Nowa Huta n/Cracow, area adjacent to the Sendzimir steelworks, Poland.
Soil sampling points (P), soil profiles (OP) and spatial distribution of heavy metals (mg kg-
1
) in
surface soil: (D) Zn, (E) Pb, (F) Cr. Kriging estimates: linear model, angle 15.
278 Fate and Transport of Heavy Metals in the Vadose Zone

6 8 10 12 14
km
o 10 12 14

o 6 8 10 12 14
Figure 13.2G-1. Site I: Nowa Huta n/Cracow, area adjacent to the Sendzimir steelworks, Poland.
Soil sampling points (P), soil profiles (* P) and spatial distribution of heavy metals (mg kg-
1
) in
surface soil: (G) Ni, (H) Cu, (I) Cd. Kriging estimates: linear model, angle 15.
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 279
Table 13.1. Physicochemical Properties of Soils in Site I (Nowa Huta n/Cracow, Area Adjacent to
Sendzimir Steelworks, Poland)
Mean
Range Percentage of the Total Area of Soils (%)
Content of fraction <0.02 J.lm
44 0-10 11-20 21-35 >35
33-58 very light light mean heavy
0 0 3.8 96.2
pH value in 1 M KCI
6.3 <4.5 4.6-5.5 5.6-6.5 6.6-7.2 >7.2
3.6-7.8 very acid acid slightly acidic neutral alkaline
7.6 22.8 15.2 44.3 10.1
Humus contents (%)
2.76 <1.00 1.01-2.00 2.01-3.50 >3.51
1.21-8.44 low mean high very high
0 16.5 77.2 6.3
Hydrolytic acidity (Hh, ceq kg-' of soil, dry weight)
2.06 <1.5 1.51-3.00 3.01-4.50 >4.51
0.52-6.98 low mean high very high
58.2 17.7 11.4 12.7
Base saturation of soils (V) (%)
86.0 <25 26-50 51-75 >75
34.8-98.6 very low low mean high
0 1.3 15.2 83.5
der: Zn > Pb Cu > Mn Ni > Cr > Cd (Table 13.2). In downtown, concentrations of
Zn, Mn, and Cd appeared to be higher than in Nowa Huta close to the Sendzimir Steel-
works, which can be explained by the contribution of other sources of air pollution.
The direct measurements of wet and dry deposition and the concentration of Zn, Pb,
and Cd in the settled dust in the area taken during 1992 (preceding the survey of the soils
for trace metals), recorded dust deposition in the sampling point 1.4 km SW from the
Steelworks to be 2 to 4 times higher than in other parts of Cracow (Table 13.3). This
proves Steelwork to be a serious source of air contamination in the Cracow area. Con-
centrations of Zn and Pb in the deposition in Site I were comparable with those recorded
in western suburbs and the city center of Cracow, downtown being the highest. These
data are in a good agreement with the routine data that were cited earlier
(Chief Statistical Office, 1993, 1995). It should e emphasized that despite similar con-
centrations, the load of metals in Site I adjacent to endzimir Steelworks is the highest,
due to high dust deposition (Table 13.3). Considering that just one defined strong source
of non point contamination occurs in the area (high and low emission from the Steel-
works), the site was selected as appropriate for the evaluation of the contamination po-
tential of emission from the steelworks. Objectives of a particular interest included: barrier
capacity of soils, forms of binding, as well as accumulation and mobility of metals of
anthropogenic origin in the vadose zone compared to the natural background.
In accordance with the kind of the emission, the metals of concern were Fe, Mn, Zn,
Pb, Ni, Cu, Cr, and Cd (Figure 13.2B,C,D,E.F,G,H,I). This order of metals aligned
with their descending concentration ranges in soils of Site 1. In Figure 13.2B-I, the
280 Fate and Transport of Heavy Metals in the Vadose Zone
Table 13.2. Mean Annual Particulate Deposition (mg km-
2
a-') and Concentration of Metals
(ng m-
3
) in Nowa Huta (Sendzimir Steelworks Area) and Cracow Downtown
a
Sampling Point. Direction and Distance
from the Steelworks
Nowa Huta, Cracow Downtown, podg6rze,
Number
b
W 1.0 km Wl0km SW10km
Contaminant N n 1992 1994 1992 1994 1992 1994
Wet and dry 12 144 106 117 123 70 77 81
particulate
deposition
(mg km-
2
a-')
Metal concentrations (ng m-
3
)
Zinc, Zn 1 12 540 325 340 542 310 363
Lead, Pb 1 12 138 160 98 160 146
Cadmium, Cd 1 12 0.0 2 0.0 3 0.0 2
Copper, Cu 1 12 30 30 50 25 20 28
Manganese, Mn 1 12 30 16 20 21 20 20
Nickel, Ni 1 12 10 14 10 8 10 4
Chromium, Cr 1 12 0.0 6 0.0 4 0.0 4
a Chief Statistical Office, 1993, 1995.
b N - number of sampling points; n - number of measurements.
Table 13.3. Wet and Dry Deposition of Dust in Nowa Huta Area (Site I) Compared to Other
Districts of Cracow (1992)
Sampling Points, Directions, and Distance from the Steelworks
Cracow
Mogila Mistrzejowice Downtown, Bielany KObierzyn
Contaminant Unit SW 1.4 km NW5km W10km W 18km SW 16 km
Annual wet and dry deposition
mg km-
2
164.2 89.1 86.0 46.6 42.9
Metal concentrations in dust deposition
mg kg-
1
Zinc, Zn 2237 1960 3104 2254 1967
Lead, Pb 449 369 580 438 502
Cadmium, Cd 50.2 26.5 50.2 16.5 11.3
Metal deposition
kg km-
2
Zinc, Zn 367.3 174.8 267.0 105.0 84.3
Lead, Pb 73.7 32.9 49.9 20.4 21.5
Cadmium, Cd 7.2 2.4 4.3 0.77 0.48
spatial distribution of ~ t a l s is presented. Figures 13.3 to 13.6 illustrate the vertical
distribution and chemical fractionation according to the binding strength of Zn (Figure
13.3), Cd (Figure 13.4), eu (Figure 13.5), and Pb (Figure 13.6). Figure 13.7 displays
the structure of these metals enrichment in the mobile/mobilizable fractions that are of
particular importance for risk assessment.
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 281
- --
Zn
Profile R
100% 0 50 100
mg/kg
150
80%
0-30
250
60%
mglkg
20-30
40%
30-50
20%
0%
50-70
0

0
E

...

" 0 >70 em
N
'"
...

Profile W
100% 0 50 100 150
80%
0-30
60%
40%
30-70
20%
70-100
0%

0
8
g
...
0

E
100-130 em
'"
...
"
Profile I
100% 0 50 100
mg/kg
150
90%
80%
70%
0-30
60%
50%
40%
30-70
30%
20%
10%
70-100
Figure 13.3. Vertical distribution and chemical fractionation of Zn in soil profiles (0-130 cm) in Site
I; Metal binding fractions: FO+ 1 (EXC): pore solution and exchangeable; F2 (CARB): carbonate
associated; F3+4 (EMRO): easily and moderately reducible Mn- and amorphous Fe-oxides; F5(OM):
oxidizable, associated with organic matter and sulfides; F6(R): residual (lithogenic matter).
Site II: Irena Glasswork Inowroclaw, Poland (Figures 13.1, 13.8)
Lead is one of the most and widespread anthropogenic pollutants. Envi-
ronmental contamination with lead is associated with processing of zinc and lead ores,
combustion of leaded gasoline in car engines, production of accumulators, paints, etc. Of
282 Fate and Transport of Heavy Metals in the Vadose Zone
Cd
Profile R
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Profile W
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Profile I
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
IFO+11

g

o 0,1 0,2 0,3
mg/kg
0-30
20-30
30-50
50-70
>70 em Pllml
30-70
70-100
100-130 em
o 0,1
0,2 mglkg 0,3
0-30
30-70
70-100
100-130 em
Figure 13.4. Vertical distribution and chemical fractionation of Cd in soil profiles (0-130 cm) in Site
I; Metal binding fractions: FO+ 1 (EXC): pore solution and exchangeable; F2 (CARS): carbonate
associated; F3+4 (EMRO): easily and moderately reducible Mn- and amorphous Fe-oxides; F5(OM):
oxidizable, associated with organic matter and sulfides; F6(R): residual (lithogenic matter).
various industries, lead crystal glassworks are a proven source of lead contamination.
The scope of this study comprises evaluating the extent of soil enrichment by lead in the
vicinity of Irena Glasswork.
Cu
Profile R
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Profile W
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Profile I
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
I]FO+1 I
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 283
o 5 10 15 20
mg/kg
________________
_
3O-5OleI ___ _
""""1I
>70cm __
o 5 10 15 20
mg/kg
30-70


20
30-70


1IIIIIIIIIIIIIF2 .. I,--FS-l
Figure 13.5. Vertical distribution and chemical fractionation of Cu in soil profiles (0-130 cm) in Site
I; Metal binding fractions: FO+ 1 (EXC): pore solution and exchangeable; F2 (CARB): carbonate
associated; F3+4 (EMRO): easily and moderately reducible Mn- and amorphous Fe-oxides; F5(OM):
oxidizable, associated with organic matter and sulfides; F6(R): residual (lithogenic matter).
Locotion, Soil C/w(acteristics, and Land Use
Site II is located in Central Poland in the flat area adjacent to Irena Glasswork, in the
NW part of Inowroclaw (Figure 13.8). Geomorphologically it belongs to plains of the
moraine upland originating from the Baltic glaciation phase. The area lies at the border
284 Fate and Transport of Heavy Metals in the Vadose Zone
Pb
Profile R
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Profile W
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Profile I
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
IFO+11
1IIIIIIIIIIIIIF2
o 5 10 15 20
________________
40
20-30
mg/kg


>70em




o 5 10 15 20

30-70

100-130 em
fFAjF3+41
IBIL..-
FS
----,
"L.....
F6
---,
Figure 13.6. Vertical distribution and chemical fractionation of Pb in soil profiles (0-1 30 cm) in Site
I; Metal binding fractions: FO+ 1 (EXC): pore solution and exchangeable; F2 (CARB): carbonate
associated; F3+4 (EMRO): easily and moderately reducible Mn- and amorphous Fe-oxides; F5(OM):
oxidizable, associated with organic matter and sulfides; F6(R): residual (lithogenic matter).
of a fallow gley podzol on the vadose zone matrix composed of sands and brown soils
formed from loess and loessial formations. This results in variable content of clay frac-
tion, ranging from 3.8 to 35.0 wt % and sand fraction occurring within the range from
78.2 to .38.8 wt %. The agricultural land represents mainly wheat complex. Dominating
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 285
Zn
100% 100% r-' 100%
80% 80% 80%
60% 60% 60%
40%
40% 40%
20%
20% 20%
0% I- ~ '--r- ~ ~
0% ~ 0%
g'-,
0 0
8
0 0 0
8
0 0 0
E
'"
....
'" '"
....
Profile R ~
"I
....
0
"
Profile W
O 0
;;;
~ Ii
Profile I 0
0
;;;
~ Ii
g 0
'" '"
"'
.... .... ....

Cd
100% r r e-
7'
100% .,--, ~
c
r
100% e-
re
~
r-
....
80%
i
80%

I
80%
; ..
60% I
60% 60%
40%
40% 40%
20%
20% 20%
0% -+- 4-
E ~
0% -+- 4- 4- ~ 0% -+- -+- 4-
Lj
g
~ 8
0
g
~ 8
g
o b1 :::!
'"
Profile R ~ g fi: "
ProfileW
o 0
~ ~ Ii
Profile I 0
g
~
- E
0
'" ~ "
....

Pb
100%
r
7'
'"'
100% ;- 100%
:>
80% 80% 80%
60% 60% 60%
40%
40% 40%
20%
20% 20%
0% 4- -+- -+- ~
g ~ 8 g
C!i '" - - E
ProfileW Pi g ~ "
0% g -+- ~ 4- 8 -+-g --l
Profile I C!i g ~ ~ Ii
Cu
100% 100% 100%
80% 80% 80%
60% 60% 60%
40%
40% 40%
20%
20% 20%
0% -+-4-4-y
0% '--r- ~ '--, 0%
g 4- ~ Lr- 8 -+-g
--l
0 0
8
0
g ~ ~ E
~
....
'"
Profile R ~ g fi: ~
g
;;; - E
Profile I 0
g ;;;
~ Ii
Profile W
~ "

