kinetics of trace metals sorption on and desorption from soils developing predictive models

LIST OF TABLES Table 2.1 Linear forms of kinetics equations adapted from Sparks, 2003; detailed description of the symbols can be found in related literature ....41 Table 4.1 Soil charac

KINETICS OF TRACE METALS SORPTION ON AND DESORPTION FROM

SOILS: DEVELOPING PREDICTIVE MODELS

by Zhenqing Shi

A dissertation submitted to the Faculty of the University of Delaware in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Civil Engineering

Winter 2006

Copyright 2006 Zhenqing Shi All Rights Reserved

UMI Number: 3205426

3205426 2006

UMI Microform Copyright

All rights reserved This microform edition is protected against unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company

300 North Zeeb Road P.O Box 1346 Ann Arbor, MI 48106-1346

by ProQuest Information and Learning Company

KINETICS OF TRACE METALS SORPTION ON AND DESORPTION FROM

SOILS: DEVELOPING PREDICTIVE MODELS

by Zhenqing Shi

Conrado M Gempesaw II, Ph.D

Vice Provost for Academic and International Programs

I certify that I have read this dissertation and that in my opinion it meets

the academic and professional standard required by the University as a

dissertation for the degree of Doctor of Philosophy

Signed:

Herbert E Allen, Ph.D

Professor in charge of dissertation

I certify that I have read this dissertation and that in my opinion it meets

the academic and professional standard required by the University as a

dissertation for the degree of Doctor of Philosophy

Signed:

Dominic M Di Toro, Ph.D

I certify that I have read this dissertation and that in my opinion it meets

the academic and professional standard required by the University as a

dissertation for the degree of Doctor of Philosophy

Signed:

Chin Pao Huang, Ph.D

I certify that I have read this dissertation and that in my opinion it meets

the academic and professional standard required by the University as a

dissertation for the degree of Doctor of Philosophy

Signed:

Donald L Sparks, Ph.D

Member of dissertation committee

ACKNOWLEDGMENTS

I express my greatest gratitude to many individuals who have helped and encouraged me to complete this dissertation

First and foremost, I wish to express my thanks to my advisor, Dr Herbert

E Allen, for his guidance and encouragement not only in this study but also my professional career I would have not been able to accomplish this research without his insightful advice and extraordinary support The knowledge and experience I have learned is much more than what I had expected

I am very grateful to Dr Dominic Di Toro who has guided me with most

of the modeling work in the past three years This research benefits a lot from his insight, broad vision and patient instructions He leads me into the amazing world of computational environmental science

My appreciation is extended to other committee members, Dr Chin Pao Huang, and Dr Donald L Sparks, and also to Dr Alexander A Ponizovsky for their assistance and invaluable advice during this study

Special thanks go to Dana Crumety for her continuous support throughout the past years I also want to thank Doug Baker, Mike Davison and Eric Eckman for their technical assistance

I appreciate the US Environmental Protection Agency and International Copper Association for the financial support of the research

Additionally, many thanks are given to my friends and colleagues, Tao Cheng, Sagar Thakali, Dave Metzler, Vaishnavi Sarathy, Kevin Rader, Yuefeng Lu,

and Yujun Yin, for their help I also thank Dr Steve Lofts, Centre for Ecology and Hydrology, Lancaster, United Kingdom, for his help on the WHAM VI program

Finally, my thanks and love are expressed to my dear parents, Qishuang Hou and Fuchu Shi, and my elder brother, Zhenshan Shi, who always encourage me and offer understanding in my journey of pursuing the doctoral degree Wherever I

