United States
Environmental Protection
Agency
Otfice of Solid Waste -
and Emergency Response
Washington, DC 20460
EPA/530/SW-87/006-F
April 1992
Technical Resource
Document

Batch-Type Procedures
For Estimating Soil
Adsorption of Chemicals

-------

-------
                                                  EPA/530-SW-87-006-F
         Technical Resource Document

BATCH-TYPE PROCEDURES FOR ESTIMATING

      SOIL ADSORPTION OF CHEMICALS
                          by

      W.R. Roy, I.G. Krapac, S.F.J. Chou, and R.A. Griffin

                Illinois State Geological Survey
                  Champaign, Illinois 61820
            Cooperative Agreement No. CR810245
                     Project Officer
                      M.H. Roulier

               Waste Minimization, Destruction
               and Disposal Research Division
            Risk Reduction Engineering Laboratory
                   Cincinnati, Ohio 45268
         Office of Solid Waste and Emergency Response
             U.S. Environmental Protection Agency
                  Washington, D.C. 20460
             Risk Reduction Engineering Laboratory
             Office of Research and Development
             U.S. Environmental Protection Agency
                   Cincinnati, Ohio 45268
                          1991                jCQi
                                             f£& Printed on Recycled Paper

-------
                                        DISCLAIMER

The studies reported in this document were funded in part by the United States Environmental Protection
Agency under Cooperative Agreement CR810245 with the Illinois State Geological Survey, Champaign, Il-
linois. This report has been subject to the Agency's peer and administrative review, and it has been ap-
proved for publication as an EPA document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.

-------
                                         PREFACE

Subtitle C of the Resource Conservation and Recovery Act (RCRA) requires the U.S. Environmental Pro-
tection Agency (EPA) to establish a federal hazardous waste management program. This program must
ensure that hazardous wastes are handled safely from generation until final disposition. EPA issued a se-
ries of hazardous waste regulations under Subtitle C of RCRA that are published in 40  Code of Federal
Regulations (CFR) 260 through 265 and 122 through 124.

Parts 264 and 265 of 40 CFR contain standards applicable to owners and operators of all facilities that
treat, store, or dispose of  hazardous wastes. Wastes are identified or listed as hazardous under 40 CFR
Part 261. The Part 264 standards are implemented through permits issued by authorized states or the
EPA in accordance to 40  CFR Part 122 and Part 124 regulations. Land treatment, storage, and disposal
(LTSD) regulations in 40 CFR Part 264 issued on July 26,1982, establish performance standards for haz-
ardous waste landfills, surface impoundments, land treatment units, and waste piles.

The Environmental Protection Agency is developing three types of documents for preparers and review-
ers of permit applications for hazardous waste LTSD facilities. These types include RCRA Technical Guid-
ance Documents, Permit  Guidance Manuals, and Technical Resource Documents (TRDs). The RCRA
Technical Guidance  Documents present design and operating specifications or design evaluation tech-
niques that generally comply with or demonstrate compliance with the Design and Operating Require-
ments and the Closure and Post-Closure Requirements of Part  264.

The Permit Guidance Manuals are being developed to describe the permit application information the
Agency seeks and to provide guidance to applicants and permit writers in addressing information require-
ments. These manuals will include a discussion of each step in  the permitting process, and a description
of each set of specifications that must be considered for inclusion in the permit.

This document is a Technical Resource Document. It was prepared for the Risk Reduction Engineering
Laboratory, formerly the Hazardous Waste Engineering Research Laboratory of the Office of Research
and Development, at the  request of and in cooperation with the Office of Solid Waste and Emergency Re-
sponse. This TRD was first issued as a draft for public comment under the title Batch-Type Adsorption
Procedures for Estimating Soil Attenuation of Chemicals (EPA/530-SW-87-006) in June 1987. The draft
TRD was also made available through the National Technical Information Service (Order No. PB87-
146155). All comments received on the draft TRD have been carefully considered and, if appropriate,
changes were made in this final document to address the public's concerns. With issuance of this docu-
ment, all previous drafts of the TRD are obsolete and should be discarded.