.... ....
I jFO+1 I
1IIIIIIIIIIIIIF2
Figure 13.7. Partition of the mobile fraction of trace metals (Zn, Cd, Pb and Cu) in soil profiles (0-
130 cm) in Site I; Metal binding mobile fractions: FO+ 1 (EXC): pore solution and exchangeable; F2
(CARB): carbonate associated; F3+4 (EMRO): easily and moderately reducible Mn- and amorphous
Fe-oxides; F5(OM): oxidizable, associated with organic matter and sulfides.
286 Fate and Transport of Heavy Metals in the Vadose Zone
0-.... ___ km
c
e2l 0
025
E
Figure 13.8. General plan of Site II: area adjacent to Irena Glasswork n/lnowroclaw, Poland).
P - Irena Glasswork; 1 - sampling points along the intersections A, B, C, 0; 2 - compact residential
area; 3 - railways; 4 - motorways.
winds blow from W, NW, and SW direction, which means that due to unfavorable loca-
tion, mainly the compact residential area of Inowroclaw City is affected with the emis-
sion from the Glasswork. The surveyed site comprised an area of 5000-m radius in NE,
NW, W, and partially SW directions from Irena Glasswork, used for intensive agricul-
tural production (mainly arable land and orchards). In the NE, NW, and W-WS direc-
tions the area is crosscut with parallel railroads and motorways.
of Antflropogenic Contarnination of Soil by Lead
The major source of soil contamination by Pb in Site II is an emission from the stacks
of Irena Glasswork. The glasswork has been in operation already for several decades,
but as a source of Pb emission it is considered since 1976, when the production of lead
crystal glass started. Pb has been emitted to the atmosphere with particulates, mainly in
the form of oxides. Another source of Pb contamination occurring in the area is leaded
petrol combustion in motor vehicles. The survey of soil contamination by Pb in Site II
was focused on the evaluation of the glasswork impact on the extent and character of
undisturbed soil contamination in this area. The results of the survey are illustrated in
Figures 13.9 and 13.10, which present spatial and vertical distribution of Pb in the un-
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 287
o -2.5cm
2.5 -5.0 em
5.0 -10.0 em
10.0 -15.0 em
15.0 - 20.0 em
35.0 - 40.0 em
Figure 13.9. Site II (area adjacent to Irena Glasswork n/lnowroclaw). Spatial and vertical distribu-
tion of lead in an undisturbed surface soil layer (0-40 cm).
disturbed surface soil layer 0--40 em, as well as chemical fractionation ofPb accumulated
in soil vs. distance from the source of emission (Glasswork stacks).
Soil Enrichment with Heavy Metals in the Areas Impacted by it
long-Term Stack Emission
Screening Survey and Methods
An extent of surface soil and the upper part of the vadose zone (subsoil) matrix con-
tamination by heavy metals in Site I (Nowa Huta, area adjacent to Sendzimir Steel-
288 Fate and Transport of Heavy Metals in the Vadose Zone
A-1a
Total
Total
8-8a
Total
C-15a
Mobile
Fractions
Mobile
Fractions
Mobile
Fractions
100% ,--"r=',.,..,.=----.,.........,.,.",...---,
80%
60%
40%
20%
0%
0-22a
Total Mobile
Fractions
A-1a
100 200 300 400 500 600
0-2,5
11111111: .::.: ...... : - : ~
2,5-5,0 1------'
mg/kg
5,0-10,0 1--_-'
10,0-15,0 1--_ ....
15,0-20,0
35,0-40,0 em
8-8a
100 200 300 400 500 600
0-2,5 11111: ..... : ..: .. ~
mg/k
2,5-5,0 1--__ --'
5,0-10,0 1--_-'
10,0-15,0 1--_ ....
35,0-40,0 em
C-15a

100 200 300 400 500 600
0-2,5
mg/kg
2,5-5,0
5,0-10,0
10,0-15,0
15,0-20,0
35,0-40,0 em
0-22a