go, they are always there to give me love and support

TABLE OF CONTENTS

LIST OF TABLES ix

LIST OF FIGURES x

ABSTRACT xvii

Chapter 1 INTRODUCTION 1

1.1 Background and Significance of the Study 1

1.2 Objective 6

References 9

2 LITERATURE REVIEW 12

2.1 Trace Metal Content in Soils and Reactivity 12

2.1.1 Copper 12

2.1.2 Zinc 13

2.1.3 Other Metals 14

2.2 Soil Properties Affecting Trace Metals Sorption and Desorption 16

2.2.1 Metal Concentration in Soils 16

2.2.2 Soil Organic Matter 18

2.2.3 Clay Minerals and Metal Oxides 20

2.2.4 Assessment of the Role of Different Soil Components 22

2.3 Solution Chemistry Affecting Trace Metals Sorption and Desorption 30

2.3.1 pH 30

2.3.2 DOM 32

2.3.3 Dissolved Calcium 34

2.3.4 Other Factors 35

2.4 Kinetic Behavior of Trace Metals Sorption and Desorption on Soils 36

2.5 Methods Used to Study Kinetics of Metal Sorption and Desorption on Soils 39

2.6 Modeling Kinetics of Metals Sorption and Desorption on Soil Particles 40

2.6.1 Kinetics Models Used for Metal Reactions with Soils 40

2.6.2 First-order Kinetic Equation 43

2.6.3 Two-site Model 45

2.6.4 Multi-site Model 48

2.6.5 Equilibrium Constraint for Kinetics Model 51

2.6.6 Theory and Modeling of Stirred-flow System 52

2.7 Conclusions on the Literature Review 54

References 56

3.1 Model Formulation 67

3.1.1 Reaction Chemistry 67

3.1.2 Basic Kinetics Model Formulations 70

3.1.3 Instantaneous Equilibrium (IE) Model 73

3.1.4 pH Dependency 75

3.1.5 Numerical Solutions for the Kinetics Model 77

3.2 Evaluation of the Kinetics Model 79

3.3 Obtaining Model Parameters 84

3.4 Relationship between Kinetics Model and Equilibrium Model 85

References 87

4 MODELING KINETICS OF CU AND ZN RELEASE FROM SOILS 88

4.1 Introduction 88

4.2 Materials and Methods 89

4.2.1 Chemicals and Samples 89

4.2.2 Soil Spiking 93

4.2.3 Stirred-flow Experiment 93

4.2.4 Experimental Matrix 97

4.3 Modeling Methods 98

4.4 Results and Discussion 99

4.4.1 Leaching Experiment 99

4.4.2 Modeling Zn Release Kinetics 115

4.4.3 Modeling Cu Release Kinetics 133

4.5 Model Tests and Assessments 145

4.6 Implications 151

References 158

5 AN ORGANIC CARBON NORMALIZED KINETICS MODEL TO PREDICT ZN SORPTION AND DESORPTION ON DIFFERENT SOILS 162

5.1 Introduction 162

5.2 Materials and Methods 164

5.2.1 Samples and Chemicals 164

5.2.2 Sorption and Desorption Kinetics Experiments 167

5.3 Model Description 168

5.4 Results and Discussion 170

5.4.1 Effect of Soil Properties 170

5.4.2 Effect of pH and Zn Loadings 183

5.5 Model Assessment and Implications 190

5.5.1 Evaluation of the Rate Coefficients 190

5.5.2 The Roles of Two Sites for Zn Sorption and Desorption Kinetics 191

5.5.3 Assessment of Model Simplifications and Limitations 198

References 205

6 DEVELOPMENT OF THE PREDICTIVE MODEL FOR KINETICS OF CU

SORPTION/DESORPTION ON SOILS 208

6.1 Introduction 208

6.2 Materials and Methods 210

6.2.1 Samples and Chemicals 210

6.2.2 Sorption and Desorption Kinetics Experiments 210

6.3 Model Description 211

6.3.1 Basic Model Formulation 211

6.3.2 Incorporation of Freundlich Equation 212

6.3.3 Incorporation of WHAM VI 217

6.4 Results and Discussion 220

6.4.1 IE Model 220

6.4.2 Freundlich Equation Based Kinetics Model 225

6.4.3 WHAM VI Based Kinetics Model 232

6.5 Model Assessment 251

6.5.1 The Role of Fe(III) and Al(III) 252

6.5.2 The Role of Active Organic Matter 257

6.5.3 The Effect of Ca 259

6.6 Other Metals 262

References 264

7 CONCLUSIONS AND RECOMMENDATIONS 266

Conclusions 266

Recommendations 269

Appendix A KINETICS MODEL FOR CU AND ZN RELEASE FROM TWO SOILS 272

A1 Zn Kinetics Model 272

A2 Cu Kinetics Model 275

B THE ORGANIC CARBON NORMALIZED ZN KINETICS MODEL 279

C WHAM VI BASED CU KINETICS MODEL 283

C1 General Information 283

C2 Instruction on Using the Computer Program 284

C3 Comments on the VB Code 285

C4 Definitions of Parameters in the EXCEL File 288

C5 Running the Model 290

LIST OF TABLES

Table 2.1 Linear forms of kinetics equations (adapted from Sparks, 2003;

detailed description of the symbols can be found in related literature) 41

Table 4.1 Soil characteristics 91

Table 4.2 Percent of Cu and Zn release at different pH and DOM concentrations in the 5-hour leaching experiment* 115

Table 4.3 Model fitting parameters for Zn release kinetics using kinetics model 127

Table 4.4 WHAM VI calculated K p at different pH 136

Table 4.5 Model fitting parameters for Cu release kinetics using kinetics model 138

Table 5.1 Soil properties 166

Table 5.2 Model parameters for different soils at pH 5.5 171

Table 5.3 Model parameters 178

Table 6.1 Model parameters for the Freundlich equation based kinetics model 232

Table 6.2 WHAM input parameters at pH 5.5 235

Table 6.3 Model fitting parameters for WHAM VI based kinetics model 235

Table A1 Definitions of Columns in EXCEL File Zn_Matapeake.xls 273

Table A2 Definitions of Columns in EXCEL File Cu_Matapeake.xls 276

Table B Definitions of Columns in EXCEL File Zn_OC.xls 280

Table C Definitions of Columns in EXCEL File Cu_WHAM.xls 288

LIST OF FIGURES

Figure 2.1 Schematic picture of metal reactions in soils and solutions in WHAM

(from Lofts and Tipping, 2000) 27 Figure 2.2 Diagram of a multireaction model (adapted from Amatcher, 1991) C

is the solute concentration in the solution, Si are the solute

concentrations in the soils, and ki are the rate coefficients 51 Figure 4.1 A device for stirred-flow experiments 94

Figure 4.2 Effect of pH (values are presented in the legend) on Cu desorption

from spiked (a) Matapeake and (b) Codorus soils .101 Figure 4.2 Continued .102 Figure 4.3 Effect of pH (values are presented in the legend) on Zn desorption

from spiked (a) Matapeake and (b) Codorus soils .105 Figure 4.3 Continued 106 Figure 4.4 Effect of DOM (concentrations are presented in the legend) on Cu

desorption from (a) spiked Matapeake soil and (b) Codorus soil

(pH=5.5) .108 Figure 4.4 Continued 109

Figure 4.5 Effect of flow rate (values are presented in the legend) on (a) Cu and

(b) Zn desorption from spiked Matapeake soil (pH=5.5) 112 Figure 4.5 Continued 113 Figure 4.6 Kinetics of Zn release from spiked Matapeake soil (pH = 5.5) (a)

Stop-flow experiment (the flow stop and restart time was indicated in the experimental part); (b) Different flow rates with two-site model Flow rate values are annotated in the figure Lines are calculated with the model .119 Figure 4.6 Continued .120 Figure 4.7 Two-site model simulation of kinetics of Zn release from spiked

Matapeake soil (pH = 5.5) at different flow rates Flow rate values are

Figure 4.8 Effect of pH (values are annotated in the figure) on the kinetics of Zn

release from spiked (a) Matapeake soil and (b) Codorus soil (Q = 1 mL min-1) Solid lines are calculated with the two-site kinetics model and dash lines are calculated with IE model .124 Figure 4.8 Continued .125 Figure 4.9 Kinetics of Zn release from spiked Matapeake soil at different flow

rates with IE model fittings (pH = 5.5) Flow rate values are annotated

in the figure Lines are calculated with the model .129 Figure 4.10 Comparison of Zn partition coefficients calculated from different

approaches (indicated in the figure) .132 Figure 4.11 Kinetics of SOM dissolution from the Matapeake soil (pH = 5.5)

Solid line is the SOM dissolution model calculation .135 Figure 4.12 Effect of DOM (values are annotated in the figure) on the kinetics of

Cu release from spiked (a) Matapeake soil and (b) Codorus soil (pH = 5.5 and Q = 1 mL min-1) Solid lines are calculated with the kinetics model and dash lines are calculated with IE model .139 Figure 4.12 Continued .140 Figure 4.13 Effect of flow rates (values are annotated in the figure) on the kinetics

of Cu release from spiked Matapeake soil (pH = 5.5) Solid lines are calculated with the kinetics model and dash lines are calculated with