-------
                                          FOREWORD

Today's rapidly developing and changing technologies and industrial products and practices frequently
carry with them the increased generation of materials that, if improperly dealt with, can threaten both pub-
lic health and the environment. The U.S. Environmental Protection Agency is charged by Congress with
protecting the Nation's land, air, and water resources. Under a mandate of national environmental laws,
the agency strives to formulate and implement actions leading to a compatible balance between human
activities and the ability of natural systems to support and nurture life. These laws direct the EPA to per-
form research to define our environmental problems, measure the impacts, and search for solutions.

The Risk Reduction Engineering Laboratory is responsible for planning, implementing, and managing
research, development, and demonstration programs to provide an authoritative, defensible engineering
basis in support of the policies, programs, and regulations of the EPA with respect to drinking water,
wastewater, pesticides, toxic substances, solid and hazardous wastes, and Superfund-related activities.
This publication is one of the products of that research and provides a vital communication link between
the researchers and the user community.

The Office of Solid Waste is responsible for issuing regulations and guidelines on the proper treatment,
storage, and disposal of hazardous wastes to protect human health and the environment from the poten-
tial harm associated with improper management of these wastes. These regulations are supplemented by
guidance manuals, technical guidelines, and technical resource documents, made available to assist the
regulated community and facility designers in understanding the scope of the regulatory  program. Publica-
tions like this one provide facility designers with state-of-the-art information on design and performance
evaluation techniques.

This Technical Resource Document (TRD) describes a number of laboratory batch procedures for assess-
ing the capacity of soils and soil components of liners for waste management facilities to adsorb chemical
constituents from solution. Procedures for both organic and inorganic constituents are described, and
their scientific basis and rationale are documented. Examples  are included to demonstrate the application
of the procedures and the use of the data in designing soil liners for pollutant retention.
                                                 E. Timothy Oppelt, Director
                                                 Risk Reduction Engineering Laboratory
                                               IV

-------
                                          ABSTRACT

This Technical Resource Document (TRD) contains laboratory procedures and guidelines for conducting
experiments using batch-equilibrium techniques to study soil adsorption of chemicals dissolved in solution
(solutes). The procedures were designed for routine use and can be used to generate data for construct-
ing equilibrium adsorption isotherms or curves. Procedures are given for inorganic and organic solutes
and volatile organic solutes.

The scientific basis for each procedural step is discussed in detail. Procedures were based on the scien-
tific literature and were developed and tested by the authors and cooperating laboratories. Examples are
given that show how to apply major procedural steps and concepts. Several types of soil materials and
solutes are featured, as well as the application of batch-adsorption data in calculations of solute move-
ment through compacted landfill liners, which is needed for estimating the thickness of liner required for
pollutant retention.

This TRD was submitted in May 1989 by the Illinois State Geological Survey in fulfillment of Cooperative
Agreement CR810245 with the U.S. Environmental Protection Agency. The TRD has been revised to ad-
dress issues that were raised during the public comment period on the draft TRD entitled Batch-Type Ad-
sorption Procedures for Estimating Soil Attenuation of Chemicals (EPA/530-SW-87-006); the revised TRD
also includes technical information that became available after the draft TRD was completed in May 1987.

-------

-------
                                          CONTENTS

Foreword
Preface
Abstract
Acknowledgments
Introduction
Chapter
 1   Adsorption Forces and Mechanisms
 2   Effects of Adsorbent Preparation
 3   Effects of Temperature
 4   Stability of Nonionic Organic Solutes in Solution
 5   Effects of Solution pH
 6   Effects of Ionic Strength
 7   Effects of Phase Separation
 8   Effects of Method of Mixing
 9   Selection of a Soil:Solution Ratio for Ionic Solutes
10   Selection of a SoihSolution Ratio for Nonionic Solutes
11   Effects of the SoihSolution Ratio
12   Constant and Variable Soil:Solution Ratios
13   Determination of the Equilibration Time
14   Construction of Adsorption Isotherms
15   Selection of Adsorption Equations
16   Application of Batch-Adsorption Data
17   Laboratory Procedures for Generating Adsorption Data
References
Appendix
 A   Summary and Chemical Composition of the Adsorbent Soils
     and Clays  Used in This Study
 B   Composition of the Metallic Waste Extract Used
     in This Study and Associated Adsorption Isotherms
 1