100 200 300 400 500 600
0-2,5
mg/kg
2,5-5,0
5,0-10,0
10,0-15,0
15,0-20,0
35,0-40,0 em
Figure 13.10. Vertical distribution and chemical fractionation of Pb in an undisturbed surface soil
layer (0-40 cm), Site II. Metal binding fractions (McLaren and Crawford, 1973): fO+ 1 (EXC): pore
solution and exchangeable; f2(CARB): specifically sorbed, carbonate-associated; f5(OM): oxidiz-
able, associated with organic matter; f3+4(EMRO) free Mn- and Fe-oxides.
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 289
works in Poland) (Figures 13.1, 13.2A) resulting from the impact of the Steelworks
stack emission was evaluated on the basis of metal accumulation and spatial distribution
in surface soil layer (Table 13.4, Figure 13.2B-I). Barrier capacity of soil, binding phases,
and metal mobility in soiVvadose zone matrix were also taken into consideration (Fig-
ures 13.3 to 13.7). The evaluation was based on the random survey of soil for trace
metals in node-points of the network of SOxSO m squares carried out according to the
EPA guidelines (Barth and Mason, 1984). The survey comprised sampling: (i) surface
soil layer in the basic 63 node-points; (ii) surface soil layer in 8 points (one of every 10)
taken in duplicate in close proximity to the basic node-points to estimate the variability
among sampling units; (iii) soil and subsoil matrix in four layers up to 30 cm thick along
the upper part of the vertical profile of the vadose zone up to 130 cm deep, in three points
selected at different distances and directions from the Steelworks with respect to the
wind rose: Wadow 64-67 (Profile W), Ruszcza 4S-48 (Profile R), and Igolomia 75-78
(Profile I) (Figures 13.2B-I, 13.3 to 13.7).
In Site II (Irena Glasswork) (Figures 13.1, 13.8), the undisturbed soil enrichment
with Pb against the distance from the source of emission and the depth of soil layer was
assessed (Figure 13.9), with a special regard to the vertical migration of Pb into undis-
turbed soil layers (Figures 13.9, 13.10). The screening survey comprised soil sampling
for lead in 2S sampling points along the 4 intersections (A, B, C, D) in the distance from
so to SOOO m from the Irena Glasswork, consecutively in 50, 100, 2S0, SOO, 1000, 2S00,
and 5000 m from the source ofPb emission (Glasswork stacks). In the investigated area,
other sources of Pb than the Glasswork occurred: Pb emission from motor vehicles,
pump stations distributing leaded gasoline, and other activities emitting Pb in the com-
pact settlement area of Inowroclaw city, located in the E, SE, and S direction from the
Glasswork. The disturbing effect of cultivation on vertical distribution of Pb in the soil
layer in the agricultural areas located in SW, W, NW, and N directions from the Glass-
work also should have been considered. To exclude effect of these factors that could
influence spatial and vertical Pb distribution in the soil layer, it was desisted from the
random sampling procedure and from soil sampling within the compacted settlements of
Inowroclaw city located in the area most affected by emission from the Irena Glasswork.
The samples were thus taken along the intersections laid out in the barren undisturbed
land in agricultural area, in the NE, NW, W, and SWW directions from the Glasswork,
not closer than SO-100 m from the motorways and buildings. In each point,S consecu-
tive soil layers from 2.5 to S.O cm thick (the uppermost two 2.5 cm thick, the rest 5.0 cm
thick), up to the depth of 20 cm, along with the layer 3S-40 cm were sampled and ana-
lyzed for Pb.
In conformity with the sampling program and a scope of the studies, the material
analyzed for trace metals comprised soil and subsoil matrix. Acid-digested (ASTM, D
5198-92, 1992) soil and subsoil matrix samples were analyzed for the total metal content
by standard methods using AAS and ICP-AES techniques (AAS Perkin Elmer 1100 B
and ICP Perkin Elmer Plasma 40).
Binding strength of metals in the selected soiVvadose zone matrix samples from Site I
was evaluated using sequential extraction (Tessier et al., 1979, modified by Kersten and
Forstner, 1986). Sequential extraction scheme partitions off exchangeable FO+1 (EXC) ,
carbonate-bound F2(CARB), easily and moderately reducible Mn-oxides and amor-
phous Fe-oxides F3+4(EMRO), oxidizable sulfidic/organic FS(OM) and residual frac-
N
\.0
o
."
III
....
fI)
III
::::I
C.
~
III
::::I
VI
"0
o
:4-
a
::I:
re
Table 13.4. Concentrations of Heavy Metals in Soils of Site I (Nowa Huta n/Cracow) Compared to the Geochemical Background in Unpolluted ~
Areas s:
Site
Fe (%) Mn
Nowa Huta from-to 1.25-2.50 254-1160
(Site I) mean 1.61 485
Geochemical from-to 0.80-2.78 380-700
background" mean 1.20 560
" Kabata Pendias and Pendias, 1992.
Heavy Metals Concentration (mg kg-l)
Zn Cu Pb Ni
32-670 8.4-22.2 10-42 11.0-22.8
120 13.8 27 16.7
30-360 4.0-53.0 19-49 10.0-104
65 19 25 25
Cd Cr
0.1-1.1 14.4-40.8
0.45 24.2
0.08-0.96 14.0-80.0
0.38 38
~
iii
:i"
-::r
fI)
<:
III
c.
~
fI)
N
o
::::I
fI)
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 291
tions F6(R), of subsequently increasing binding strength. For the selected soil samples
from Site II, that was investigated several years earlier than other sites, the 5-step se-
quential procedure by McLaren and Crawford (1973) was used, which partitions off
exchangeable (fO+1), specifically sorbed, mainly carbonate-bound (2), organic (f5), free
oxides (f3+4), and residual (f6) fractions. In general, it well corresponds with a chemical
fractionation of metals derived from the procedure by Tessier et al. (1979) modified by
Kersten and Forstner (1986).
Metal Distribution in Soil VS. the Duration and Extent oj Emission
Survey of soils for heavy metals in Site I showed distinct impact of a stack emission
from the Steelworks on the trace metal distribution in the undisturbed surface soil layer
(0 to 40 cm) (Figure 13.2B-I). The maps of equipotential lines generated by SURFER
5.01 Mapping System for the surveyed metals display clear correlation with the emission
from the Steelworks for Zn, Cr, Fe, Cd, and Pb. At the same time, weak impact on Mn
and Cu concentrations and practically no effect on Ni content in the area adjacent to the
Steelworks in the direction of dominating winds was observed (SW area of Site I in the
close proximity to the Steelworks). Zn in the whole impacted area occurred in mean
concentrations markedly (about twofold) higher with respect to the mean geochemical
background for the similar soils from the unpolluted areas (Table 13.4) (Kabata-Pendias
and Pendias, 1992). Besides, Zn was also the only element showing a uniform back-
ground concentration range of 50 to 100 mg kg-
1
in the nonimpacted area and the dis-
tinct borders of the impact of the Steelworks emission, while no other competitive sources
of emission occurred (Figure 13.2D). The Steelworks were apparently responsible for
the elevated concentrations of Fe in the soil compared to the geochemical background
(Figure 13.2B), though another area in SE of the site (Igolomia) also displays concen-
tration anomaly. Of other 8 surveyed metals, maximum and mean contents of Cd oc-
curred somewhat in excess with respect to the background concentrations in unpolluted
soils, both in the impacted area of the Steelworks (SW) and in the Igolomia site (SE)
(Figure 13.21). The Steelworks was also a sole source of a substantial (two- to threefold)
enrichment of the surface soil layer in the adjacent area by Cr (Figure 13.2F). The back-
ground concentrations of Cr in the area were, though, very low and both the maximum
and mean concentrations of this element were placed within the concentration range
occurring in the unpolluted soils (Table 13.4). The analysis of the equipotential map of
heavy metal distribution in the surface soils of Site I gives evidence of occurrence of
other sources of metal enrichment in soils, comparable or even of a higher extent than
the Steelworks. Such area of significantly elevated concentrations of Fe, Mn, Pb, Ni,
Cu, and Cd was identified in Igolomia, SE of Site I (Figure 13.2B,C,E,G,H,I). The no
longer existing actual source of this anomaly is probably an old metal processing work-
shop, which can be assumed from the character and limited area of enrichment. Pb and
Ni displayed a vast area of elevated concentrations compared to the local background
(Figure 13.2E,G). The distribution of these metals suggests overlapping effect of several
enrichment sources, in particular motorways besides the Steelworks and Igolomia aban-
doned site. In general, from the spatial distribution of surveyed metals and their concen-
trations in soils of Site I in the vicinity of the Steelworks, it can be assumed that the
significance of the Steelworks with respect to soil contamination by metals in Cracow
292 Fate and Transport of Heavy Metals in the Vadose Zone
region is overestimated. Besides Zn and Cd in the limited area, all other surveyed metals
emitted from the Steelworks did not exceed the concentrations occurring in unpolluted
soils, despite Steelworks being in operation for more than 40 years. Other sources of
elevated concentrations of metals with respect to the geochemical background appeared
to be of comparable (Igolomia: Fe, Pb, Cd) or higher extent (lgolomia: Mn, Ni, eu;
motorways: Pb, Ni; railway: eu).
Comparison of the geostatistic analysis of Zn, Pb, and Cd occurrence in the Site I
fitted by GEO-EAS computer program (Twardowska, 1995) and by SURFER Version
5.01 Surface Mapping System applied in this study showed a very good consistency and
thus proved high compatibility of both software.
Barrier Capacity Of a Surface Soil Layer
Strong barrier capacity of a surface soil layer with respect to metals was proved by a
distinctly higher content of these species in contaminated surface soil than in deeper
parts of the soil profile. A vertical distribution of metals in the surface soil layer clearly
reflects an extent of anthropogenic impact and either anthropogenic or geogenic origin
of a metal. Good illustration is the vertical distribution of surveyed metals in soil profiles
in Site I (Figures 13.3 to 13.6). The elevated concentrations of Zn caused by the emis-
sion from the Steelworks occur actually only in the surface soil layer, while Zn content
in subsoilloessial matrix is at the uniform background level (Figure 13.3). The excess of
Zn concentration in the surface layer appeared to be more than 5 times higher compared
to the layer 20 to 30 cm in the profile R (Ruszcza) closest to the Steelworks and located
in the disused area. This indicates accumulation of metals from wet and dry deposition in
the uppermost layers of undisturbed soil profile. Land cultivation (plowing) causes ho-
mogenization of metal concentrations within the cultivated layer. Nevertheless, also in
the cultivated area, the uppermost soil layer directly exposed to the deposition will be
more enriched with metals. Therefore, any error or inconsistency in sampling procedure
can lead to serious errors in evaluation of the environmental impact of the particular
source of emission.
A pattern of a vertical distribution similar to that of Zn show also other metals of a
distinct anthropogenic impact proven by the pattern of spatial distribution, in particular
Pb and Cd (Figures 13.4, 13.5). For metals of a low or negligible anthropogenic impact
(e.g., eu) the species concentration along the whole profile is uniform (Figure 13.6). For
the evaluation of the extent of soil contamination, besides spatial and vertical distribu-
tion of metal concentrations in soils, no less important is metal mobility. This quality is
characterized by metal fractionation according to the binding strength (Figures 13.3 to
13.7), in particular by mobile/mobilizable fractions (Figure 13.7). Metal mobility and its
assessment will be discussed further.
The correct site-, use-, and target-specific risk assessment, cost-effective cleanup ac-
tions, and correct evaluation of barrier capacity of a surface soil layer with respect to the
continuous anthropogenic impact of a particular metal from the stack emission require
more detailed information concerning spatial and vertical distribution of species in this
layer itself, both in undisturbed and cultivated soil of different types.
The pattern of metal accumulation and migration in the undisturbed surface soil layer
was exemplified by the study of soil contamination in the vicinity of Irena Glasswork
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 293
after 7 years' production of lead glass (Site II) (Figures 13.1, 13.8). The survey of the
surface layer (0 to 2.5 cm) of undisturbed soils for lead in the vicinity of Irena Glasswork
after 7 years' production of lead glass (Site II) showed its extensive enrichment with Pb
species (Figure 13.9). The highest concentrations of Pb and their longest range occurred
along the intersection A-A, the lowest in the directions C-C, in conformity with wind
frequency. In all the investigated area, Pb concentrations in the uppermost soil layer
exceeded the natural background contents (20 mg kg-I). The distance range from the
source of emission of estimated concentrations above the level considered permissible
(100 mg kg-I) (Kabata-Pendias and Pendias, 1992) was from about 3000 m (NE, inter-
section A-A) to below 500 m (W, intersection C-C).
The vertical distribution of Pb shows, though, that high accumulation of this metal
occurs predominantly in the uppermost 0 to 2.5 cm layer of surface soil. The Pb concen-
tration decreases sharply (approximately in half) in the next 2.5 to 5 cm layer and gradu-
ally in the subsequent layers, generally in accordance with a logarithmic pattern. In the
layer 35 to 40 cm, all the Pb concentrations were below 50 mg kg-I. In the area up to
about 500 m radius around the Glasswork, the concentrations of Pb in the deepest sur-
veyed layer 35 to 40 cm were still above the background level. Up to over twofold higher
Pb contents in this layer were observed in the closest proximity (in the radius of about
250 m) from the source of emission (Figures 13.9, 13.10). The pattern of a vertical distri-
bution of Pb confirms, on the one hand, a very strong barrier capacity of the uppermost,
humic layer of the soil. On the other hand, it suggests lack of a leak-tightness of this layer
and permanent gradual "unloading" of the uppermost layer due to the vertical migration
of a metal. Considering the pattern of Pb vertical distribution, maximum mean concentra-
tions of this metal in the cultivated surface soil layer relevant to the reported values in the
undisturbed soil layer are supposed to be considerably lower due to averaging effect of
cultivation. The averaged values will range from about 80 mg kg-
I
in the direction of the
lowest wind frequency (W, NW, SW), to about 160 mg kg-
I
in the directions of the most
frequent winds (E, NE, SE). Due to continuous emission of metals, also in cultivated
soils the uppermost layer will display higher metal content resulting from accumulation of
dry and wet deposition in this layer in the periods between cultivation treatment of soils.
The presented data lead to a conclusion that for risk assessment at land use such as
playgrounds, sport fields, or recreation areas, where transfer pathways are through the
direct uptake and the risk receptors are children, the metal concentrations in the upper-
most 0 to 2.5 cm layer of soil should be considered.
Heavy Metal Binding Strength and Mobility in Soils
For quality-safe target-specific actual and potential risk assessment, the identification
of metal mobility in soils is of fundamental importance. The evaluation of the total spe-
cies concentration in the matrix does not give enough information, due to different bind-
ing strength of various physiochemical associations, in which the metals occur in soils.
This feature determines metal ability to mobilization under different exposure condi-
tions with respect to different direct risk receptors, of which humans (adults and chil-
dren), farm and wildlife, soil organisms and groundwater should be specified. At present,
many authors emphasize a necessity for determination of metal fractions of different
mobility as a requirement for risk assessment (Gupta et aI., 1996; McGrath, 1996; Ure,
294 Fate and Transport of Heavy Metals in the Vadose Zone
1996). For this purpose, sequential extraction schemes for distinguishing metal-binding
fractions appear to be an extremely useful tool. The concept of this technique is that
elements occur in the soil matrix in various pools of different binding strength which can
be assessed by different reagents (Ure, 1996).
Since 1973, more than 10 sequential extraction procedures using different extractants
and defining from three to nine extraction steps to identify "forms" of metal binding,
have been elaborated. Among them, those developed by McLaren and Crawford (1973),
Tessier et al. (1979), Kaszycki and Hall (1996), and Ure (1996) are of general or specific
use. The chemical extraction sequences by many authors are still subject to arguments
and controversy concerning the selectivity of extractants and the redistribution of met-
als among phases during fractionation (e.g., Tessier and Campbell, 1991; Xiao-Quan
and Bin, 1993; Tack and Verloo, 1996). The indisputable advantage of this method lies
in the possibility of evaluating actual and potential availability of metals for site- and
use-specific risk receptors, as well as of estimating long-term effects on metal mobility of
the changing controlling factors. The attempts of many authors are focused on using a
sequential extraction procedure mainly for the identification of chemical associations of
pollutants in different mixed organic/inorganic matrices. The greatest virtue of the chemi-
cal extraction sequences, though, is a possibility to differentiate between the fractions of
different binding strength onto particular matrix and to compare different matrices par-
titioning with respect to the fractions of adequate binding strength. Mechanisms of metal
binding onto these matrices can be different; e.g., metal bonding on the matters predomi-
nantly organic like peat and predominantly mineral like fly ash (Twardowska and Kyziol,
1996). In general, for these purposes the optimum sequential fractionation procedure
should be simple both analytically and conceptually and display an order of a consecu-
tive increase of binding strength. These properties are found in one of the most widely
applied sequential extraction procedures, that of Tessier et al. (1979), modified by Kersten
and Forstner (1986) for partitioning sediment samples. This procedure has been also
used directly for metal speciation in soils (e.g., Harrison et aI., 1981; McGrath, 1996),
also in this study for Site I (Figures 13.3 to 13.7).
Comparison of the chemical fractionation of zinc in the anthropogenically impacted
surface soil layer, and in consecutive subsoil layers where the Zn concentrations reflect
the natural background level, indicates occurrence of considerable qualitative as well as
quantitative changes (Figures 13.3, 13.7). With respect to distribution in the surveyed
unpolluted subsoil layers (>70 to 130 cm) in mg.kg.-
1
, Zn association with fractions of a
different binding strength followed the order:
F6(R) > F2(CARB) > F3+4(EMRO) F5(OM) FO+l(EXC)
The residual fraction accounts for about 50%, while 25 to 30% is associated with
mobile carbonate-bound fraction. Variable amounts of Zn are bound with mobilizable
F3+4(EMRO) and strongly bound with oxidizable fractions F5(OM): the rate of Zn
associated with both these fractions ranged from about 15% to approximately 30%. The
association of Zn with easily and moderately reducible Mn- and Fe-oxides comprised 14
to 25%, while in oxidizable fraction (generally associated with organic matter and sul-
fides, here in particular with organic matter), it occurred in a minor amount (1.2 to
5.0%). The role of the exchangeable fraction in Zn binding was negligible in all layers of
the surveyed soil profiles.
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 295
In overlaying intermediate subsoil layer 20 to 70 cm, both Zn concentrations and
chemical fractionation appeared to be similar to that in the underlying layer >70 to 130
cm, though a tendency to increase of Zn rate associated with reducible fraction was
observed.
In the surface humic soil layer weakly impacted by the emission from the Steelworks
(profiles Wand I), the enrichment of most binding fractions with Zn occurred, but to
the different extent. It resulted in the rearrangement of the relative and absolute parti-
tioning order according to the sequence (Figures 13.3, 13.7):
F3+4(EMRO) > F6(R) F2(CARB) > F5(OM) FO+1(EXC)
The association of Zn with mobilizable reducible oxide-bound fractions and oxidiz-
able fraction associated with organic matter in the surface soil increased up to 51 to 56%
of the total Zn, the reducible oxide-bound fraction being dominant (41-48%). The amount
of Zn associated with carbonate-bound fraction was generally stable along the soil pro-
file and did not show enrichment in the surface soil, which resulted in the substantial
decrease of the proportion of mobile fraction in comparison with the subsoil layer. The
results of Zn partitioning are generally in line with those reported by other authors for
unpolluted and geochemically polluted soils (McGrath, 1996) and sediments (Tack and
Verloo, 1996). All these matrices showed domination of Zn associated with reducible
oxide fractions, generally minor role of oxidizable fraction, high rate of stable residuum,
and relatively low proportion of mobile fractions associated with carbonates. The role of
the exchangeable fraction in zinc binding appears to be negligible. To summarize, in the
surface soil layer enriched with zinc anthropogenically, from 30 to 35% of this species
comprised immobile lithogenic material, while 65 to 70% of the total concentration com-
prises mobile or mobilizable fractions of different binding strength (Figure 13.7).
Partitioning of cadmium (Figures 13.4, 13.7) in the surveyed soil profiles followed the
sequence:
- in subsoil layer 30 to 130 cm:
F6(R) > F3+4(EMRO) > FO+1(EXC) > F2(CARB) F5(OM)
- in surface soil layer 0 to 30 cm:
FO+1(EXC) z F3+4(EMRO) > F6(R) z F2(CARB) F5(OM)
In the subsoil layer, 35 to 45% of Cd was associated with residual fraction. Reducible
oxide-bound Cd comprised 27 to 30%. The mobile exchangeable and carbonatic frac-