IE model .141 Figure 4.14 Effect of pH (values are annotated in the figure) on the kinetics of Cu

release from spiked Matapeake soil (Q = 1 mL/min) The effect of

different pre-equilibration times is shown on the plot for pH = 5.5

Solid lines are calculated with the kinetics model and dash lines are

calculated with the IE model 142

Figure 4.15 Comparison of Cu partition coefficients calculated from different

approaches (indicated in the figure) .144 Figure 4.16 Reaction rates of different processes for Zn release from the

Matapeake soil at pH 5.5 and flow rate 1 mL/min 147

Figure 4.17 Changes of particulate Zn concentrations in (a) fast and (b) slow sites

simulated using the two-site kinetics model for the Matapeake soil (pH

= 5.5) The time scales are different for two figures .149 Figure 4.17 Continued .150

Figure 4.18 Release of (a) Cu and (b) Zn from Matapeake soil predicted using the

kinetics model with (dashed lines) and without (solid lines)

re-adsorption at pH = 5.5 and flow rate = 1 mL/min 153 Figure 4.18 Continued .154 Figure 4.19 Release of (a) Zn and (b) Cu from Matapeake soil predicted using the

kinetics model at pH = 5.5 in the batch system (flow rate = 0) .156 Figure 4.19 Continued .157 Figure 5.1 Effect of SOM content on the kinetics of Zn sorption and desorption

on different soils (pH = 6.0 and influent [Zn] = 1.78 - 1.90 mg/L) (a) [SOC] = 0.76% and 2.32%; (b) [SOC] = 3.43% - 7.15% SOC

concentrations are presented in the figure Solid lines are model

calculations 173 Figure 5.1 Continued .174 Figure 5.2 Linear regressions of the Zn sorption rate coefficients vs SOC at pH

5.5 Dashed lines are regressions with zero intercept values and solid lines are regressions with non-zero intercepts (a) The slow site: the linear regression with zero intercept (ka1= 5.46 × 10-3 × [SOC], R2 = 0.965) and without zero intercept (ka1= 5.57 × 10-3 × [SOC] - 5.57 × 10-6, R2 = 0.966), (b) the fast site: the linear regression with zero

intercept (ka2= 9.34 × 10-2 × [SOC], R2 = 0.713) and without zero

intercept (ka2= 6.99 × 10-2 × [SOC] + 1.19 × 10-3, R2 = 0.841) .175 Figure 5.2 Continued .176

Figure 5.3 Kinetics of Zn sorption and desorption on different soils at (a) pH =

6.0, (b) pH = 6.5, and (c) pH = 6.0 with double soil particle

concentration (Influent [Zn] = 1.70 – 1.90 mg/L) SOC concentrations are presented in the figure Dilution curve corresponds to the blank

experiment without soil particles Solid lines are model calculations .180

Figure 5.3 Continued .182

Figure 5.4 Kinetics of Zn sorption and desorption on the Matapeake soil at

different influent Zn concentrations ([Zn]) and solution pH (pH

values are presented in the figure) (a) [Zn] = 0.83 – 0.92 mg/L; (b)

[Zn] = 1.73 – 1.91 mg/L; (c) [Zn] = 3.62 mg/L Solid lines are model calculations 185 Figure 5.4 Continued .186 Figure 5.4 Continued .187

Figure 5.5 The comparison of organic carbon normalized partition coefficient

Kp,OC calculated by WHAM VI and the kinetics model (symbols

indicate values calculated using the kinetics model; dashed line is the WHAM VI prediction) .189 Figure 5.6 Kinetics of Zn sorption/desorption on two sites with the two-site

model simulation at pH 6.0 and [Zn] = 1.73 – 1.91 mg/L (a) [SOC] = 0.76%; (b) [SOC] = 3.43% and (c) [SOC] = 7.15% 195

Figure 5.7 Kinetics of Zn sorption/desorption with the change of solution pH

(Sorption pH = 6.47 and desorption pH = 5.45; [Zn] = 1.80 mg/L) (a) Effluent Zn concentration vs time; (b) Particulate Zn concentration vs time 196 Figure 5.7 Continued .197

Figure 5.8 Distribution of Zn among soil components for the Boonton Bergen

County soil at different pH predicted using WHAM VI ([Zn] = 1.80 mg/L) 199

Figure 5.9 WHAM VI predicted Zn sorption behavior at different pH (a)

Equilibrium partition coefficients vs particulate Zn concentrations; (b) sorption isotherm ([SOC] = 2.32%) .201 Figure 5.9 Continued .202 Figure 5.10 The titration isotherm for Zn complexation by DOM at different pH

values ………203

Figure 6.1 Kinetics of Cu sorption and desorption on the Matapeake soils at

different pH (influent [Cu] = 1.4 - 1.7 mg/L) (a) pH = 5.5; (b) pH = 6.0 and (c) pH = 6.5 Lines are IE model fits 222 Figure 6.1 Continued .223 Figure 6.1 Continued .224 Figure 6.2 Kinetics of Cu sorption and desorption on the Matapeake soils at

different pH (influent [Cu] = 1.4 - 1.7 mg/L) (a) pH = 5.5; (b) pH = 6.0 and (c) pH = 6.5 Lines are kinetics model fittings .226 Figure 6.2 Continued .227 Figure 6.2 Continued .228

Figure 6.3 Kinetics of Cu sorption and desorption on the Matapeake soils at pH

5.5 with influent [Cu] = 0.85 mg/L Line is the kinetics model fitting .229

Figure 6.4 Kinetics of Cu sorption and desorption on two different soils at pH 5.5

with influent [Cu] = 1.6 mg/L: (a) [SOC] = 3.43%; (b) [SOC] =

7.15% Lines are kinetics model fittings .230 Figure 6.4 Continued .231 Figure 6.5 Kinetics of Cu sorption and desorption on the Matapeake soils at pH

5.5 (influent [Cu] = 1.7 mg/L): (a) Effluent Cu concentration vs time; (b) WHAM VI predicted partition coefficient vs time 236 Figure 6.5 Continued .237 Figure 6.6 Kinetics of Cu sorption and desorption on the Matapeake soils at pH

6.0 (influent [Cu] = 1.4 mg/L): (a) Effluent Cu concentration vs time; (b) WHAM VI predicted partition coefficient vs time 238 Figure 6.6 Continued .239 Figure 6.7 Kinetics of Cu sorption and desorption on the Matapeake soils at pH