 3
 5
11
15
17
21
25
27
31
35
39
49
53
59
65
67
75
87
95

99
                                                vii

-------
                                              FIGURES
  5
  6

  7
  8

  9
 10
 11

 12
 13

 14
 15
 16

 17

 18
 19

 20

 21

 22

 23
 24

 25

 26

 27

 28

 29

30
31
 Effect of air-drying on concentration of exchageable manganese
 Relationship between concentration of exchangeable potassium in Harpster clay loam
 and moisture content
 Effect of oven-drying at 105°C on concentration of water-soluble organic carbon in
 an Israeli calcareous clay loam
 Adsorption isotherm of acetophenone by fresh field-moist and air-dried samples of Crane Island
 alluvium
 Arsenate adsorption isotherms by Catlin soil at 15°, 25°, and 35°C, and at pH 6.6
 Zinc, copper, and cadmium adsorption from a Du Page County landfill leachate
 by kaolinite at 25°C at various pH levels
 Chromium (VI) adsorption by kaolinite at 25°C at various pH levels
 Langmuir-type maximum for adsorption of arsenic as AS(V) and As(lil) by amorphous
 iron hydroxide
 Effect of pH on adsorption of triazines by a Ca-montmorillonite sample
 Effect of pH on adsorption of different ionizabie organic solutes by an illite sample
 Adsorption of PCB aroclor 1242 by a synthetic goethite, a Cecil clay, and EPA-14 soil samples
 as a function of pH at 24°C
 Ratio of concentration to activity versus ionic strength for some common ions
 Effect of pore size and number of continuative filtrations of 100-mL aliquots of
 HCB-saturated water on concentration of HCB in filtrates
 Distribution of arsenic concentrations in solutions that were either centrifuged or filtered
 National Bureau of Standards rotary extractor
 Distribution of arsenic concentrations after 24 hours of contact with different soil
 materials as a function of soihsolution ratio
 Distribution of cadmium concentrations after 24 hours of contact with different soil
 materials as a function of soil:solution ratio
 Adsorption isotherm of o-xylene by Catlin soil at 23°C and at pH 6.1
 Adsorption isotherms of dichloroethane and tetrachloroethylene by Catlin soil at 23°C
 and at pH 6.1
 Relationship between the linear Freundlich constant (Kd) and soihsolution ration
 as a function of percent adsorption (lower range)
 Relationship between the linear Freundlich constant (Kd) and soihsolution ration
 as a function of percent adsorption (upper range)
 Effect of soil:solution ratio on cadmium adsorption by a Sangamon Paleosol sample at
 pH 6.1 and at 22°C
 Cadmium adsorption by a Sangamon Paleosol sample
 Distribution of pH values of arsenate solutions after 24 hours of contact with different
 soil materials as a function of soihsolution ratio
 Distribution of pH values of cadmium solutions after 24 hours of contact with different
 soil materials as a function of soihsolution ratio
 Distribution of pH values of solutions of zinc slurry extract after 24 hours of contact with different
 soil materials as a function of soihsolution ratio
 Distribution of the ionic strength of solution containing either arsenate or cadmium
 after 24 hours of contact as a function of soihsolution ration
 Freundlich constant (K) for two PCB isomers versus sediment concentration with
 and without prewashing to remove nonsettling particles
 Freundlich constant (K) for the adsorption of Aroclor 1242 by four different soils at 23°C
 as a function of soihsolution ratio
Aroclor 1242 adsorption isotherms by 5 soils at 23°C using various soihsolution ratios
Adsorption of dieldrin, tetrachloroethylene, and  1,2-dichloroethane by Catlin soils at
various soihsolution ratios
  5

  5

  6

  8
 11

 18
 19

 19
 20
 20

 20
 21

 25
 26
 28

 32

 33
 36

 36

 37

 38

 40
 41

 42

 43

 43

 44

 45

 45
46

47
                                                VIII

-------
                                     FIGURES (continued)