tions accounted for 19 to 22% and 12 to 17% of Cd, respectively, while the amount of Cd
bound to oxidizable organic fraction was negligible.
High content of humic organic compounds in the surface soil layer did not enhance
binding Cd with oxidizable organic fraction, whereas the amounts of Cd bound in ex-
changeable, reducible, and carbonate fractions significantly increased as compared to
subsoil. Except the oxidizable fraction, the Cd distribution among the fractions of differ-
ent binding strength was almost uniform, showing great resemblance to the pattern of
Cd partitioning in the moderately polluted soil studied by Harrison et al. (1981). Frac-
296 Fate and Transport of Heavy Metals in the Vadose Zone
tionation of Cd adsorbed onto peat; i.e., predominantly organic matter, displayed domi-
nance of binding mainly onto FO+1(EXC) fraction of the weakest binding strength
(Twardowska and Kyziol, 1996). It could be therefore admitted that a significant part of
the most labile FO+ 1 (EXC) fraction may be associated with organic matter, besides that
of clay minerals. In the case of humic-rich matters, the attribution of metal binding to ion
exchange mechanism is questionable. This supports an assumption, expressed also by
Kersten and Forstner (1988) and Tack and Verloo (1996), that the mechanism of metal
binding onto different or transformed matrices also considerably differs, while the most
reliable parameter for comparison is binding strength, adequate to the related fractions
(Twardowska and Kyziol, 1996). It should be also emphasized that in general, chemical
fractionation of Cd in soil and sediments indicates its predominant binding onto mobile
and easily reducible phases (Harrison et al., 1981; Kersten and Forstner, 1988; Forstner,
1992; McGrath, 1996; Tack and Verloo, 1996). Therefore, Cd is susceptible to
remobilization resulting from the changes of the chemical environment. The partitioning
of Cd is thus also subject to strong changes. The reported data on soils and sediments are
consistent with respect to role and significance of exchangeable, reducible oxide-associ-
ated and carbonate associated mobile and mobilizable fractions in Cd binding (Figures
13.4, 13.7). A bigger difference in the reported data is concerning oxidizable and re-
sidual fraction (Forstner, 1992; McGrath, 1996; Tack and Verloo, 1996).
Copper occurrence in the soils of the Site 1, as shown by the spatial distribution and
concentration range (Figure 13.2H, Table 13.4) displays weak impact of the emission
from the Steelworks and in the surveyed soil profiles is of predominantly geogenic ori-
gin. It results in uniformity of Cu distribution and partition along the profiles (Figures
13.5, 13.7). From 44 to 54% of Cu is stably bound in the residual fraction. The predomi-
nant part of mobilizable species was found in oxidizable fraction (25 to 33%), that seems
to be geogenically specific for soils and sediments and is in conformity with other sources
(Tack and Verloo, 1996; McGrath, 1996). It should be mentioned that also in some
anthropogenic materials, such as municipal solid wastes, domination of specific linkage
of Cu to organic matter was observed (Prudent et al., 1996), though not all the matrices
show the same binding pattern (Twardowska and K yzioL 1996). The partitioning of Cu
in the surveyed profiles, both in surface soil and subsoil layers, follows the general order:
F6(R) > F5(OM) > F3+4(EMRO) F2(CARB) >FO+1(EXC)
In the deeper part of Profile I (Igolomia), predominant binding to the reducible fraction
occurred (Figures 13.5, 13.7). No visible increase of Cu association with oxidizable or-
ganic matter-bound fraction was observed, despite much higher content of organic mat-
ter in this layer.
Distribution of lead (Figures 13.6, 13.7) reflects its low mobility in soil and subsoil
profIles. Opposite to Cd, association of Pb with mobile fractions, both exchangeable
and particularly carbonate-bound appeared to be very low. In all layers of the profiles,
including surface soil layer, mobile exchangeable and carbonate-bound fractions com-
prised 2.5 to 5.5% and 0.0 to 3.2% of total Pb, respectively. The highest, though vari-
able, enrichment of subsoil with Pb occurred in the residual (25 to 66%) and reducible
oxide-bound fractions (25 to 66%). The highest lead binding in the residual fraction (56
to 66%) and the lowest in reducible oxide-bound one (22 to 29%) was observed in the R
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 297
(Ruszcza) profile, in the area of the highest impact of the Steelworks. In two other
profiles Wand I located in less impacted areas the proportions of the residual and
reducible fractions were 25 to 43% and 52 to 66%, respectively. The total Pb contents
in the subsoil (> 30 cm) of the R profile ranged from 10 to 12 mg kg-I, while the concen-
tration range in the subsoil of these two other profiles was 10 to 22 mg kg-
I
and could be
thus assumed as falling within the uniform background concentrations. Therefore, it is
rather unlikely that enrichment of residual fraction and decrease of reducible one was
induced anthropogenically, and probably reflects the geogenic variability of the area in
this respect.
In the subsoil layer > 70 cm, Pb binding onto organic matter is generally very low or
negligible, but increases in the upper transitional subsoil layer (70 to 50 cm).
In the surface soil layer, the proportion of Pb associated with organic fraction consid-
erably increases (to 13-23%), which shows good correlation both with the content of
organic matter in the soil profile and exposure to the anthropogenic impact. Besides
higher rate of organic-bound fraction and general quantitative increase of Pb-enrich-
ment, dependent upon the distance and direction with respect to the emission source, no
substantial changes in partitioning of this metal in soil profile was observed.
Partitioning of Pb with respect to binding strength and predominant chemical asso-
ciations in the surveyed soil profiles followed the sequence:
- in subsoil layer 30-130 cm:
F3+4(EMRO) >< F6(R) FO+1(EXC) ~ F2(CARB) ~ F5(OM)
- in surface soil layer:
F3+4(EMRO) >< F6(R) > F5(OM) FO+1(EXC) ~ F2(CARB)
Comparison of chemical fractionation of Pb in the surface soil layer in Site I (Figures
13.6, 13.7) and in Site II adjacent to the Irena Glasswork (Figure 13.8), where the up-
permost 0 to 2.5 cm soil layer is highly contaminated by lead (about one order of magni-
tude compared to the surface soil layer in Site I) (Figure 13.9), displays clear influence
of the extent of anthropogenic impact on Pb distribution among the fractions (Figure
13.10). Partitioning of Pb in the least contaminated soil samples (C-15-a and D-22-a)
shows high enrichment of the stable bound residual fraction (62 to 63%). The rest of
species was almost equally partitioned over mobile and mobilizable fractions of different
binding strength: exchangeable fraction comprised 8 to 9%, the fraction associated with
organic matter 12 to 15%, while the rest was distributed among oxide-bound and specifi-
cally sorbed (mainly carbonate-bound) fractions:
F6(R) F5(OM) > F1+0(EXC) >< F2(CARB) >< F3+4(EMRO)
In the most contaminated area (samples A-1-a and B-8-a), considerable changes of Pb
distribution occur:
F6(R) "" F5(OM) F3+4(EMRO) >< F2(CARB) > FO+1(EXC)
298 Fate and Transport of Heavy Metals in the Vadose Zone
Comparison between the samples of the highest (A-I-a) and the lowest (D-22-a) Pb-
contamination showed that the most anthropogenically enriched fractions appeared to
be those associated with organic matter (46%) and stable residuum (28%). Much weaker
anthropogenic impact displayed, in the descending order, fractions: specifically sorbed
F2(CARB) (14%), oxide-bound F3+4(EMRO) (8%), and exchangeable FO+1(EXC)
(4%). The anthropogenic enrichment follows, therefore, the sequence:
FS(OM) > F6(R) F2(CARB) > F3+4(EMRO) > FO+1(EXC)
The chemical fractions associated with Pb in the soil samples taken from Site II cannot
be directly compared with those in Site I due to use of different sequential extraction
methods (by McLaren and Crawford, 1973, in Site II and by Tessier et ai., 1979, modi-
fied by Kersten and Forstner, 1986, in Site I). The analysis of Pb partition in both sites,
though, clearly shows that the highest enrichment due to the anthropogenic impact (stack
emission) occurs in the oxidizable organic matter-bound and stable residual fraction.
Mobile chemical associations with exchangeable and carbonate fraction, as well as with
reducible step associated mainly with manganese oxides and amorphous iron oxy-
hydroxides, are subject to the anthropogenic enrichment to much lesser extent.
To conclude, heavy metal fractionation in surface soil and the vadose zone matrix
differs substantially with respect to binding strength. Surface soil has high barrier prop-
erties, which cause enrichment of this layer with heavy metals in the areas impacted by
anthropogenic emission. In general, anthropogenic enrichment occurs in all binding frac-
tions, though at a different rate. The highest increase has been observed in mobilizable
fractions, which results in the elevation of hazard to higher extent than it can be assumed
from the quantitative changes. This leads to the conclusion that for quality-safe risk
assessment, not only quantitative but also qualitative transformations of metal associa-
tions caused by anthropogenic impact should be considered.
Monitoring Program Requirements for Risk Assessment from
large-Area Soil Contamination by Trace Metals from
Anthropogenic Sources
The results of soil survey in two anthropogenically impacted sites show the impor-
tance of assessing such parameters as (i) actual and potential land use and risk receptors;
(ii) the thickness of an averaged surface soil layer to be exposed to a direct contact with
risk receptors and the form of a contact; (iii) the fractions of the total metal content in
soil actually available and implying a risk for the risk receptor; (iv) the fractions of the
total metal content in the soil potentially available (mobilizable) and probable conditions
of the metal(s) mobilization. These parameters are essential for a quality-safe monitor-
ing and evaluation of an extent of soil contamination by trace metals.
The monitoring requirements are based on the character of a vertical distribution of
metals in anthropogenically impacted soil from the large-area emission, showing high
accumulation in the uppermost layers of soil, 1-2 cm thick. In the areas undisturbed by
depth-averaging cultivation treatment (e.g., lawns in childrens' playgrounds, meadows
used as grazing areas) this layer will be directly exposed to the contact with receptors
(children, farm animals, and wildlife). In agricultural land, the direct receptors (e.g.,
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 299
plants, food crops, fodder crops) will be exposed to a concentration of metal averaged by
cultivation treatment.
Considering the association of metals in soil matrices with "pools" displaying different
binding strength, which reflects direct and potential availability to the different recep-
tors, application of the sequential extraction procedure gives an essential opportunity to
avoid false-positive errors in actual risk assessment. Overestimating the risk may be
avoided through excluding the rate of metal stably bound in the residual fraction. The
correct risk assessment requires an identification of metal-binding fractions directly avail-
able to the particular receptors, e.g., mobile fractions FO (pore solution), F1 and F2
displaying the weakest binding strength, and thus susceptible to leaching and ground-
water contamination. In general, for the actual risk assessment, evaluation of mobile,
mobilizable on/after uptake and immobile (stable) fractions provides adequate required
information. For this purpose, the sequential extraction is a proven, reliable tool. The
testing in the frame of BCR-interlaboratory studies of two extraction procedures, to be
considered as standards by ISO (Quevauviller et al., 1996; Ure, 1996), confirms both
their reliability and usefulness for risk assessment needs. As has been shown above,
sequential extraction also provides valuable information on quantitative and qualitative
changes in partition of a metal in question, resulting from the anthropogenic impact.
For the potential risk assessment, long-term prognosis of heavy metal release and
selection of the optimal remedial/cleanup actions, not only metal fractionation according
to binding strength, but also identification of the geogenic and anthropogenic chemical
associations of pollutants, in particular in mobilizable fractions [easily and strongly re-
ducible F3+F4(EMRO) and oxidizable (FS) (OM)] are required. The definition ofpa-
rameters, transforming equilibria conditions in matrix, as well as external or internal
factors controlling these transformations (e.g., pH, Eh) are also a prerequisite for the
correct life cycle risk assessment.
The results presented here show the need for a differentiated approach to actual and
potential risk assessment from the large-area sources of emission such as stack emission,
and point out the pitfalls of data inconsistency in their evaluation. The monitoring pro-
gram for quality- fe risk assessment and the selection of a both efficient and cost-effec-
tive remedial strate minimizing the adverse consequences oElong-term emission should
be highly use-specific and target-oriented.
Monitoring data on trace metal enrichment in soil and vadose zone matrix caused by
wet and dry depositio from industrial sources (mainly stack emission) in the vicinity of
operating industrial lants (Sendzimir Steelworks, Site 1, and Irena Glasswork, Site II)
showed an essenti role of chemical fractionation of metals in adequate evaluation of
soil contaminati n. A substantial part of the total metal load originating from the anthro-
pogenic industrial sources is stably bound in the residual fraction. In some emissions,
though, anthropogenic contaminants occur in more labile forms than the species of the
lithogenic/geogenic origin, which adequately increases the risk (e.g., anthropogenic en-
richment of oxidizable fraction with Pb in Site II). Taking into consideration at risk
assessment not only contaminant concentration, but also its chemical fractionation with
respect to binding strength could highly improve the classical principle of preliminary
evaluation of contaminated sites based on soil threshold values. Application of scientifi-
cally proven critical values would also greatly enhance site- and use-specific models of
exposure assessment. Up to now, these values are a weak point of the best-constructed
300 Fate and Transport of Heavy Metals in the Vadose Zone
exposure assessment models. Metal fractionation in soil for risk assessment and man-
agement has been taken into account in a three-level concept by Gupta et al. (1996).
EVALUATION OF A LARGE-AREA DESERTED INDUSTRIAL SITE
Investigation of a large deserted industrial site as a potential human risk was pre-
sented in the case study on an abandoned industrial area of Marktredwitz in Germany,
impacted by the long-term emission of Hg and Sb from an old chemical plant (Site III)
(Figures 13.1, 13.11). The major issue facing old contaminated sites sanitation require-
ments is the need of a quality-safe evaluation of such areas, taking into account both
interests of the environment and nature on one side and economy and industry on the
other. Therefore, an optimum model of investigation and assessment of chemical pollu-
tion of the site is to be use- and site-specific, in accordance with particular criteria in
view of the defined protection objectives, which are determined by further use of the
decontaminated area, and corresponding human sensitivities. In Germany, the efforts
directed to elaboration of reliable long-term risk assessment methods resulted in devel-
oping several models of different applicability.
The proposed approach to the assessment of the human risk potential originating
from deserted industrial sites has been exemplified in a case study of the large-area soil
contamination by mercury and antimony in Marktredwitz city, North Bavaria, FRG
(Figures 13.1, 13.11). The study, conducted by the research group of the GSF-Institute
of Ecological Chemistry, FRG, has been focused on a site-specific risk assessment and
selection of the adequate preventive/remedial action. Unlike the studies in Sites I and II,
oriented to one selected measurement endpoint (soil), this study was of a complex char-
acter: measurement endpoints included soil, water, air, sediments, dust, plants; while
target risk receptors were human: adults and children.
Site Characteristics
Site IV is a typical urban area of a city that started to develop in the industrialization
period of the end of the eighteenth century as a residential area of one of the oldest plants
in Germany, chemical factory Marktredwitz (CFM). Hence the central position of the
plant in the town, which is surrounded by a railway (Figure 13.11). The CFM area
adjoins the Kossene river course, which belongs to the Elbe River drainage basin. The
river was regulated in e 1930s to intercept frequent floods. The reclaimed old riverbed
is also adjacent to CFM. ain wind directions are Wand SW. Dominating types ofland
development are individu I houses with gardens.
Sources of Heavy M11 Contamination in the Area
The major s o u r c ~ heavy metal contamination in the area is now the abandoned
industrial site of a more than 200-year-old former chemical factory Marktredwitz
(CFM) founded in 1788 (Figure 13.11) which used to produce a great variety of inor-
ganic and organic chemicals, among them Hg- and Sb-based compounds and herbicides
(Table 13.5) before closure in 1985 for ecological reasons. Maximum concentrations of
heavy metals found in soils of the area (Table 13.6) reflect an extent of the environmen-
tal damage.
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 301
Figure 13.11. General map of Site III: Marktredwitz urban area with chemical factory site, North
Bavaria, F.R.G. with the location of measurement points for Hg.
Table 13.5. Compendium of the CFM Catalog: Heavy Metal-Bearing Chemicals
CFM - Products
Inorganic and metalorganic products:
Hg numerous products, under it Hg, Hg
2
C1
2
, HgX2 (X=CI, Br, I, CN, SCN),
Hg(N0
3
h, HgO, RHgCI (R=CH
3
, C
2
H
s
, C
6
H
s
), etc.
Sb potassium antimony tartrate, potassium antimony citrate
As Hg
3
(As0
3
h, Hg
3
(As0
4
h
Zn Zn
3
P2
Cu 3Cu(OHhCuCl
2
Among 12 groups of CFM products, Hg-bearing chemicals comprised 10 different
groups, of them Fusariol accounted for 5%. One group represented Sb compounds, of
which over 90% consisted of antimonyl tartrate (tartar emetic). Mercury-
bearing chemicals prevailed in the toi CFM production: the rate of Hg-products ac-
counted for 94%, and Sb-products, 6 Yo. The Hg- and Sb-products delivery from CFM
to customers increased since 1961 to 1982 from 67.0 to 116.1 t and from 0.7 to 9.6 t,
respectively.
Registered trademark of the Chemical Factory Marktredwitz, Inc., Marktredwitz, Germany.
302 Fate and Transport of Heavy Metals in the Vadose Zone
Table 13.6. Detected Heavy Metal Contaminants from CFM and Their Maximum
Concentrations in Soil
Metal As Cd Co Crt Cu Hg
C (mg kg-
1
) 684 1.64 18.8 560 673 6,140
Metal Mn Ni Pb Sb Sn Zn
C (mg kg-
1
) 770 390 19,364 36,400 13.2 2,532
Monitoring Strategy
A complex character of the old contaminated site evaluation determined a broad pro-
gram of preliminary investigations undertaken in the framework of the Marktredwitz
project. Due to historically long-term impact of contaminants under the changing condi-
tions of area development/management and extent of anthropopression, detailed pre-
liminary studies were required to most accurately define the monitoring strategy. These
studies comprised measured endpoints and risk receptors, sampling points and param-
eters assuring quality-safe evaluation of the site.
The preliminary studies undertaken in the framework of Marktredwitz project (Site
III) were focused on making explicit the factors of concern for site evaluation, in par-
ticular the kind and pathway of pollutants, past and future area development, availabil-
ity, and adequacy of the existing database. The target task was to elaborate an optimum
sampling and measurement program adequate for a reliable use-specific risk assessment.
The studies comprised historical background investigation and elaboration of a geo-
graphical information system (GIS) for an investigated site. The objectives of the his-
torical investigations were to identify precisely vs. time: (i) contaminants inventory:
industrial products in the area, delivered amounts; (ii) pathways of contaminants: aquatic
(surface- and groundwater), terrestrial (controlled and uncontrolled waste disposal and
use), air (wet and dry particulate deposition); (iii) uncontrolled waste disposal and use
as common fill or soil amendment; and (iv) causes of environmental damage from un-
controlled sources.
To identify possible uncontrolled disposal of contaminated material in the Marktredwitz
area, the inventory of industrial and other sites (e.g., quarries) where such material could
have been disposed was elaborated. For this purpose also, the available archival aerial
views were investigated to detect changes of the cityscape in time, where contaminated
material could have been involved as a common fill (relocation of the river bed, urban
area development, road construction, changes of land use, position and condition of in-
dustrial sites in time, land leveling).
For the contaminated SIte eva tion, a GIS visualizing any kind of data with refer-
ence to their Gauss- Kruger or loca oordinates in the Windows style appeared to be
particularly suitable. It served as a spa ally allocated data bank of required information:
general, pathways and input/output 0 contaminants (emission, imission, utilization),
sampling and analysis (e.g., Figures 1 .1 L 13.12).
Survey of Transfer Pathways and Risk Receptors
A sampling program was designed with use of the GIS spatially related database ob-
tained from the preliminary historical investigations. It was focused on deriving compre-
,"",,,,,7)'1'\1 Metal Contamination in Industrial Areas and Old Deserted Sites 303
antimony (new values)
mercury
railway installations
- river KOsselne
III residential area