6.5 (Influent [Cu] = 1.5 mg/L): (a) Effluent Cu concentration vs time; (b) WHAM VI predicted partition coefficient vs time 240 Figure 6.7 Continued .241

Figure 6.8 Kinetics of Cu sorption and desorption on the Matapeake soils at pH

5.5 with influent [Cu] = 0.85 mg/L: (a) Effluent Cu concentration vs time; (b) WHAM VI predicted partition coefficient vs time .242 Figure 6.8 Continued .243

Figure 6.9 Kinetics of Cu sorption and desorption on the Boonton Bergen soil at

pH 5.5 (influent [Cu] = 1.6 mg/L; [SOC] = 3.43%): (a) Effluent Cu

concentration vs time; (b) WHAM VI predicted partition coefficient

vs time .244 Figure 6.9 Continued .245

Figure 6.10 Kinetics of Cu sorption and desorption on the Boonton Union soil at

pH 5.5 (influent [Cu] = 1.7 mg/L; [SOC] = 7.15%): (a) Effluent Cu

concentration vs time; (b) WHAM VI predicted partition coefficient

vs time .246 Figure 6.10 Continued .247 Figure 6.11 The active organic matter vs SOC concentration: (a) percentage plot;

(b) mass concentration plot .249 Figure 6.11 Continued .250 Figure 6.12 The Cu partition coefficient vs Cu loading in soils (a) pH 5.5; (b) pH

6.0 and (c) pH 6.5 253 Figure 6.12 Continued .254 Figure 6.12 Continued .255

Figure 6.13 The Cu partition coefficient vs Cu loading at different SOC

concentrations (pH = 5.5 and 46% SOC is active) The SOC

concentrations are presented in the legend 258

Figure 6.14 Kinetics of Cu sorption/desorption on the Matapeake soil at pH 5.5

Lines are model calculations at different Ca concentrations (Model parameters were obtained by fitting the experimental data at Ca

concentration of 3 mM Solid line is the model fitting and dash lines are the model predictions of different Ca concentrations) .260

Figure 6.15 Kinetics of trace metals sorption and desorption on the Matapeake soil

(pH = 6.0; Influent metal concentration (µM): [Ni]=26.4, [Zn]=27.5, [Cu]=22.7, [Cd]=23.7) .263

ABSTRACT

Understanding the kinetics of trace metals sorption and desorption on soils

is important for better prediction of metal behavior in the environment In this

dissertation, the effect of solution chemistry and soil composition on trace metal sorption and desorption kinetics was investigated Based on the experimental results, predictive kinetics models were formulated and successfully used to describe the kinetics of trace metals sorption and desorption on a variety of soils

The kinetics experiments were conducted with a stirred-flow method A number of soil samples, which cover a wide range of soil properties from USA and European countries, were selected to test the effect of soil composition The effects of solution chemistry, pH, metal loadings, and dissolved organic matter (DOM)

concentration, were also systematically examined Most of kinetics experiments were run with two trace metals, copper and zinc, which demonstrated very different kinetics behaviors The solution pH has significant effect on the Cu and Zn sorption and desorption kinetics which can be accounted by the proton competition DOM greatly enhanced Cu release but had little effect on Zn, attributed to formation of strong Cu-DOM complex decreasing the Cu ion re-adsorption on soils Among all soil

components, soil organic matter (SOM) is the dominant phase controlling Cu and Zn sorption and desorption kinetics The effect of residence time was also tested

The kinetics models were formulated and successfully used to describe the

Cu and Zn sorption and desorption kinetics under different solution chemistry and soil compositions The models are based on the mass balance of the flow system and incorporate the chemical reactions between metals and soils For Zn, a two-site

kinetics model, including one fast and one slow site, was necessary The soil organic carbon (SOC) is the major model parameter accounting for the effect of soil properties

on Zn sorption and desorption kinetics The organic carbon normalized sorption rate coefficients can be applied to different soils based on their SOC concentrations One set of rate coefficients can be applied to different influent Zn concentrations, which corresponds to the linear portion of the Zn sorption isotherm The sorption rate

coefficients were dependent on solution pH which was accounted for by Zn and proton competition The same model parameters can be applied to different flow rates Cu showed different kinetics behavior and the one-site kinetics model was sufficient DOM effect can be accounted for by DOM complexation of copper ions, which

decreased their sorption Besides the kinetics model assuming the linear binding, the nonlinearity of Cu binding to soils was also considered by incorporating the

equilibrium models into the kinetics model, such as Windermere Humic Aqueous Model (WHAM) VI The equilibrium model was used to predict the partition

coefficient at every Cu loading The WHAM VI based kinetics model, which

considered the nonlinear metal binding and heterogeneity of soils, successfully

described Cu sorption and desorption kinetics for a variety of experimental conditions

sorption and desorption kinetics under different solution chemistry and soil compositions.

Chapter 1

INTRODUCTION

1.1 Background and Significance of the Study

Metals in the environment occur in several forms: dissolved in surface and ground waters, incorporated in or sorbed on the surface of minerals in rock, sand, and soils or as well as bound to soil organic matter Cadmium (Cd), copper (Cu), nickel (Ni) and zinc (Zn) are trace metals widely used in industry, transportation, agriculture and are being released into the environment Cadmium, Cu, Ni, and Zn enter surface waters both in particulate and in dissolved forms from natural and anthropogenic sources Natural processes contributing trace metals into soils, water and air include soils mineral weathering, water and wind erosion of rocks and soils, volcanic activity, and biological transfer of elements The major anthropogenic sources of trace metals are mining and smelting, agricultural application of fertilizers, pesticides, and

ameliorants containing trace metals, transportation, and municipal and industrial wastes

Contamination of the environment with trace metals is a serious problem

in many regions of the USA and all over the world Soils can sorb trace metals and

through leaching and in the particulate form through erosion (Bergkvist, 1986; McColl

et al., 1986; Rasmussen, 1986; Bergkvist, 1987; Bergkvist et al., 1989) Much of the

trace metals entering the surface waters are usually released from contaminated solid particles The accumulation of the metals in the waters is highly dependent on the weathering of these solid particles The trace metals leaching out of the soil particles into soil solutions (pore water) can become available to plants and animals in soils