32  Adsorption of Aroclor 1242 by altered Vandalia till and unaltered Vandalia till at 23°C using
    various soihsolution ratios                                                                  48
33  Distribution of Freundlich constants (K) and exponents (1/n) associated with arsenic,
    cadmium, lead, and PCB (Aroclor 1242) adsorption isotherms                                   50
34  Cadmium adsorption isotherm with a Vandalia till sample with the amount
    adsorbed associated with each isotherm data point shown                                      51
35  Distribution of cadmium adsorption data for a Tifton sandy loam                                 47
36  Distribution of arsenate adsorption data at 23°C for different soil samples using different
    soihsolution ratios                                                                         52
37  Adsorption of cadmium by 5 soil materials at 22°C as a function of contact time                   54
38  Adsorption of arsenic by 11 soil materials as a function of contact time                           54
39  Equilibration times of Ba, Pb, and Zn from a laboratory extract of the Sandoval zinc slurry
    with the Sangamon Paleosol and the Cecil clay sample                                         56
40  Adsorption of o-xylene, dichloroethane, and tetrachloroethylene by Catlin soil as
    a function of contact time                                                                   56
41  Adsorption of arsenic by a kaolinite clay sample at 25°C, described by traditional linear
    Langmuir, double-reciprocal Langmuir, and Freundlich equation                                 66
42  Lead adsorption by Cecil clay loam at pH 45 and at 25°C, described by a linear Freundlich
    equation forced through the origin                                                           71
43  Predicted distance of lead migration in Cecil clay loam after 35 years, based on three
    approaches                                                                               73
44  Flow diagram of the procedures for generating batch-adsorption data                            76
B1  Barium adsorption isotherm at 21 °C with the Sangamon Paleosol from the metallic
    waste extract                                                                              "
B2  Lead adsorption isotherms  at 24°C of two soils using the metallic waste extract                   100
B3  Zinc adsorption isotherms at 24°C of two soils using the metallic waste extract                   100
                                                IX

-------
  5
  6
 10
 11

 12
 13
 14
 15
A1
A2
A3
A4
B1
                                        TABLES
 Effect of drying on exchangeable manganese in for Hawaiian soils
 pH of soil-water slurries (1:2 v/v) made with field-moist and oven-dried (110°C) samples
 Effect of sample pretreatment on the Freundlich partition coefficients (/O of acetophenone
 adsorption                                                      '
 Effect of temperature on Freundlich adsorption constants (Kf) for phenanthrene
 cc-naphthol
 First ASTM sensitivity analysis for cadmium and arsenic
 Cadmium adsorption data from second ASTM interlaboratory sensitivity analysis with
 an NBS rotary extractor as the mixing device
 Arsenic adsorption data from second ASTM interlaboratory sensitivity analysis with
 an NBS rotary extractor as the mixing device
 Soihsolution ratio determination for the Sangamon soil and Vandalia ablation till using cadmium
 as the adsorbate
 Soil: solution ratios for the Sangamon Paleosol and the Cecil clay loam sample using an extract
 of the Sandoval zinc slurry
 Equilibration times for adsorption of arsenate by soil materials
 Equilibration times for adsorption of Ba, Pb, and Zn from a Sandoval zinc slurry extract bv the
 Sangamon Paleosol and Cecil clay
 Equilibration times for adsorption of the PCB Aroclor 1242  by Catlin soil
 Data reduction for arsenic adsorption by a kaolinite clay sample
 Lead adsorption data for a Pb(NO3)2 salt and the Cecil clay
 Approaches for estimating minimum liner thicknesses on the basis of adsorption
 Summary of adsorbents
 Selected physicochemical characteristics of clays and soils used in developing this TRD
 Major element composition (in oxide form) of clay and soils used in developing this TRD
Trace element concentrations in clays and soils used in developing this TRD
Chemical constituent concentrations obtained by the ASTM-A water shake extraction
performed on the Sandoval zinc slurry
                                                                                                7
                                                                                                7
 12
 27