Figure 13.12. Site III: Marktredwitz area with the location of circles D=130 m to define sampling
paints for Hg and Sb.
hensive and reliable data for evaluation of the extent and propagation of contamination in
the areas suspected of being polluted. In these investigations, a human as a risk receptor
was a target assessment endpoint, while groundwater was not considered at this stage due
to the lack of elevated concentrations of site-specific pollutants in drinking water.
The area to be surveyed for Hg could be roughly estimated on the basis of the avail-
able qualitative information, while for Sb no such estimation was possible due to insuffi-
cient data and weak correlation between the occurrence both metals. The sampling
area was thus planned, starting from the old CFM site and define according to the main
transfer pathways; i.e., the Kosseine river flow (E) and predomina t wind direction (S,
SW), though E direction of wind transport also occurs. As a joint) effect of the major
pathways, in the E direction from the old CFM site, "hot spots" of the highest extent of
contamination were expected, while in the W direction a somewhaylesser contamination
could not be excluded. With the help of the geostatistical analysts' and available data on
Hg, the distance of 130 m in diameter was found to be sufficient for the reliable evalua-
tion of the contaminant expansion. On designing the sampling network, the maximum
distances of sampling points accounting for 130 m were therefore generated by means of
circles centered in proven contaminated points (Figure 13.12). The maximum distance
of 130 m was assumed to be valid also for Sb. Thus, an extensive sampling of the sur-
veyed area could be accomplished with a minimum effort.
For the quantitative exposure assessment, all relevant transfer pathways comprising
soil, food crops, indoor and outdoor ambient air were sampled. Soil samples for analy-
sis were prepared through averaging of a sufficient number of random samples from the
304 Fate and Transport of Heavy Metals in the Vadose Zone
respective area, therefore the results represented mean values for the area. In total, about
200 areas were sampled. The sampling was performed in accordance with ISO/DIN-
10381. The soil exposed to air contamination was taken from the top layer 0 to 10 cm.
The soil enriched with contaminants from a long-term impact was sampled also from the
layer 10 to 30 cm, if risk receptors were children. Additionally, in some points layers of
30 to 60 cm were also taken.
Sampled food crops grown in contaminated house gardens comprised mainly veg-
etables and fruits. Particulate samples were taken from the indoor ambient air in the
living rooms and outdoor of the residential houses, and a fraction <2.5 pm (lung pen-
etrating dust) was separated for analysis. In total, 800 soil samples, 200 plant samples,
and 146 indoor air and dust samples were collected.
In the Marktredwitz site, a human biomonitoring was conducted simultaneously with
sampling the transfer pathways, i.e., soil, food crops, and indoor air. The residents of the
areas supposed to be highly contaminated were invited to participate voluntarily in the
investigations, submitting 24 hours urine and giving blood samples. In total, 264 volun-
teers participated in the monitoring, among them the most sensitive group: 0 to 4-year-
old children accounted for 8 volunteers giving blood, and 21 giving urine. The collected
261 blood samples and 264 urine samples were examined for Hg and Sb.
In conformity with the sampling program and a scope of the survey, the material
analyzed for Hg and Sb comprised atmospheric particulate, soil, plants, river sediments,
indoor and outdoor air, human blood and urine. Samples were analyzed for Hg and Sb
content by standard methods using ICP-MS techniques (ICP Perkin Elmer Sciex ELAN
5000 coupled with a high-resolution quadruple mass-spectrometer Finnigen Mat).
The results of the screening/monitoring survey were introduced into the GIS Geo-
graphical Information System. Currently, GIS is used as a spatially allocated data bank
of required input information on the contaminated site. Further development of the sys-
tem will comprise the integration of a geostatistical estimation of contamination with a
quantitative estimation of exposition as equipotential maps.
Human Risk Potential Assessment
Approach to Human Risk Potential Assessment
For evaluation of Hg contamination in the Marktredwitz area, the method of use-
specific and site-specific quantita ve exposure assessment (QEA) has been applied. The
QEA method adopted by the Insti te of Ecological Chemistry is an individual site-
specific qualitative exposure estimati n model comprising standardized scenarios with
defined uptake rates through the tra sfer paths for different kinds of soil utilization
developed at the Institute Fresenius GmbH, Taunstein in Germany (Simmleit et al.,
1997). The application and of the QEA model were a part of the
Marktredwitz case study. The essence of the model and the procedure to be followed for
assessing the risk potential will be presented and discussed in the example of evaluation
of Site III contamination by Hg in the background of a critical discussion of other risk
assessment concepts.
The Marktredwitz study and applied concept of risk potential assessment exemplifies
a new site-specific approach which is aimed at finding a sound compromise between the
economy and environment. To facilitate the understanding of differences and specificity
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 305
Area A - protect:
Area B - tolerate:
Area C - remediate:
unlimited, local general and multifunctional possibility of use,
limited, but site- and protected object-related possibility of use,
area of toxicity, where a damage of the protected object (plants,
animals, humans, ecosystems) might occur; protective means are
necessary
Figure 13.13. Definition of the Three Areas according to Eikmann, Kloke, and Lilhr (1991).
of this concept, the risk evaluating models most widely applied in Germany (Three-
Area-System and Critical Values by Ewers/LAGA) will be discussed below.
Fundarnental Remarks
Soil contamination can be assessed in two essentially different ways: (i) use-nonspe-
cific, protecting soil multifunctionality; (ii) use-specific, which considers further use of a
decontaminated area. Most experts consider a use-nonspecific assessment unrealistic.
Of site-related and use-specific methods, a Critical Value System by LAGA and Three-
Area-System or Eikmann/Kloke list are of particular significance. Other classical prin-
ciples of contaminated sites evaluation are based on soil threshold values derived from
experience and summarized in assessment lists (e.g., Dutch list, Swiss list). Though widely
used by environmental engineers, only a few lists are toxicologically proven. An interest-
ing approach for three-level risk assessment was proposed by Gupta et al. (1996).
Three-Area-System
The Three-Area-System, known also as an EikmannlKloke list (Eikmann et al., 1991)
has been elaborated to provide a risk assessment based on the land use and sensitivity of
objects to be protected against soil contamination. Soil pollution is attributed to three
increasing levels (A, B, C) prescribing adequate actions (Figure 13.13). The kind of
action following the determination of soil contamination is divided into three categories,
depending on the seven soil utilization modes and related risk receptors (e.g., playgrounds
for children, house gardens). The latest issue of the Three-Area-System (Eikmann and
Kloke, 1993) comprises a list of 19 heavy metals, 5 nonmetal and 3 chloroorganic com-
pounds, while protection measures are focused on human. The system, nevertheless, uses
empiric compilation of soil values (e.g., for Hg: Table 13.7) and does not incorporate soil
parameters into the EikmannlKloke list. Therefore, It n be considered as a guideline
for use-specific, but site-nonspecific assessment without to icological background.
Use-Specific Critical Values by EwerslLAGA (Ewers and 1993)
The use-specific critical values (PW) and threshold the necessity of
preventive actions (MW) proposed by the Ewers/LAGA committee, contrary to
those of the Three-Area-System, are based on the toxicological data (Hassauer et al.,
1993). The respective values for mercury are presented in Table 13.8. However, these
values also display some weak points, the major of them being: (i) entirely exposure
mode "oral soil intake" by the most sensitive risk receptors (little children) is considered;
306 Fate and Transport of Heavy Metals in the Vadose Zone
Table 13.7. Guide Values for the Use-Dependent Evaluation of Soils Polluted by Hg, According
to Eikmann and Kloke (1993)
No. Use Protected Object BW
mg Hg kg-
1
0 Multifunctional human beings, BWI 0.5
others
Playground for children human beings BW II 0.5
BW III 10
2 Private gardens human beings BW II 2
BW III 20
3 Sport grounds and football fields human beings BW II 0.5
BW III 10
4 Parks and leisure grounds human beings BW II 5
BW III 15
5 Industrial and trade areas human beings BW II 10
BW III 50
6 Agricultural land, others BW II 10
orchards and market gardening BW III 50
7 Ecosystems not used agriculturally others BW II 10
BW III 50
Table 13.8. Use-Specific Critical Values (PW) and Cleanup Threshold Values (MW) for
Assessing Soil Contamination by Hg, According to Ewers/LAGA (Ewers and Viereck-
Goethe,1993)
PW (mg Hg kg-
1
)
MW (mg Hg kg-
1
)
Playground for
Children
4
10
Residential
Area
8
20
Parks and
Leisure Grounds
20
(ii) threshold values indicating the necessity of preventive action are based upon the
rates of resorbed pollutant dose, while currently there is no standard method for evalu-
ation of resorbable fraction in soil; (iii) the values do not provide an evaluation of "risk
situation" in the legal and juridical sense. Hence, the values by Ewers/LAGA are of an
entirely informative character and hardly meet the requirements of recommendations
for preventive/remedial action.
Applied Model: Quantitative Exposure Assessment (QEA)
Bosic Concept
/
A concept of quantitative exposure assessme (QEA), applied for evaluation of soil
contamination by Hg and Sb in the old site of t e Chemical Factory Marktredwitz is
aimed to avoid the simplifications and disadvantag s of the approaches presented above.
The method is comparable to the UMS model; that is a German abbreviation of "Toxico-
logical Assessment of the Human Exposure to ollutants from Contaminated Sites"
(Simmleit et aI., 1997; Hempfling et aI., 1991). he exposure scenarios accommodated
by different federal expert groups were recently published by Stubenrauch et al. (l994a, b).
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 307
For the evaluation of individual areas in the vicinity of CFM in the Marktredwitz site
evaluation, the QEA model modified for the most sensitive land-use scenario, "Living
Area with a House Garden," has been used in accordance with the ARGE Foconl
Fresenius Institute. It follows the gradual procedure, from exposure estimation and toxi-
cological data (Figure 13.14A) to calculating a risk value describing the degree of haz-
ard (Figure 13.14B).
In Table 13.9, the transfer/exposure pathways of Hg to two groups of risk receptors
within the scenario "Living Area with a House Garden" and intake rates used for QEA
calculations are incorporated.
The intake rate of pollutants by human from a particular area has been determined by
three factors: (i) frequency and mode of land use; (ii) pollutant concentration in contact
media (soil, air, and plants); (iii) ubiquitous pollutant concentration in foodstuffs and
contact media, including drinking water. Ubiquitous background contamination of hu-
man (UBI) refers to an average intake rate from foodstuff, drinking water, and air. To
evaluate the total risk value (RW), UBI values should be added to the site-specific in-
take rate. There is, though, still lack of statistically confident UBI data; consequently,
just estimated value can be used. The reported data on Hg intake and resorbed dose vary
considerably. WHO (1990) estimates a total resorbed Hg concentration in the range
from 0.08 pg kg-
1
d-
1
up to 0.28 pg kg-
1
d-
1
per 70 kg body weight. In Germany, an
overstated UBI value of 0.08 pg kg-
1
d-
1
for Hg was accepted for land use modes such as
playground for children, park and sports ground (for protective reasons). At present,
taking into account the quality of foodstuff of a proper origin and the indoor residence
time, the UBI value for the use scenario "Living Area with a House Garden" has been
reduced by the Fresenius Institute to 0.025 pg kg-
1
d-
1