Both sorption and desorption processes are important for controlling the metal behavior in the soil and solution systems The sorption process directly affects the metal distribution among different soils components and thus the future desorption from soils The equilibrium of trace metals partition between soils and solutions has been studied extensively and several equilibrium models have been developed to predict the partitioning of trace metals between solid and solution phases However, the sorption and desorption of trace metals on the soil particles appear to be a slow process and the equilibrium between solid and solution may not be attained in soils (Sparks, 1989 and 2001) There have been extensive equilibrium studies of metal sorption on soils for many years (Jenne, 1968; Shuman, 1975; McBride and Blasiak,

1979; Elrashidi and O’Connor, 1982; Zachara et al., 1992; Weng et al., 2001; Tipping

et al., 2003), but much fewer studies have focused on the kinetics of trace metal

sorption/desorption on soils (Yin et al., 1997; Strawn and Sparks., 2000;

Sukreeyapongse et al., 2002; Zhang et al., 2004)

Knowledge of the kinetics reactions for metals between soils and solutions can be important for predicting metal behavior since an equilibrium assumption may

not be appropriate The kinetic behavior of metals in the field may be affected by different processes including metal reactions between soils and solutions, the mobility

of solutions, solution diffusion, or uptake by organisms The metal reactions between

soils and solutions do not achieve equilibrium instantaneously (Yin et al., 1997; Strawn and Sparks, 2000; Zhang et al., 2004) In laboratory column studies, leaching

of trace metals from soils is kinetically controlled rather than by instantaneous

equilibrium (Kedziorek et al., 1998) However, the importance of this kinetic reaction

to control the metal behavior may differ for different chemistry and hydrological

conditions (Zhang et al., 1998; Ernstberger et al., 2005; Voegelin et al., 2001) Thus a

quantitative understanding of the rates of metal sorption and desorption on different soils at varying solution chemistry would be quite useful for developing models to accurately predict the fate and transport of trace metals in the environment for the levels of contamination available in the soils

One challenge for the kinetics modeling in the soil system is the

heterogeneity of soils, and little progress has been made in previous studies In soils trace metals are bound to different components including organic matter, clay

minerals, and metal oxides-hydroxides to different extents So, sorption and

desorption of the metals may include various chemical reactions proceeding at

different time scales with different mechanisms Taking into account all the main processes controlling the metals sorption and desorption on soils is most important to predict the impact of soil compositions on the metal contamination of surface waters

2000), metals reactions with soils may depend on metals content, SOM, iron and manganese oxides-hydroxides and clay fraction It is essential to assess the role of different sorbents in soil components to control the reactions of metals with soils at different conditions

In natural conditions, the solution compositions can be very different Solution pH, DOM concentration and other cations (e.g Ca2+, Al3+) may be the

principal solution properties affecting the metals reactions How these solution

parameters affect metal partition equilibrium has been extensively studied, and some speciation models have been developed to calculate the metal speciation at different conditions, such as MINEQL+ (Schecher and McAvoy, 1998) and WHAM (Tipping, 1994) However the understanding of the kinetics effect of solution chemistry is still insufficient A systematical study on how these factors affect the kinetics of metal sorption and desorption on soils is necessary in order to develop predictive models that can be applied to different solution chemistry conditions

Another important concept, which has been overlooked by most of the previous studies, is the speciation of metals in the reaction systems Even in the solutions, the metal speciation can be different at different pH, DOM concentrations The reactivity of different species may differ greatly Thus the total dissolved metals

may not be a good indicator for the reactive metals in the solutions (Allen et al.,

1980) For example, copper availability to plants and animals is found to be related not to total copper concentration in solutions but to the copper ion concentration (Di

Toro et al., 2001; McBride, 2001) In the soils, due to their heterogeneity, the metal

may be bound by a variety of sorbents and form different metal species Recently spectroscopic techniques have been used to identify the speciation of metals in soils, which demonstrated that the speciation of metals can vary considerably among soils

(Manceau et al., 1996; Roberts et al., 2002) Thus, in the kinetics modeling, the metal

speciation should be carefully taken account of since different metal species may have very different kinetic behaviors

Some kinetics models that have been used to describe metal sorption and desorption kinetics in soil systems (Sparks, 1989 and 1999) are reviewed in Chapter 2

in detail Briefly, the rates of the reactions on soil constituents have been described using various models, such as ordered models (zero-, first-, and second-order kinetics equations), parabolic diffusion, two-constant rate, Elovich, and differential rate

equations However, the rate constants obtained in these models were not constant but changed with experimental conditions Most of previous modeling was mainly based

on individual curve fitting and then correlation analysis The mechanisms controlling metal sorption and desorption in these models was not explicit and their usefulness is limited A new kinetics model, which specifically incorporates the reaction chemistry,

is needed in order to better predict the metal behavior in the environment

Furthermore, it has been recognized that the reactivity of different trace metals was different in the environment The cation properties, such as electron configuration, water exchange rate, and the first hydrolysis constant, etc., vary among metals, which may be the reasons accounting for their different behaviors For

Zn binding is very linear Cu can form strong complexes with DOM but Zn does not The pH can also affect Cu and Zn binding to humic substances differently All these factors make kinetics modeling complicated A robust kinetics model should be able

to predict different metal behaviors according to different chemical reactions

Overall, little progress has been made on the kinetics modeling of trace metals in soil systems compared the extensively used equilibrium models The

predictive kinetics models, which can be used at different solution chemistry and soil compositions, are highly desired This will improve the existing approaches based on equilibrium assumptions

1.2 Objective

The objective of this study is to develop kinetics models to predict trace metals (mainly Cu and Zn) sorption and desorption on soil particles The effects of soil properties and solution chemistry on kinetics of trace metals sorption and

desorption were studied Key soil and solution properties were selected The models are based on the minimal number of input parameters that provide reasonable accuracy

of model predictions The developed models will form an important component part

of the “Unit World” model of metal behavior in the environment being designed in the Center for the Study of Metals in the Environment, University of Delaware The results are expected to be critical for the development of soil screening procedures for ecological risk assessment