 29

 29

 31

 34
 55

 57
 57
 60
 70
 74
 95
 96
 96
 97

99

-------
                                    ACKNOWLEDGMENTS

The authors acknowledge the partial support of the U.S. Environmental Protection Agency, Cincinnati,
Ohio, and Dr. Michael H. Roulier, project officer of Cooperative Agreement CR810245. We thank Dr. Cal-
vin C. Ainsworth, formerly with the Illinois State Geological Survey (ISGS), for his efforts during the first
year of this project, Dr. Randall E. Hughes for the mineralogical characterizations, Terence Beissel and
Robert Arns for their technical support, and members of the Analytical Chemistry Section of the ISGS for
the adsorbent characterizations. Several laboratories contributed directly and indirectly to this TRD
through their participation in American Society for Testing and Materials (ASTM) D34.02 on Waste Dis-
posal round-robin testing of batch-adsorption procedures: Dr. Gregory Boardman (Virginia Polytechnic In-
stitute and State University), Dr. Chester Francis (Oak Ridge National Laboratory, Tennessee), Dr. Marc
Anderson (University of Wisconsin), Dr. William A. Sack (West Virginia University), and Otis E. Michels
(Daily and Associates Engineers, Peoria, Illinois). Dr. John J. Hassett of the University of Illinois is grate-
fully acknowledged for several informal discussions that helped to refine this document. The suggestions
made by Dr. Kenneth J. Williamson of Oregon State University, Dr. P.S.C. Rao of the University of Flor-
ida, and Dr. Thomas C. Voice of Michigan State University are also acknowledged and appreciated.
                                               XI

-------

-------
                                        INTRODUCTION

The capacity of geological materials to attenuate potential pollutants has been studied by many re-
searchers, especially during the past 30 years. A potential application of such research is the design and
evaluation of compacted soil or clay liners for attenuation of chemical constituents of leachates from
waste-management facilities, such as landfills and surface impoundments. The batch-adsorption or static-
equilibration technique has often been used in laboratory studies to assess the capacity of soils and soil
components to remove chemical constituents from solution. Batch procedures vary considerably from one
another in terms of experimental conditions and research objectives and sometimes yield different results
even when the same soils, solutes, and concentrations are studied.

The simplicity of the batch-adsorption technique accounts in part for its popularity. With this technique, an
aqueous solution containing solutes of known composition and concentrations is mixed with a given mass
of adsorbent for a given period of time. The solution is then  separated from the adsorbent and chemically
analyzed to determine changes in solute concentration. The amount of solute adsorbed by the adsorbent
is assumed to be the difference between the initial concentration (before contact with the adsorbent) and
the solute concentration after the mixing period. Although the approach is relatively simple, several experi-
mental parameters can affect the adsorption of a given constituent.

For inorganic solutes, these parameters include contact time, temperature, method of mixing, soil:solution
ratio, adsorbent moisture content, solution pH, hydrolysis, and the composition and concentration of other
dissolved constituents in the solution (White, 1966; Barrow  and Shaw,  1975,1979; Helyar et al., 1976;
Hope and Syers, 1976; Griffin and Au, 1977; Barrow, 1978; Roy et al., 1984,1985). For organic solutes,
similar parameters may also affect adsorption (Bailey and White, 1970; Graver and Hance, 1970; Dao
and Lavy, 1978; Koskinen and Cheng, 1983; Horzempa and DiToro, 1983). Dissolved organic carbon, ad-
sorbate volatility, photodegradation, biodegradation, and compound stability can also affect adsorption
data associated with organic solutes (Harris and Warren, 1964; Scott et al., 1981; Chou and Griffin, 1983).

In the batch-technique studies cited above, equilibration time, a basic experimental parameter, has varied
from 30  minutes to 2 weeks. Soil:solution ratios used in batch procedures have varied from very dilute sys-
tems (1:100,000)  to 1:1 pastes. These particular experimental conditions were probably appropriate for
the specific system under study and for the intended use of the data. However,  this diversity in experimen-
tal conditions can make comparisons of data between studies difficult. Furthermore, few well-docu-
mented, comprehensive sources are available on conducting batch-adsorption experiments. The only
standardized batch-adsorption procedures designed for routine use are the procedural guidelines outlined
by EPA (1982) and the standard methods developed by the American Society for Testing and Materials
 (ASTM) D-18, D-34, and E-47. Results from D-34.02 round-robin testing of batch-adsorption procedures
 indicated coefficients of variation greater than 140% during initial testing; these were reduced to less than
 10% by application of standard procedures and equipment between laboratories (Griffin et al., 1985).