To evaluate tolerable resorbed dose of Hg, a resorption rate of 7% for oral and 80%
for inhalative intake of inorganic Hg was assumed (WHO, 1991). As all available data
from the Marktredwitz area refer to total Hg concentrations, the provisional inorganic
Hg values were estimated as 0.1 % Hg
t
in soil and 10% Hg
t
in air, according to analyses
by GKSS/LAGA (1993). The observations of methyl-mercury are not actually neces-
sary, unless the concentrations of organic Hg exceed 0.05 pg kg-
1
d-
1

The provisional guide values for Hg used by Hassauer et al. (1993) for the toxicologi-
cal evaluation of total body dose rates were derived from the lowest observed adverse
effect level (LOAEL) during the subchronical and chronical animal testing and the low-
est observed impact level of an epidemiological study. These values were defined as
" ... total body doses of a pollutant, which-at the state of the art-on their own do not
exhibit adverse effects on health, or give only little rise to inducing health risks" (Hassauer
et al., 1993). The provisional guide values used in this study accounted for 0.08 pg kg-
1
d-
1
(TRD oral) and 0.07 pg kg-
1
d-
1
(TRD inhalative). These values are supposed to be
questionable and subject to chan at t lead to different results of evaluation.
Assessment of risk caused by the presume 'te contamination is aimed at evaluating
actual and potential risk situations with regard t O ~ h e legal and juridical term "risk." This
term defines a situation that would provoke a da age, i.e., a violation of the integrity of
protected objects if no preventive actions are take (LAGA, 1991).
In order to evaluate exposure, uptake pathway specific risk indices are calculated as a
ratio of an intake pathway-specific dose (PDI) to tolerable absorbed dose (TRD). The
sums of indices calculated this way are then being referred to the ubiquitous background
308 Fate and Transport of Heavy Metals in the Vadose Zone
Quantitative Exposure Assessment for the Scenario "Living Area with a House Garden"
(According to HEMPFLING et aI., 1994)
soil- oral;
Determination of the
relevant exposure pathways
plant - oral; indoor air - inhalative; indoor dust - inhalative
Determination of the
pollutant concentration in contact media
/ / \ '"
soil plant indoor air indoor dust
Calculation of the
daily intake rate (DlR) of contact media
user groups:
-little children
- adults
----
different exposure pathways
Calculation of the potential daily
resorbed intake dose (POI) for
/ " different user groups different intake pathways
Figure 13.14A. Flow c rt of the q titative exposure assessment (QEA) for Hg in the neighbor-
hood of the former CFM for the expos e scenario "Living Area with a House Garden: Calculation
of the potential daily resorbed intake ose POI.
!
I
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 309
Quantitative Exposure Assessment for the Scenario "Living Area with a House Garden"
(According to HEMPFLING et al., 1994)
Risk assessment by calculating the
specific risk indices (RI) for each intake pathway using the

potential daily resorbed dose (POI)
for the inhalative and oral
tolerable resorbed pollutant dose (TRD)
for the inhalative and oral
intake pathway intake pathway
calculation of the total risk index (GRI) by summing up the intake pathway
specific risk indices considering the background burden
from food, water, air,
Calculation of the risk value (RW) as a base of assessment
RW = risk index: risk factor
(where the risk factor (GF) depends on the security
factors used fOI derivation of TRD-values)
Deduction of site- and use-specific recommendations
dependent on
/
\
the background burden quality of the data base
0,25 $ RW" 1
RW<! 1
a risk cannot be excluded, however it is in the range
of the ubiquitary background burden
a growing concern must be considered; preventive
measures, respectively changes of use,
are to be examined
the probability of heath impact is high; remediation
action is necessary
Figure 13.148. Flow chart of the quantitative exposure assessment (QEA) for Hg in the neighbor-
hood of the former CFM for the 'Living Area with a House Garden: Calculation
of the risk value RW. /
\
310 Fate and Transport of Heavy Metals in the Vadose Zone
Table 1 3.9. Survey of the Exposure Pathway-Specific Intake Rates for the Scenario "Living Area
with a House Garden"
Establishing the Pathway Specific Intake Rate
User Exposure Medium -
Group Exposure Pathway BW
a
Little soil - oral 10 kg
children plant - oral 10 kg
indoor air - inhalative 10 kg
indoor dust - inhalative 10 kg
Adults plant - oral 70 kg
indoor air - inhalative 70 kg
indoor dust - inhalative 70 kg
a BW: body weight.
b Homegrown fruits and vegetables (dry weight).
C Dust retention: 75%.
Daily Exposure Intake Rate
Intake Frequency (average per year)
1 g 200 d a-
1
55 mg kg-
1
d-
1
3,7 g d-
1
,b
all-year
370 mg kg-
1
d-l,b
3 mL d-
1
21 h d-
1
0,26 mL kg-
1
d-
1
3 mL d-
1
21 h d-
1
0,26 mL kg-
1
d-
1
,c
20 g d-
1
,b
all-year 290 mg kg-
1
d-
1
20 m d-
1
21 h d-
1
0,25 mL kg-
1
d-
1
20 ml d-
1
21 h d-
1
0,25 mL kg-
1
d-l,c
concentration (UBI). This method, however, is not sufficient for deriving well-founded
recommendations of protective measures related to the specific risk receptors.
To fill the gap between the toxicological statement on the one hand and the legislative
requirements on the other, Konietzka and Dieter (1994) proposed to assess a risk threshold
value with regard to the safety factors used for determination of the TRD values. The
safety factors are expected to be slightly below the LOAEL of sensitive individuals.
According to this proposal, risk indices are transformed into risk values depending on
the reliability of data used for estimation of provisional guide values. Estimation of the
provisional guide value for the oral uptake of inorganic mercury was based upon oral
uptake data for rats, with a safety factor of 200, while the respective values for inhalative
uptake of inorganic Hg were derived from LOAEL for humans, with a safety factor of
20. The Fresenius Institute determines risk threshold values by using provisional guide
values for oral uptake with a safety factor of 10 and for inhalative uptake with a safety
factor of 4.5, referring to Konietzka and Dieter (1994).
A risk value (RW) calculated as the total risk index to the risk factor ratio, allows
distinguishing between the lack of risk and possible risk for the receptors. To describe
the situation for the land-use scenario "Living Area with a House Garden," weighted
means are used as risk factors for oral and inhalative uptake.
In Figure 15.15, the three RW ranges for Hg of a different probability of risk situa-
tions requiring adequate actions are presented. At the lowest stage (RW < 0.25) the risk
cannot be entirely neglected, but no specific action is required. At the highest stage of
risk, preventive or remedial action has to be taken.
Estimation oj Soil Values jor a Scenario "Living Area with a House Garden. "
Marktrec1witz Area (Data up to 1992)
The risk assessment using I sensitive scenario "Living Area with a House
Garden" applied in Mark edwitz reqUl s database quality for transfer/uptake path-
ways (soil, plant, indoor concentrations) be evaluated. From Figure 13.16 one can
conclude that soil was the only medium exa ined in 61 % of all the areas (quality level
0,25 :s;
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 311
RW
RW
RW
<
<
;:0:
0,25
1,0
1,0
hazard in the range of the average background burden
risk with a growing concern
danger (in the sense of police regulations)
Figure 13.15. Scheme of risk values.
PER:ENTAGE
OF SINGLE
AREA
70
60
50
40
30
20
10
0
ANALYTICAL
DATA FDR All
EXPDSURE
PATHWAYS
QUALITY OF DATABASE
A B C
----+
DB:RfASING DEGREE OF QUALITY
D
ANALYTIC AL
DATA DNlY
FOR ONE
EXPOSURE
PATHW AT
Figure 13.16. Database quality for a single area QEA for the scenario "Living Area with a House
Garden': decrease of quality from A to 0, expressed in percent of the area surveyed for Hg in soil.
D). The problem of the result compatibility in time for different uptake pathways in
individual fields was not considered.
The results of an exposure assessment carried out within the project "Site-specific
assessment of the contamination in the vicinity of the Marktredwitz Chemical Plant,"
displayed a virtual significance of the indoor inhalative uptake path for the used sce-
nario. Quantitative exposure assessment for individual areas was accomplished using
either actual measurements or statistical values generated from indoor analyses. The
indoor concentrations of mercury were found to be strongly influenced by the living
habits of residents (e.g., frequency of room cleaning, airing, etc.) and also dependent
upon the time of sampling, conducted in 1987 to 1991, due to the altered exposure con-
ditions (demolition and removal of contaminated CFM buildings). Another problem
was a small amount of indoor data available. These reasons led to the concept of using
indoor data as statistical parameters for evaluating Hg concentrations in air and indoor
suspended dust for the QEA of the single areas. Mercury concentrations in food/fodder
crops have been also evaluated on a unjformst:atis . al basis and then used to calculate
critical values. Consequently, Marktredwitz site-speci IC critical values of Hg concen-
trations in soil were estimated in a toxicologically base way. Risk values for soil con-
tamination by Hg, derived from the QEA single area a s s e ~ ~ m e n t for the land use scenario
I
I
312 Fate and Transport of Heavy Metals in the Vadose Zone
Deduction of site specific assessment values by means of quantitative exposure
assessment (scenario: "Living Area with a House Garden")
Formulation of the question: Ic
Bo
= f(RW) atgivenRwl (ascertained empirically)
I
I
Ir
RW = 0.25
total risk index
GRI
RI pathway
0.5 0.75 1.0
specific risk index for each pathway
RI (oral) Ri (inhalative)
potential daily intake dose POI pathway
Input: Pollutant concentration in contact media taken from
statistical analysiS of site-specific data (90 th percentile,
average value)
c plant = f (c SOil) C plant = constant
....
C indoor air
= constant
c indoor dust = constant

exposure scenario (living area with house garden)
oral: soil little children
plant little children, adults
drinking water not considered in Marktredwitz
inhalative: indoor air and dust little children, adults
mercury concentrations in soil depending on the risk values for the receptor group little
children:
C IOit [mg Hg/kg soil] 33 78 127 176
assessment value (Hg) RPW MSW(I) MSW(II) G
RW (risk value) 0.25 0.5 0.75 1.0
Figure 13.17. Deduction of site-specific assessment values by means of quantitative exposure and
risk assessment. RPW: risk testing value; MSW (I, 11): low- and high frequency examination
threshold values; G: value indicating danger. The calculation of these assessment values was
carried out in cooperation with the Institute Fresenius Ltd., Taunusstein.
"Living Area with a House Garden," served as a basis for the evaluation of this land-use
mode (Figure 13.17).
As the estimated soil values (33 to 176 mg kg-I) were so-called "field means" (FDW)
and incorporated statistical error of 74%, for exposure calculations, the corrected soil
values were used as input for the user/risk receptor group "little
children" were calculated for two versions of i\ut data for Hg in plants: (i) Hg concen-
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 3 I 3
tration in food/fodder plants is proportional to the relevant concentration in soil; and (ii)
constant Hg concentration in plant = 90% of the average for useful plants per field value.
In both versions, concentrations of Hg in all species of useful plants, suspended dust and
air were considered as input data. The critical values estimated for both versions are
presented in Table 13.10.
These site- and use-specific critical values, based on the data up to 1992 and consider-
ing the variability of on-site concentrations, appeared to be surprisingly high and in most
cases significantly exceeded the concentrations of Hg in Marktredwitz soil. These val-
ues, though, are consistent with the results of a human biomonitoring. No markedly
heightened mercury concentrations were found in numerous samples of body fluids (hu-
man blood and urine) taken in 1987 and 1982. The aforementioned critical values were
derived from the toxicological models with a high factor of confidence with respect to
the body transfer pathways, i.e., resorption of the pollutant (Hg). Consequently, only
higher Hg concentrations would cause distinct physical damage. The lack of an adequate
amount of data for children should however, be pointed out (Figure 13.18). Only five of
them were in the age range of 2 to 4 years, and just one of them was checked for Hg
concentrations in soil. For this addressee, a field mean including analytical error of 74%
is below 33 mg Hg/kg. It should be added that only a few of the examined group were
living in heavily contaminated localities, and the information concerning habits and liv-
ing conditions was limited. Another human monitoring conducted in 1995 tried to avoid
these weak points.
The RPW and G values derived for the land-use mode "Living Area with House
Garden" have three basic advantages: (i) they have a toxicological fundament, being
based upon the exposure analysis and toxicologically related Hg-TRD values by Hassauer
et al. (1993); (ii) they enable differentiation between lack of a risk and a possible danger
for risk receptors; and (iii) they consider a site-specific situation in Marktredwitz.
Four critical values derived site-specifically are linked to the different Hg concentra-
tions in soil. The risk value of 0.25 corresponding to 33 mg Hg kg-
1
in soil already con-
siders a heightened site-nonspecific background concentration. It should be emphasized
that the differentiation between RPW and G values has no toxicological reason. Their
aim is just to define the time interval of taking measures. The interpretation of soil values
corresponding to the different risk level is given in Figure 13.19.
On application of these values to the evaluation of the Marktredwitz site based on the
land-use mode "Living Area with a House Garden," some insufficiencies had to be con-
sidered. The most important of them were: (i) lack of a complete set of data for the area
adjacent to CFM, that had to be substituted with average means; (ii) many soil values
were acquired using averaged samples for a whole evaluated area; (iii) the toxicological
evaluation of Hg contamination was subject to change. Therefore, the evaluation of con-
tamination by Hg should be modified accordingly.
Human Biomoniioring Data
Besides models of exposure assessment, a quality-safe risk assessment requires hu-
man biomonitoring data to be incorporated into investigations. These data enable an
estimation of the internal exposure of a l r G m ~ n to ntaminants. To evaluate properly
the internal exposure data originating from the soil co tamination, other sources con-
314 Fate and Transport of Heavy Metals in the Vadose Zone
Table 13.10. Deduction of Site-Specific Assessment Values for the Scenario "Living Area with a
House Garden," Based on Uniform Data; the Assessment Data Underlaid with Gray Were Used
to Evaluate House Gardens in Marktredwitz
C plant = 0,0084 c 5011
(mg Hg kg-
1
soil)
Cplant =
0,016 mg Hg kg-l plant
User Group: Little Children User Group: Little Children
Risk Values Assessment C soil (mg Hg kg-
1
soil)
RW Values Variant 1
0.25 RPW
Hg
a
33
0.5 MSWHg(l) b 78
0.75 MSWHg(1I) 127
1.0 G
Hg
c
176
a RPW: risk testing value.
b MSW (I, II) low- and high-frequency examination threshold values.
C G: value indicating danger.
C soil (mg Hg kg-
1
soil)
Variant 2
34
87
139
191
NUMBffiOF AGE STRUCTURE OF PROBANDS
PROBAMlS BLOOD SAMPLING 428