To achieve the proposed objective, this research is carried out with

following three steps The first step is the study of kinetics of Cu and Zn release from two soils In this part, key solution parameters are systematically evaluated and selected A two-site kinetics model is formulated and used to model Cu and Zn release kinetics This model is based on the linear binding isotherm of Cu and Zn, which is able to account for the pH, DOM and flow rate effect in the solutions The second step is focused on improvement and validation of the Zn two-site kinetics model The kinetics model is applied to both sorption and desorption kinetics

processes, different soil compositions and different solution chemistry (varying Zn loading and pH) This model is still based on the linear binding isotherm but the sorption rate coefficients are normalized based on the SOC concentration This model can be applied to different soils based on their SOC concentration The final step is the improvement of the Cu kinetics model The nonlinearity of Cu binding on soils is carefully considered and WHAM (Windermere Humic Aqueous Model) VI (Tipping

et al., 2003) is built into the kinetics model to predict the Cu nonlinear binding to

soils The WHAM based kinetics model is a more general model accounting for the heterogeneity of soils and nonlinear binding of metals In addition, the kinetic

behavior of some other trace metals, Cd and Ni are also compared with Cu and Zn

Following this introduction, Chapter 2 provides a literature review of the reaction chemistry of trace metals with soils, and equilibrium and kinetics modeling Chapter 3 discusses the kinetic modeling methods, including most of model equations

release from two soils, in which the effect of solution pH, DOM and flow rates are tested Chapter 5 develops a more general kinetics model for Zn sorption and desorption on different soils with the organic carbon normalization of sorption rate coefficients Chapter 6 focuses on incorporating the nonlinear Cu binding behavior and heterogeneity of SOM into the kinetics model for Cu sorption and desorption on different soils Chapter 7 summarizes the major findings of this study and proposes the recommendations for the future efforts

References

Allen, H E., Hall, R H and Brisbin, T D (1980) Metal speciation: effects on aquatic

toxicity Environ Sci Technol 14, 441-443

Bergkvist, B (1986) Leaching of metals from a spruce forest soil as influenced by

experimental acidification Water, Air Soil Pollut 31, 901-916

Bergkvist, B (1987) Soil solution chemistry and metal budgets of spruce forest

ecosystems in S Sweden Water, Air Soil Pollut 33, 131-154

Bergkvist, B., Folkeson, L and Berggren, D (1989) Fluxes of Cu, Zn, Pb, Cd, Cr and

Ni in temperate forest ecosystems A literature review Water, Air Soil Pollut

47, 217-286

Di Toro, D M., Allen, H E., Bergman, H L, Meyer, J S., Paquin, P R and Santore,

R.C (2001) Biotic Ligand Model of the Acute Toxicity of Metals 1

Technical Basis Environ Toxicol Chem 20, 2383-2396

Elrashidi, M A and O’Connor, G A (1982) Influence of solution composition on

sorption of zinc Soil Sci Soc Am J 46, 1153-1158

Ernstberger, H., Zhang, H., Tye, A., Young, S and Davison, W (2005) Desorption

kinetics of Cd, Zn, and Ni measured in soils by DGT Environ Sci Technol

39, 1591-1597

Jenne, E A (1968) Controls on Mn, Fe, Co, Ni, Cu and Zn concentrations in soils

and waters: The significant role of hydrous Mn and Fe oxides Adv Chem

Ser 73, 337-387

Kabata-Pendias, A (2001) Trace elements in soils and plants CRC Press, Boca

Raton, Florida

Kedziorek, M A M., Dupuy, A., Bourg, A C M and Compere, F (1998) Leaching

of Cd and Pb from a polluted soil during the percolation of EDTA: laboratory column experiments modeled with a non-equilibrium solubilization step

Environ Sci Technol 32, 1609-1614

Manceau, A., Boisset, M C., Sarret, G., Hazemann, J L., Mench, M., Cambier, P and

Prost, R (1996) Direct determination of lead speciation in contaminated soils

by EXAFS spectroscopy Environ Sci Technol 30, 1540-1552

McBride, M B and Blasiak, J J (1979) Zinc and copper solubility as a function of

pH in an acid soil Soil Sci Soc Am J 43, 866-870

McBride, M.B (2001) Cupric ion activity in peat soil as a toxicity indicator for

maize J Environ Qual 30, 78-84

McColl, J G and Pohlman, A A (1986) Soluble organic acids and their chelating

influence on Al and other metal dissolution from forest soils Water, Air and

Soil Pollut 31, 917-927

Rasmussen, L (1986) Potential leaching of elements in three Danish spruce forest

soils Water, Air and Soil Pollut 31, 377-383

Roberts, D R., Scheinost, A C and Sparks, D L (2002) Zinc speciation in a

smelter-contaminated soil profile using bulk and microspectroscopic

techniques Environ Sci Technol 36, 1742-1750

Sauvé, S., Hendershot, W and Allen, H E (2000) Solid-solution partitioning of

metals in contaminated soils: Dependence on pH and total metal burden

Environ Sci Technol 34, 1125-1131

Shuman, L M (1975) The effect of soil properties on Zn adsorption by soils Soil

Sci Soc Am J 39, 455-458

Sparks, D L (1989) Kinetics of soil chemical processes Academic Press, San

Diego, CA

Sparks, D L (1999) Soil physical chemistry (second edition) CRC Press, Boca

Raton, FL

Strawn, D G and Sparks, D L (2000) Effect of soil organic matter on the kinetics

and mechanisms of Pb (II) sorption and desorption in soil Soil Sci Soc Am J

64, 144-156

Sukreeyapongse, O., Holm, P E., Strobel, B W., Panichsakpatana, S., Magid, J and

Hansen, H C B (2002) pH-dependent release of cadmium, copper and lead

from natural and sludge-amended soils J Environ Qual 31, 1901-1909

Tipping, E (2002) Cation binding by humic substances Cambridge University Press,

Cambridge

Tipping, E., Rieuwerts, J., Pan, G., Ashmore, M R., Lofts, S., Hill, M T R., Farago,

M E and Thornton, I (2003) The solid-solution partitioning of heavy metals

(Cu, Zn, Cd, Pb) in upland soils of England and Wales Environ Pollut., 125,

213-225

Voegelin, A., Vulava, V M and Kretzschmar, B (2001) Reaction-based model

describing competitive sorption and transport of Cd, Zn, and Ni in an acidic

soil Environ Sci Technol 35, 1651-1657

Weng, L., Temminghoff, E J M and van Riemsdijk, W H (2001) Contribution of

individual sorbents to the control of heavy metal activity in sand soil Environ

Sci Technol 35, 4436-4443

Yin, Y., Allen, H E., Huang, C P., Sparks, D L and Sanders, P F (1997) Kinetics

of mercury(II) adsorption and desorption on soil Environ Sci Technol 31,

496-503

Zachara, J M., Smith, S C., Resch, C T and Cowan, C E (1992) Cadmium sorption

to soil separates contaminating layer silicates and iron and alumimun-oxides

Soil Sci Soc Am J 56, 1074-1084

Zhang, H., Davison, W., Knight, B and McGrath, S (1998) In situ measurement of

solution concentrations and fluxes of trace metals in soils using DGT

Environ Sci Technol 32, 704-710

Zhang, H., Lombi, E., Smolders, E and McGrath, S (2004) Kinetics of Zn release in

soils and prediction of Zn concentration in plants using diffusive gradients in

thin films Environ Sci Technol 38, 3608-3613

Chapter 2 LITERATURE REVIEW

2.1 Trace Metal Content in Soils and Reactivity

2.1.1 Copper

Generally copper (Cu) content in soils ranges from 2 to 250 mg/kg

(Sparks, 1995) For uncontaminated soils Cu content in different soil types usually varies 1 to 50 mg/kg (Kabata-Pendias, 2001) The mean values vary from 13 to 24 mg/kg In contaminated soils at old mining areas, close to the smelters, the content of