 This Technical Resource Document (TRD) incorporates the experience gained during those inter- labora-
 tory testing programs and the interactions with the scientists and laboratories affiliated with ASTM. The
 ASTM procedure D4646 was reviewed and voted on by the committee members. Comments were re-
 ceived from 96 people active in research,  government, industry, and waste management. This TRD de-
 scribes  a number of batch-adsorption procedures for inorganic and organic solutes, documents their
 scientific basis, and recommends procedural steps. Examples are given that demonstrate how to apply
 each procedural step. Chapter 16 shows how adsorption data can  be used in designing or evaluating soil
 liners for pollutant retention. Chapter  17 contains the actual procedures; before attempting them, the
 reader should study the preceding chapters. Most of the procedural steps recommended here were tested
 in the authors' laboratory with a variety of soils, solutions containing several solutes, and aqueous ex-
 tracts of actual wastes. Characteristics of the soils, clays, and waste are described in the appendices.

 The collection of accurate and meaningful adsorption data is not a simple task. Even though the proce-
 dures described  here are fairly easy to use and precise, some "scatter" in the data will  inevitably occur,

-------
and the origins of the dev.at.ons will elude clear-cut explanations. The investigator is encouraged to perse-
vere and repeat the procedures as the situation demands. Perseverance is well warranted Groundwater
and surface water must coexist with the by-products of our civilization, and the acquisition of hiqh-qualitv
adsorption data is essential to the protection of water quality                                  H"a»iy

-------
                                        CHAPTER 1

                          ADSORPTION FORCES AND MECHANISMS

Before adsorption studies are discussed, a brief review is presented of the physicochemical forces and
mechanisms thought to be responsible for the adsorption of ions and molecules. Adsorption from solution
at the solid-liquid interface is a complex and imperfectly understood phenomenon.

These physicochemical forces can be broken down into eight categories (Reinbold et al., 1979; Griffin
and Roy, 1985; see references in Voice and Weber, 1983):
• London-van der Waals Attractive forces arise from momentary dipoles about atoms or molecules
caused by small perturbations of electronic motions. These dipoles induce small dipoles in neighboring at-
oms of opposite sign. Although the momentary dipoles and induced dipoles are constantly changing posi-
tion and sign, the net result is a weak attraction (4 to 8 kJ/mol for small molecules and atoms). These
forces are important in adsorption of organics and are generally attributed to the nonideal behavior in
gases. They also have been partially treated by quantum mechanical perturbation theory, which uses po-
larizabilities, ionization potentials, and the magnetic susceptibilities of the interacting atoms to explain vari-
ous phenomena such as adsorption.
• Coulomblc-electrostatic-chemical  An electrostatic force results from a charged surface due to iso-
morphous substitution in the mineral lattice (permanent charge) or protonation of surface oxygen and OH
groups (pH-dependent charge) and an oppositely charged species, which maintains the electroneutrality
of the surface. In layer silicates, substitution of octahedrally or tetrahedrally coordinated cations by cations
of lower valence results in a net negative charge. This excess charge can bring about the formation of a
diffuse layer of positively charged atoms or molecules about the colloid; the density of this layer is greater
at the surface, and then decreases exponentially to the level of the bulk solution. This type of reaction is
important in adsorption of inorganic ions and ionized organic molecules.

• Hydrogen bonding  A hydrogen atom is bonded to two or more other atoms; the "bond" is generally
conceived as an induced dipole phenomeon. No universal agreement has been reached on the best de-
scription of the hydrogen bond (Huheey, 1978), but it may be considered as  the asymmetric electronic dis-
tribution of the 1s electron of the hydrogen atom by very electronegative atoms (e.g., F, O, S, Cl).
Hydrogen bonding is probably more than simply an exaggerated dipole-dipole or ion-dipole interaction,
since these concepts do not account for molecular geometry in some cases  (see Huheey, 1978; Cotton
and Wilkinson, 1980). In reality, hydrogen bonds  may be delocalized covalent bonds, i.e., resonance
bonds or multiple-center bonds (Huheey, 1978). The energy of this attraction ranges from 8 to 42 kJ/mol.