400
350
300
250
200
150
100
50 6
<4
10
4-7 7-10 10-20 20-50 >50
AGE (YEARS)
Figure 13.18. Age structure of probands of the human biological Hg-monitoring in the surround-
ings of CFM (blood samples taken between 1987 and 1992).
tributing to the exposure have to be considered and differentiated (Table 13.11). The
human biomonitoring data compared to the toxicologically proven reference values serve
as a base for recommendations. The evaluation of human biomonitoring data is still in
progress and subject to a comprehensive statistical analysis when completed. Here, the
partial results are presented and discussed.
The concentrations of Sb in the blood and urine were found to be predominantly at
the edge of a detectable level, and only in a few samples occurred in the range of a few pg
dm -3, although concentrations of Sb in soil were in some areas very high. It can therefore
be concluded that high soil contamination by Sb has not been reflected in its occurrence
in the body fluids (Figure 13.20). The reasons for it were a short Sb half-period of 70
hours in a human body and a of probands in the age when
/
Heavy Metal Contamination in Industrial Areas and Old Deserted 315
Interpretation of the site specific assessment values for the mercury concentration in soil c soli
(= average value of an area plus additionally 74 % for potential measurement errors) deduced
by means of QEA on a uniform data base for the mercury concentrations in the contact media
(analytical data gained from measurements between 1987 and 1992):
scenario:"Living Area with a House Garden" I site: Marktredwitz
risk testing value: RPW
H9
= 33 mg Hg/kg soil
C soli < RPW
Hg
(RW < 0,25)
mercury concentration in soil, up to which no human health hazard above the average is
expected
RPW
H9
S C soli < MSWHg(l) (0,25 S RW < 0,5
mercury concentration in soil, posing a stress to human health over the average without short- or
medium-term need of actions.
low-frequency examination threshold value I MSWHg(l) = 78 mg Hg/kg soil
MSWHg(l) S c soli < MSW
Hg
(lI) (0,5 S RW < 0,75)
mercury concentration in soil from which investigations and health examinations should be
started: an increasing stress for the human health occurs.
low-frequency examination threshold value II MSWHg(lI) = 127 mg Hg/kg soil
MSWHg(lI) SCIoli < G
H9
(0,75 S RW < 1,0)
mercury concentration in soil from which frequent investigations and health examinations should
be started:checking if danger is looming.
value indicating danger: = 176 mg Hg/kg soil
csoll ~ G
H9
( R W ~ 1,0)
mercury concentration in soil from which a high probability of danger for health occurs:
remediation of contaminants impact should be started immediately
Figure 13.19. Interpretation of site-specific risk assessment values for the mercury concentration
in soil.
they ingest soil (crawling infants). Therefore, an important transfer pathway did not
contribute to the exposure.
The problem of a low participation of the most sensitive group of population in blood
sampling is common for almost all investigations, while the participation in urine sam-
pling is definitely higher. The alternative sampling of hair as a sink for Sb was renounced
due to the methodological and time problems.
The results of the blood and urine analysis for Hg were categorized in accordance
with the FRG Health Ministry proposals (Figure 13.21). The measured values appeared
to differ significantly from those obtained in the frameworks of the German extensive
environmental survey (UWS) conducted in 1990 and 1992 (Table 13.12). The Hg con-
tents in blood of probands from the Marktredwitz area were distinctly elevated in com-
parison with the UWS, though none of the probands was classified with the category III.
The preliminary results show therefore, that soil contamination by Hg is represented in
the blood of the inhabitants, while the data on urine did not differ considerably from the
UMS results.
316 Fate and Transport of Heavy Metals in the Vadose Zone
Table 13.11. Factors to Be Considered for Human Biomonitoring"
1. Choice of appropriate parameters
relevant contaminants
toxicokinetics, investigation material
consideration of other important possibilities of exposure with respect to the contami-
nants to be investigated (profession, tobacco smoking, amalgam fillings, etc.)
2. Choice of the probands
criteria of choice
criteria of exclusion
investigation of a group of comparison
criteria of choosing a group of comparison
3. Recording of data relevant for an estimation (questionary)
personal data (age, gender, separately: name and address)
time of living, living conditions
profession(s)
smoking habits
personal customs and behaviors
use of the garden
consumption of the grown vegetables and fruits
4. Criteria of judging the measured data of human biomonitoring
reference values
orientation values
toxicologically derived values of exposure referring to the effect (values of human
biomonitoring)
a According to Ewers and Suchenwirth, 1996.
The presented investigation and evaluation of the deserted industrial area (Site III)
shows that quality-safe assessment can be accomplished with limited efforts, provided
an adequate investigation procedure as well as site- and use-specific exposure assess-
ment is applied.
The complex investigation procedure should comprise detailed historical studies aimed
at identification of possible sources of contamination and characterization of the site;
visualization of data by the geographic information system; and estimation of the results
with use of adequate methods. Of the methods currently in use, models for the Quanti-
tative Exposure Assessment (QEA) appears to be the best fit to the purpose of site- and
use-specific estimation of human risk potential originating from contaminated soils.
Estimation based entirely on the transfer pathways can lead to an overestimation of
the actual exposure; for the reliability of the evaluation, a strong linkage between the
toxicologically proven threshold values for the contaminant uptake and the results of the
exposure assessment is indispensable. As such values are available only for a limited
number of contaminants, the best solution to the problem is a human biomonitoring
which provides empirical data on the actual internal burden of human risk receptors.
REMEDIAL CONCEPTS
The environmental issues exemplified in case studies show the need for a differenti-
ated approach to the selection of appropriate, site- and use-specific remedial actions. On
remedial action analysis, based on estimated extent of contamination by heavy metals
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 317
MJM BER OF
PROBANCS
120
100
80
60
40
20
a
119
It)
o
NUMBEROF
PROBANJS
90
80
70
60
50
40
30
20
10
a
0.1
3
-
a
1
a a a a a a a a a
N It) Mit)...,. It) It) It)
N ..; "4' iii
ANTIMONY CON::ENTRA TION IN BlOOD [II9/L]
a a a
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
ANTIMONY CONCENTRATION IN URlt [1I91L]
Figure 13.20. Antimony concentrations in blood and urine samples.
a a
1
a
a
1.1 1.2
The categories for evaluating the contents of heavy metals in human samples (Krause et aI.,
1987), edited by the Institute of Water, Soil and Air Hygiene, are referred to the measured
values in blood and urine. Three categories have been distinguished:
Category I in nonrisky value
Category II
Category III
higher value, no evident health risk, examination is recommended
distinctly elevated value, a health risk cannot be excluded in the long run,
cleanup is necessary to remove or at least reduce the sources of a risk
Figure 13.21. Estimation based on the guide values for human
"\
318 Fate and Transport of Heavy Metals in the Vadose Zone
Table 13.1 2. Comparison Between the Human Biomonitoring in Marldredwitz in 1966 and the
Environmental Survey (UWS), 1990-1992
Number of Measured
Studies Persons Value Category I Category II Category III
Hg in blood <3
3-10 >10
(lJg L-')
CFM-Investigation 134 83,6% (112) 16,4% (22) 0% (0)
UWS (1990-1992) 3958 97,9% 2,1% 0,03%
Hg in urine <5 5-20 >20
(lJg L-')
CFM-Investigation 134 100% (134) 0% (0) 0% (0)
UWS (1990-1992) 3958 96,8% 3,0% 0,2%
and assessed requirements for such action (further use of the land, groundwater rotec-
tion) , a set of efficient and cost-effective options elaborated also by the auth rs was
considered. These options are in line with the current trend to developing metho s, which
clean up the soil without destroying its properties and fertility, be applied situ. Such
methods are generally very cost-effective. To a great extent, they are based on the vul-
nerability to mobilization of metals bound onto mobile or easily mobilizable phases. Re-
moval of metals bound in these fractions would render the soil harmless in the most
effective way.
Basic options that have been analyzed with regard to the cost-effective application for
trace metal control in the large-area contaminated sites was polluted soil sanitation by
leaching methods. Studies conducted within this task comprised the selection of the novel
effective extractants. The reuse of waste material as extractant was also considered.
Here, only general concepts are presented. The detailed discussion of methods and data
are subject to being published elsewhere, part of them having already been published
(Fischer et al., 1994; Leidmann et al., 1995).
With regard to the remedial actions for decontamination of soils and prevention of
contaminated water infiltration from the waste layer to the groundwater, an application
of agricultural and hydrolyzed food engineering residues (silage effluents, residues of a
brewing industry, slaughtering offal) have been found promising. These residues can be
used either directly (Leidmann et al., 1995) or as raw materials for preparation of
extractants for the sanitation of metal polluted soils by leaching method (Fischer et al.,
1994). The extraction rates obtained with grass silage juice from two adsorbents (bento-
nite and peat) were: Cd 74.7%, Zn 55.7%, Cu 53.5%, Ni 38.9%, Cr 12.7%, and Pb 8.9%.
The efficiency of metal extraction with hydrolysate applied to contaminated soils was
adequate: Cu 50.3% and Ni 38.7%, at initial concentration of metals in soils 279 mg Cu
kg-
I
and 54 mg Ni kg-I.
Many other natural or reused materials are now under investigation by numerous
authors, in order to apply them either as metal extractants or adsorbents. These options
offer the benefit of both the decontamination of soils and leachates and the treatment/
disposal of the waste material in an ecologically tolerable way.
Growing interest in mild cleanup technologies results in extending their practical use.
Many of these technologies are currently a routine practice, while others need more
research and efforts to make them practically applicable.
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 319
- -- ----_ .. ----........... - ... -----.-------
SUMMARY
From the presented case studies, investigations, and evaluation concepts, several gen-
eral conclusions can be derived, which are summarized as follows. Optimization of moni-
toring programs for heavy metal contaminated site evaluation should be based on a
use-specific and site-specific approach. Among the required data on factors and criteria
for evaluation of contaminated sites, vertical distribution of a metal in the soil profile,
species mobility/availability to the risk receptors and its possible transformations in time
are to be considered for estimation of scientifically proven threshold values. The site-
and use-specific, exposure-based evaluating model for assessment of the human risk
potential originating from old landfills and deserted industrial sites appears to provide a
reliable quality-safe estimation of the actual risk resulting from a contaminated site. The
remedial methods using appropriate waste material as heavy metal extractant in a con-
trolled, environmentally friendly way are promising cost-effective options for decon-
tamination/preventive actions preserving soil properties and fertility.
REFERENCES
ASTM Designation: D 5233-92. Standard Practice/or Nitric Acid Digution 0/ SoLid WlUte, 19
133-135.
Barth, D.S. and B.J. Mason. SoiL SampLing QuaLity AJdurance UderJ Guide. Cooperative Agreement
CR 81050-01, U.S. EPA Environmental Monitoring Systems Laboratory, Office of R&D, Las
Vegas, NY, March 1984, p. 104.
Chief Statistical Office. Environment Protection 199J. In/ormation and StatuticaL Data, Warsaw, 1993,
p. 450 (in Polish).
Chief Statistical Office. Environment Protection 1995. Information and Statistical Data, Warsaw,
1995, p. 490 (in Polish).
Eikmann, Th. and A. Kloke. Nutzungs- und schutzgutbezogene Orientierungswerte fur Schad-
Stoffe in Boden-Eikmann-Kloke-Werte, Kennzahl3590. BOdendchutz - erganzharu Handhuch,
D. Rosenkranz, G. Bachmann, G. Einsele, and H.-M. HarreB, Eds., Erich-Schmidt-Verlag 14.
LEg. Xl93, Berlin, 1993 (in German).
Eikmann, Th., A. Kloke, and H.P. Luhr. Grundlagen und Wege zur Ermittlungvon Bodenwerten
fur das Drei-Bereiche System, in IWS-Schriftenreihe Band 13: Ahleitung von Sanierungdwerten/iir
Icontaminierte BO'den, Erich Schmidt Verlag, Berlin, 1991, pp. 279-360 (in German).
Ewers, U. and L. Viereck-Goethe. Ableitung von wissenschaftlich begruendeten nutzungs- und
schutzgutbezogenen Pruefwerten fur Bodenverunreiningungen, Bericht im Au/trag dU BayStMLU
in Vergindung mit der Arheitgruppe Prue/werte dU LAGA-AuddchUdded "ALtlMten. "Hygiene-Institut des
Ruhrbereites, Gelsenkirchen, 1993, p. 79 (in German).
Ewers, U. and R. Suchenwirth. Expositionsabschatzung: Human-Biomonitoring Modell-
rechnungen. UWSF-Zeitdchrift/iir UmweLtchemie und OlcotoxilcoLogie 8(4), pp. 213-220, 1996 (in
German).
Fischer, K., P. Riemschneider, D. Bieniek, and A. Kettrup. Food engineering residues: Amino
acid composition of hydrolysates and application for the decontamination of metal polluted