Cu appears to be up to 300-800 and up to 2000-4000 mg/kg in urban orchards,

sludged, irrigated, or fertilized farmland (Kabata-Pendias, 2001)

Cu usually occurs in soils adsorbed by organic matter, Fe and Mn oxides (hematite, goethite, birnessite), amorphous Fe and Al hydroxides and clay minerals Occlusion, coprecipitation, and substitution are involved in nonspecific adsorption of

Cu Among all soil components, the ability of soil organic compounds to bind Cu is

well recognized Stevenson et al (1981) stated that maximum amount of Cu2+ that can be bound to humic and fulvic acids is close to the content of acidic functional groups In general, this corresponds to the sorption of 48 to 160 mg Cu per gram of

humic acid According to Bloom et al (1979), humic acids strongly immobilize Cu2+

ion in direct coordination with functional oxygen-atoms of organic substances

Schilling and Cooper (2004) found that carboxyl and hydroxyl functional groups were

most significant in Cu binding in the soil, with the aid of chemical modification and

13C CP-MAS NMR spectroscopy

Cu is abundant as free and complexed ions in soil solutions of all types of soils Concentration of Cu in soil solutions varies from 3 to 135 µg/L (0.047 to 2.1 µM), depending on soil types and the experimental techniques (Kabata-Pendias, 2001) Complexation of Cu2+ with soluble organic substances may govern the

bioavailability and the migration of this metal in soils and surface waters CuCO3 is reported to be the major inorganic soluble form of Cu in neutral and alkaline soil solutions (Sanders and Bloomfield, 1980)

Cu is one of the 3d transition metals and the Cu2+ ion has the [Ar]3d9electron configuration The first hydrolysis constant (pK1) of Cu2+ ion is 8.00 and the

water exchange ratio is 1×9 1/s (Tatara et al., 1997; Sekaly et al., 2003)

2.1.2 Zinc

Zinc (Zn) is a common element and occurs naturally in soils as an element

or as a mineral Zn content in uncontaminated surface soils in the US and other countries ranges from 17 to 125 mg/kg (Kabata-Pendias, 2001) In the highly

contaminated sites, Zn concentrations can be much higher, which are mainly present

as Zn minerals (Roberts et al., 2002) Zinc concentrations in soil solutions range from

The reactivity of Zn with different soil components is inconclusive Some researchers have reported that Zn binding by clay minerals and hydrous Fe and Al oxides, are likely to be the most important factors controlling Zn solubility, while

complexation with SOM and precipitation may be less important (Zyrin et al., 1976; Kabata-Pendias et al., 1995; Adb-Elfattah et al., 1981) Organic complexing and

precipitation of Zn as hydroxide, carbonate, and sulfide are reported to be of much less importance (Kabata-Pendias, 2001) However, recent researches have demonstrated

that SOM is the major sorbent controlling Zn reactions in soils (Weng et al., 2001; Tipping et al., 2003) It is likely that the importance of different soil components for

Zn binding may vary at different conditions

Zn2+ ion has the [Ar]3d10 electron configuration The pK1 of Zn2+ ion is 8.96 and the water exchange ratio is 7×107 1/s (Tatara et al., 1997; Sekaly et al.,

2003)

2.1.3 Other Metals

Cadmium (Cd) and nickel (Ni) are another two common trace metals which are environmentally relevant As one of the most eco-toxic metals in soils, Cd concentrations lie between 0.06 and 1.1 mg/kg (Kabata-Pendias, 2001) The sorption

of Cd by soil components has been extensively studied and leads to some

generalizations: Cd activity is strongly affected by pH in all soils; in acid soils, the organic matter and oxides may largely control Cd solubility, and in alkaline soils, precipitation of Cd compounds is likely to account for Cd equilibrium

Ni content in soils ranges from 0.2 to 450 mg/kg and in the US ranges from <5 to 150 mg/kg (Kabata-Pendias, 2001) The highest Ni contents are always in clay and loamy soils In soils, Ni mainly occurs in organically bound forms and soil

Ni carried in the oxides of Fe and Mn is also important and available for plants Ni concentrations in soil solutions vary from 3 to 25 µg/L (Kabata-Pendias, 2001)

Ni is the 3d transition metal and Ni2+ has the [Ar]3d8 electron

configuration The first hydrolysis constant of Ni2+ ion pK1 = 9.86 and the water exchange rate equals 3×104 1/s (Tatara et al., 1997; Sekaly et al., 2003) In the

periodic table Cd is just below the Zn and Cd2+ has the [Ar]4d10 electron

configuration The first hydrolysis constant of Cd2+ ion pK1 = 10.3 and water

exchange ratio equals 3×108 1/s (Tatara et al., 1997; Sekaly et al., 2003)

It is possible that the trace metal kinetics parameters can be related to the

basic cation characteristics Sekaly et al (2003) tried to correlate the rate constants of

trace metals (Co, Cu, Ni and Zn) dissociation from humic substances to the basic cation properties They found that the lability of the metal complexes follows the reverse order of the ligand field stabilization energy (LFSE), except for Cu The rate

of the water exchange falls into the similar order of LFSEs (weak field) Furthermore, the experimentally observed fastest dissociation rate coefficients of Co, Cu, Ni and Zn

complexes shows the similar trend, with the exception of Cu Likewise, Fasfous et al

(2004) also observed, except for Cu, the dissociation constants for metal complexes follow the similar trend predicted from the LFSEs (weak field) and the inert

complexes are those with large LFSEs Although more research is needed, a general model for different trace metals may be possible based on the cation properties