soils. FrueniUd J. AnaL. Chem. 350, pp. 520-527, 1994.
Forstner, U. Riverine and estuarine sedimentation of pollutants and leaks from sludges and
wastes: Analytical, prognostic, experimental and remedial approaches. Land Degradation RehahiL.
4, pp. 281-296, 1992.
GKSS/LAGA. Unterduchung ded Gefahrdungdpotentia& von queclcdiLherlcontaminierten Standorten in
Bayern. Olctoher 1992 Forschungsbericht im Auf trag des Bayerische Landesamt fur
320 Fate and Transport of Heavy Metals in the Vadose Zone
Umweltschutz. GKSS-Forschungszentrum Geesthacht und Landesgewerbeanstalt Nurnberg,
1993 (in German).
Gupta S.K., M.K. Vollmer, and R Krebs. The importance of mobile, mobilizable and pseudo total
heavy metal fractions in soils for three-level risk assessment and risk management, Sci. TotaL
Environ. 178, pp. 11-20, 1996.
Harrison, RM., D.P.H. Laxen, and S.J. Wilson. Chemical associations oflead, cadmium, copper
and zinc in street dust and roadside soils. Environ. Sci. Techno!. 15, pp. 1378-1383, 1981.
Hassauer, M., F. Kalberlah, J. Oltmanns, and K. Schneider. Basisdaten Toxikologie fur
umweltrelevante Stoffe zur Gefahrenbeurteilung von Altlasten. UBA-Berichte 4/93, E. Schmidt,
Berlin, 1993 (in German).
Hempfling, R., L. Heming, and S. Stubenrauch. Quantitative Exp0.JitiolLJaIJdchiitzung und-bewertung
fur dad Umfet;} der Chemi.lchen Fabrik Marktredwitz auf Badi.l von GSF-au/bereiteten Daten. Februar
1994. Bericht im Auf trag des GSF - Forschungszentrums fur Umwelt und Gesundheit. Institut
Fresenius GmbH, Taunusstein, 1994 (in German).
Hempfling, R, S. Stubenrauch, U. Mayer, and S. Simmleit. Fallbeispiele fur die Altlastenbewertung
mittels UMS, inALtfMtenbewertung-DatenanaLY.Je und Gejahrenbewertung, S. SChUlte-,stede, R
Freitag, A. Kettrup, and W. Fresenius, Eds., ECOMED-Verlag, Landsberg, 199 (in Ger-
man).
Kabata-Pendias, A. and H. Pendias. Trace Element.J in Soil! and Plant.J, 2nd ed., CRC ress, Inc.,
Boca Raton, FL, 1992, p. 365.
Kaszycki, C.A. and G.E.M. Hall. Application of Phase Selective ntial Extraction
Methodologies in Surficial Geochemistry, in EXTECH I: A MuLtidi.lcipLinary Approach to MadJive
SuLphide Re.Jearch in the RUJty Lake-Snow Lake GreenJtone Belt.J, Manitoba. GeowgicaL Survey of Canada,
BuLL. 426, G.F. Bonham-Carter, A.G. Galley, and G.E.M. HalL Eds., 1996, pp. 155-168.
Kersten, M. and U. Forstner. Chemical fractionation of heavy metals in anoxic estuarine and
coastal sediments. Wat. Set: TechnoL. 18, pp. 121-130, 1986.
Kersten, M. and U. Forstner. Assessment of Metal Mobility in Dredged Material and Mine
Waste by Pore Water Chemistry and Solid Speciation, in Chemi.ltry and Biowgy of SoLid Wadte.
Dredged MateriaL and Mine TaiLingJ, W. Salomons and U. Forstner, Eds., Springer-Verlag,
Berlin/Heidelberg, 1988.
Konietzka, Rand H.H. Dieter. Kriterien fur die Ermittlung gefahrenverknupfter chronischer
Schadstoffzufuhren. GeJundheitJweJen 56, pp. 21-28, 1994 (in German).
LAGA, Erfassung und Sanierung von Altlasten. Mitteilungen der LA GA. Bd.35, E. Schmidt Verlag,
Berlin, 1991 (in German).
Leidmann, P., K. Fischer, D. Bieniek, and A. Kettrup. Chemical Characterization of silage
effluents and their influence on soil bound heavy metals. Intern. J. Environ. Ana!. Chem. 59, pp.
303-316, 1995.
McGrath, D. Application of single and sequential extraction procedures to polluted and unpol-
luted soils. Sci. TotaL EflIJiron. 178, pp. 37-44, 1996.
McLaren, R.G. and D.V. Crawford. Studies on soil copper. I. The fractionation of copper in soils.
J. Soil Sci. 24, pp. 172-181, 1973.
Prudent P., M. DomeizeL and C. Massani. Chemical sequential extraction as decision-making
tool: Application to municipal solid waste and its individual constituents. Sci. TotaL Environ.,
178, pp. 55-61, 1996.
Quevauviller, Ph., Ed. Leaching/extraction tests for environmental risk assessment, Special issue,
Sci. TotaL Environ. 178, p. 145, 1996.
Simmleit, N., S. Stubenrauch, U. Mayer, and R Hempfling. UMS-Modell: Einzelflachenbezogene
quantitative Expositionsabschatzung und Gefahrenbeurteilung von Altlasten, in
ALtfMtenbewertung-DatenanaLY.Je und GejahrenbewertUflg. S. Schulte-Hostede, R Freitag, A. Kettrup,
and W. Fresenius, Eds., ECOMED-Verlag, Landsberg, 1997 (in German).
Heavy Metal Contamination in Industrial Areas and Old Deserted Sites 321
---------"-"-----"--"-" ""---- - - - - - " - - - ~ - - - -"-----"-
Stubenrauch, S., R. Hempfling, N. Simmleit, T. Mathews, and P. Doetsch. Abschatzung der
Schadstoffexposition in Abhangigkeit von Expositionsszenarien und Nutzergruppen. I.
Grundlagen und Vorschlage zur Ableitung von Aufnahmeraten am Beispiel von Trinkwasser,
UWSF - Z Umweltchem. Okotox. 6, pp. 41-49, 1994a (in German).
Stubenrauch, S., R. Hempfling, S. Simmleit, T. Mathews, and P. Doetsch. Abschatzung der
Schadstoffexposition in Abhangigkeit von Expositionsszenarien und Nutzergruppen. II.
Vorschlage fur orale Aufnahmeraten von Boden, Badewasser und Nahrungsmitteln des
Eigenbaus, UWSF - Z Umweltchem. Okotox. 6, pp. 165-174, 1994b (in German).
Tack, F.M. and M.G. Verloo. Impact of single reagent extraction using NH
4
0Ac-EDTA on the
solid phase distribution of metals in a contaminated dredged sediment. Sci. Total Environ. 178,
pp. 29-36, 1996.
Tessier, A., P.G.C. Campbell, and M. Bisson. Sequential extraction procedure for the speciation
of particulate trace Metals. Anal. Chem. 51, pp. 844--851, 1979.
Tessier, A. and P.G.C. Campbell. Comment on Pitfalls of Sequential Extractions by P, M.V. Nirel
and F.M.M. MoreL Water Red., 24, pp. 1055-1056, 1990. Water Red., 25. pp. 115-117, 1991.
Twardowska, I. Areas of long-lasting anthropopression: Assessment and monitoring of pollution
potential to soil and ground water. SPIE, 2504, pp. 253-264, 1995. \
Twardowska, I. and J. Kyziol. Binding and chemical fractionation of heavy metals in typical ptat
matter. FredenilM J. AnaL Chem. 354, pp. 580-586, 1996. i
Ure, A.M. Single extraction schemes for soil analysis and related applications, in Special Istue:
Harmonization of Leaching/Extraction Tests for Environmental Risk Assessment/ Ph.
Quevauviller, Ed., Sci. Total Environ. 178, pp. 3-10, 1996.
WH 0 (World Health Organization). Environmental Health Criteria 101, Methylmercu .
natwnal Programme on Chemical Safety. World Health Organization, Geneva, 1990.
WH O. Environmental Health Criteria 118, Inorganic Mercury. IPCS, I nternatwnal Programme on Chemi-
cal Safety. World Health Organization, Geneva, 1991.
Xiao-Quan, S. and C. Bin. Evaluation of sequential extraction for speciation of trace metals in
model soil containing natural minerals and humic acid. Anal. Chem. 65, pp. 802--807, 1993.
INDEX
A
Acid ammonium acetate-EDTA
(AAAc-EDTA) extraction 253,
256, 266, 260-262, 267-269
Acid-base changes in the rhizosphere
229-232
Acid forest soils 29-55
Aging 5, 6, 11, 19, 20
Aliphatic acids exuded by roots 230
ALnUd 231
Aluminum 20, 23, 24, 41, 45, 185,
225, 232, 233, 240
effect on roots 239
hydrous oxides 29, 61
lomc 207
Aluminum hydroxide 29, 61
Aluminum oxide 13, 14, 24, 110, 181
Amino acids 230, 234
Ammonium ion 231
Ammonium nitrate 127
Amphiboles 239, 240
Antimony 300, 301, 302, 303, 304,
306
in the blood and urine 314, 317
Aromatic acids exuded by roots 230
Arsenate 10, 12
Arsenic 10, 301, 302
deposition 29
B
Barrier capacity of a surface soil layer
292-293
Batch versus flow-through systems
71-74
Bentonite 108
Beryllium
distribution coefficients for 4
Bicarbonate 186
Binding strength 293-298
Biotite transformation to kaolinite
240
Biphasic sorption reactions 13
Breakthrough curves (BTCs) 76,
77-78, 83, 92, 100, 103, 104,
105, 133
Brunauer-Emmett-Teller (BET)
isotherm model 19
C
Cadmium 6, 9, 13, 19, 23, 59-85, 91,
92, 100, 102, 160, 164, 166, 168,
169, 171, 209, 253, 257, 263, 264,
266, 267, 269, 278, 279, 280, 282,
291, 292, 302, 318
desorption kinetics 82-83
isotropic exchangeability 5
retention 96
sorption 63
reversibility 80-83
uptake by soils 4
Cadmium-calcium exchange isotherm
101
Cadmium carbonate 22, 23
Cadmium sulfate 160, 161
324 Fate and Transport of Heavy Metals in the Vadose Zone
Cadmium sulfide 161
Calcareous agricultural soils 177-197
Calcium 23, 68, 100, 152, 164, 165,
167, 169, 194
IOmc 108, 112, 114, 115, 121, 152,
161, 186, 205, 207
Calcium carbonate 186, 204, 205
deposition 29
fractionation 296
in sludge 252, 255, 268
in surface soils 203
mobile fraction 285
partitIOning 295
solubility 162
sorption 210
USEP A drinking water quality
standard 59
Calcium chloride 68, 70, 79, 112,
127, 132
Calcium nitrate 66, 68, 70, 80, 112,
127, 132
Calcium oxide 181
Carbohydrates exuded by roots 230
Carbon 169, 171
flux 229
transformations 234
Carbon dioxide 167, 189, 231
loss 186
Cardiomyopathy 147
CaJuarina 231
Cesium 2,6
desorbed fraction 5
sorption 14, 15
Chelate effect 15
Chlorine ions 64-66, 79
Chromium 19, 92, 100, 164, 234,
253, 257, 263, 264, 266, 269,
277, 279, 280, 291, 318
deposition 29
in sludge 252, 255, 259, 268
sorption 4
Chromium hydroxides 31
Clover 240
Cobalt 3, 6, 8, 19, 20, 22, 23, 65, 81,
160, 162, 234, 302
deposition 29
desorption 5, 7
distribution coefficients 4
isotropic exchangeability 17
sorption 16-17
Cobalt hydroxide 20, 31
Complexation capacity 208
Convective-dispersion transport
equation (CDE) 99
Copper 13, 14, 92-94, 97, 102-104,
105, 160, 164, 168, 170, 185, 206,
207, 209, 210, 212-219, 233, 278,
279, 280, 292, 296, 301, 302, 318
adsorption by calcium carbonate
204
bioavailability 127-145
fractionation 283
in sludge 259
in surface soils 203
ionic 114, 205-207
mobile fraction 285
mobility 127-145
oxidized states 234
retention 96, 107-122
solubility 162
sorption on noncrystalline aluminum
oxide 4
Copper-hydronium exchange
equilibrium 122
Copper-magnesium exchange isotherm
103
Copper-sodium exchange 109
Corn (Zea mayd) 231, 240, 255
Cupric chloride 108
Cupric hydroxide 108, 110, 130
Cupric nitrate 112
Cupric oxide 149, 150, 151, 152, 153,
154, 155, 156
Cupric sulfate 103, 130, 161
D
Dendrobaena veneta 141, 142, 144
Diffuse double-layer 30
Diffuse ion association 30-32
Diffuse ion complex 11
Diffusion-controlled kinetic reactions
8-24
Dissolved organic carbon (DOC)
131-137, 145, 148-150, 152,
154, 210
Douglas fir 231
Drinking water 307
E
Earthworms 139-142
EDTA 115, 122, 194, 195, 207 See
auo Acid ammonium acetate-
EDT A extraction
Electronegativity 12, 218
Electron spin resonance (ESR)
spectroscopy 13
F
Fatty acids exuded by roots 230
Feldspars 239, 240
Ferric hydroxide 110
Ferric oxide 159, 161, 181
Fick's second law 9-10
Flame atomic absorption spectrometer
132
Flow-through systems 73
compared to batch systems 84
Foliar manganese content of conifers
232
Food crops 304
Food engineering residues 318
Forest soil 264
French beans 232
Freundlich equation 93, 102, 119,
120, 121 See auo Two Species
Freundlich equation; van
Bemmelen-Freundlich equation
Fulvic acids (FA) 110-111, 115-119,
121, 168, 241
G
Galena 189
Gardens, contaminated 304, 307-315
GEOCHEM 64, 149-150, 154, 155
Geographical information systems
(GIS) 302, 304
Gibbs free energy function 10
Gibbsite 20
Goethite 70, 108, 164
Grass silage 318
Grid model 79
Gypsum 190
H
Index 325
High resolution transmission electron
micrography (HRTEM) 23
Hordeum vulgare 233
Human biomonitoring data 304,
313-316
Human risk potential assessment
304-316
Humic acids (HA) 5, Ill, 195
Hydride generation atomic absorption
spectrometry (HG-AAS) 148,
149, 150, 151, 152, 154
Hydrogen 233
effect on roots 239
IOmc 114, 115, 121, 164, 167, 168,
171, 172, 189, 232
release 244
Hydrous ferric oxide (HFO) 6
Hysteresis 5, 6, 8, 96
I
Indoor air 304, 310-312
Inner-sphere surface complexation
11-13, 15, 30, 32, 61-62
Ion chromatography 41
Ion exchange retention 100-105
Ionic radius 8, 12, 19, 218
Ionic strength 12, 14, 62-63, 66,
67-69, 79, 84, III
Iron 8, 19, 41, 45, 160, 164, 165, 167,
169, 188, 189, 195, 212, 215, 233,
240, 276, 291, 292
content of the shoot and roots of
dwarf French beans 232
distribution coefficients 4
hydrous oxides 29
Iomc 161, 185, 207
oxidized states 234
preCIpItation 162
solubility 162
Isomorphic substitution 12
328 Fate and Transport of Heavy Meta!s in the vadose Zone
Siderite 162
Siderophores 230, 234
Silicates 182
Silicon 22, 185
backscattering 19
hydrous oxides 29, 61
Silicon dioxide 181
Silicon hydroxide 29, 61
Sludge See Sewage sludge
Sodium 65, 109, 150, 152
lOmc 112, 121
Sodium chloride 66
Sodium nitrate 66, 127
Soil
depth 137
formation 178-180
fractionation 236-239
Soil-plant systems 160-161
Solid-solution ratio 71-73
Sorption kinetics of trace elements
1-25
Soybean 231, 240
Stack emissions 274-300
Sterols exuded by roots 230
Strontium 9
Sulfate 64, 150, 152, 182, 188
Sulfides 182, 185
formation 172
insoluble 161-162
Sulfur 228, 233
Surface precipitation 18-24
T
Tartaric acid (TA) 115, 121
Tartrate 115-119, 121
Tin 302
Titanium 22
Titanium dioxide 22, 23, 181
Trees 232-235
growth and rhizosphere chemistry
242
Two Species Freundlich (TSF)
equation 129-136, 138, 143
v
Valence charge 12
van Bemmelen-Freundlich equation
34, 39-40
van Bemmelen-Freundlich isotherms
42-46
Vanselow equation 109, 120-122
Vermiculite 240
Volcanic rock 182
chemical composition 182
w
Waste sites 273-319
Wastewater 211-218, 252
Weathering in soil 239-242
Wheat 240
x
X-ray absorption fine structure
(XAFS) spectroscopy 9-10, 11,
13, 19-20, 24
X-ray absorption spectroscopy (XAS)
12-13, 19-20
X-ray photoelectron spectroscopy
(XPS) 9
Z
Zinc 9, 13, 38, 65, 79, 82, 100, 102,
160, 161, 164, 166, 168-171, 185,
188, 204-207, 209, 212, 213, 215,
217-219, 233, 253, 257, 263-267,
269, 277, 279-280, 291-295, 301,
302, 318
content of the shoot and roots of
dwarf French beans 232
distribution coefficients 4
in sludge 252, 255, 259, 268
mobile fraction 285
retention 96
solubility' 162
surface soils 203
uptake by soils 4
Zinc sulfide 161

Anda mungkin juga menyukai