2.2 Soil Properties Affecting Trace Metals Sorption and Desorption

Soils consist of different components including SOM, metal (hydro)oxides and clay minerals which are responsible for metal binding The extent of metal

binding by these components is important for understanding the kinetics of metal reactions in soil systems Different components in soils may contribute to metal sorption and desorption to different extent It is thus necessary to assess the relative importance of soil sorbents for the binding of trace metals in soils

2.2.1 Metal Concentration in Soils

The metal concentration in soils may directly affect both the amount and the rate of metals sorption and desorption The latter may be expected since the rate of any chemical reaction increases with the increase in the concentrations of reactants Moreover, the effect of metal concentration on the sorption and desorption rate may be more complex since metals may be bound to different soil compounds with different reaction rates Even for the same soil component, e.g the SOM, the reaction rates may be different at different metal concentrations since SOM is reported as the

heterogeneous mixture with a series of binding sites Thus more interesting is how the metal concentrations in the soils will affect the distribution of the metal in soil binding sites and thus the reaction (sorption/desorption) rates

Tipping (2002) demonstrated how Cu saturated the different binding sites

of humic substance with the increase of Cu loading using the Windermere humic aqueous model (WHAM) At low loading, the Cu preferentially bound to the strong tridentate sites, which have the highest affinity for Cu With the increase of Cu

loading, more binding sites were filled and Cu was distributed to the weaker binding sites, the more abundant bidentate sites Since the Cu distribution is different at different Cu loadings, it is expectable that at different Cu loading the kinetics of Cu reactions may vary since the Cu was bound to different functional groups

Furthermore, at different metal loadings, metal distribution among organic and

mineral components in soils may be different

It has also been recognized that, even with same metal concentrations, the metals in aged soils behavior differently from the metals in freshly reacted soils

Zhang et al (2004) showed that the Zn release from fresh spiked soils was much faster

than that from field aged soils So the absolute soil metal concentration alone may be not a good index since reaction time is an important factor to control the metal

reactivity in soils

The metal concentration in soils also affected the binding behaviors of metals with soils A linear sorption isotherm has been often observed at low metal concentrations for some metals In the linear isotherm region, the partition

coefficients are the constant at constant pH, irrespective the metal loading This will simplify the kinetics model since the effect of metal loading can be neglected For

dependency of partition coefficients on metal loadings in soils In this situation, more complex models are necessary to explain the kinetic behaviors at different metal

concentrations

2.2.2 Soil Organic Matter

SOM consists of non-humic substances such as amino acids,

carbohydrates, and lipids, and humic substance, a series of high-molecular-weight, brown to black substances formed by secondary synthesis reactions (Stevenson, 1994) SOM is able to bind trace metals and thus affects metal sorption and desorption on soils SOM content in most mineral soils is only 0.5 to 5%, but up to 80% of the soil cation exchange capacity (CEC), especially of sandy soils, is due to SOM (Sparks, 1995)

Humic substances contribute the most to the cation-binding properties of SOM Complexation of metals with SOM is attributed mainly to carboxylic and

phenolic groups available in the molecules of SOM The ability of these groups to

bind metals decreases in the following sequences (Kinniburgh et al., 1996):

Carboxylic groups: H+>>Pb2+>Cu2+>Cd2+>Ca2+

Phenolic groups: H+>> Cu2+>Pb2+> >Cd2+>>Ca2+

The stability of the humic complexes of various metals was also reported

to follow the Irving-Williams order: Mg<Ca<Cd≈Mn<Co<Zn≈Ni<Cu<Hg (Mantoura

et al., 1978) Thus, the binding ability of H+ with SOM is much stronger than that of metal ions That is why the pH greatly influences metal binding by SOM The

average pKa for the carboxylic group and phenolic group is 4.5 and 10 respectively (Martell and Smith, 1977) Thus at low pH, the most dissociated groups are

carboxylic groups and at high pH phenolic group shows more importance Different trace metals may bind to these sites differently Lu and Allen (2002) reported that Cu complexation with humic substance is principally through the replacement of protons

by Cu ions at the phenolic binding sites in their experimental conditions Moreover, the most part of Ca and Mg is bound by carboxylic groups especially at high

concentrations For Zn, the acidic carboxylic group was responsible for the majority

of Zn binding (Cheng, 2003) The different binding properties of trace metals with humic substance indicated the kinetics of trace metal reactions with humic substance may be very different

Several researchers evaluated the importance of SOM on different trace

metals’ reactions between solid and solution phases Yin et al (1997) observed that

Hg(II) adsorption and desorption rate coefficients were inversely correlated with SOM content For Cu, some researchers found high dependence of partition of Cu on SOM

content in soils (Impellitteri, 2000), but some did not find any dependence (Janssen et

al., 1997; Sauvé et al., 1997) by the statistical analysis More recent findings

supported the importance of SOM controlling Cu binding in soils both by application

of spectroscopic techniques (Flogeac et al., 2004) and equilibrium models (Weng et

al., 2001) With the combination of electron paramagnetic resonance (EPR), extended

X-ray absorption fine structure (EXAFS), and X-ray absorption near-edge structure

(XANES) spectroscopies, Cu was found to bind to organic matter coated onto the

mineral fraction of soil particles (Flogeac et al., 2004)

The heterogeneity of SOM has been well recognized The spectroscopic study on humic substance has also demonstrated the molecular structure of humic substances is very complicated and a series of functional groups are able to complex trace metals These studies have been reviewed by Stevenson (1994) The modeling

of metal reactions with SOM is thus a big challenge Recent modeling efforts have demonstrated that the heterogeneity of SOM can be simplified and the SOM of

different type soils behave similarly (Weng et al., 2001; Gustafsson et al., 2003; Tipping et al., 2003)

2.2.3 Clay Minerals and Metal Oxides

The clay minerals may significantly affect retention and release of trace metals in soils since (a) they contain minerals and compounds adsorbed on their surface (organic matter, iron and manganese hydroxides) able to bind trace metals and (b) they have high specific surface area and consequently large number of available binding sites Oxides-hydroxides of Fe, Mn and Al are present in most soils revealing high specific surface and reactivity with the respect to trace metals Surface

complexation mechanism is always assumed for binding the trace metals on oxides surfaces

Clay fraction or amorphous Fe and Al contents and CEC have been

assumed to be the main properties influencing Zn desorption rate (Dang et al., 1994;