•  Ligand exchange-anion penetration-coordination  Many atoms or molecules form coordinated com-
plexes with ligands that range in complexity from simple linear molecules to  extensive chelate complexes.
The coordinated complexes may carry a net charge that may be localized on some part of the complex.
These complexes may in turn be bonded to surfaces by hydrogen bonding or by polyvalent cation bridges
linking the complex to a charged surface. The possible geometrical arrangements of coordinated com-
plexes bonded to mineral faces are diverse.  The  bonded coordinated complexes may be displaced by
other coordinated complexes that better satisfy electroneutrality requirements (i.e., are stronger complex-
ing agents) while being constrained  by steric limitations. The energy of ligand exchange reactions with in-
organic ions ranges from 8 to 60 kJ/mol.
 • Chemisorption In this adsorption process an actual chemical bond, usually covalent, is formed be-
tween the molecule and surface atoms. A molecule undergoing chemisorption may lose its identity as the
 atoms are rearranged, forming new  compounds at the demand of the unsatisfied valences of the surface
 atoms. The enthalpy of chemisorption (~A H>29 kJ/mol) is much greater than physical adsorption. The
 basis of much catalytic activity at surfaces is chemisorption, which may organize molecules into forms
 that readily undergo reactions. Chemisorption and  physical adsorption are often difficult to distinguish
 from one another because a chemisorbed layer may have a physically adsorbed layer upon it. Moreover,
 some ligand exchange reactions are chemisorption processes.

-------
•  Dipole-dfpole or orientation energy This interaction results from the attraction of a permanent dipole
tor another permanent dipole. The resulting energy of attraction is less than 8 kJ/mol.

•  Induction or dipole-induced dipole This type of interaction results from the attraction of an induced
dipole brought about by either (1) a permanent dipole or (2) a charged site or species. The energy of at-
traction is less than 8 kJ/mol, but this force often adds to coulombic interactions.

•  Hydrophobic effect The exact nature of this adsorption mechanism is uncertain. Some investigators
believe that hydrophobic adsorption is primarily an entropically driven mechanism brought about by the
destruction of the physical cavity occupied by the solute in the solvent, and from the partial loss of struc-
  1! Au??f "^l60"'68Labout the solute' ordered by van der Waals forces (Horvath et al., 1976; Sinanoglu
and Abdulnur, 1965). Other researchers feel that the hydrophobic effect is the result of simple partitionino
Nonpolar organic solutes tend to migrate from the aqueous phase to hydrophobic surfaces on the adsorb-
ent (Dzombak and Luthy, 1984; Chiou et al., 1979,1983; Griffin and Roy, 1985).

-------
                                         CHAPTER 2

                           EFFECTS OF ADSORBENT PREPARATION

The process of preparing samples taken in the field for laboratory investigations can directly influence ana-
lytical results. Adsorbent samples (e.g., soils and clays) are usually dried so that they can be homog-
enized and stored until needed. However, studies have shown that the method of drying the sample may
alter its chemical properties, which in turn can influence the results of batch-adsorption procedures.

An early paper by Fujimoto and Sherman (1945) concluded that the concentration of exchangeable man-
ganese in 23 Hawaiian soils tended to increase as the samples were dried. A portion of their results is
given in table 1. The differences in the amount of exchangeable manganese in field-moist and air-dried
samples were minimal compared with the changes that occurred upon oven-drying or autoclaving. They
also found that the amount of exchangeable manganese tended to increase as the duration of air-drying
increased until about 8 to 10 weeks (fig. 1). Luebs et al. (1956) found that the amount of exchangeable K+
in 13 Iowa soils increased when the soils were air-dried for 2 months. However, a reduction in the mois-
ture content of the soils from 25 to 10% was required before appreciable changes in exchangeable K+
could be detected (fig. 2).

Drying soil samples may also have an effect on the stability of the organic matter in soils. Air-drying soils
generally stimulates soil microorganism respiration when they are rewetted, and Stevenson (1956) con-
cluded that the degree of metabolic activity varies directly with the concentrations of free amino acids and
other nitrogenous materials released during air-drying.

Birch (1958), continuing the work of earlier investigators, found that when either oven-dried or air-dried
soils were remoistened, a portion of the organic matter dissolved,  and that the magnitude of this decompo-
sition depended directly on the amount of organic matter present in the soil. He later concluded (Birch,
1959) that this decomposition after rewetting was caused primarily by microbial decomposition of water-
soluble organic matter.

An alternative hypothesis was proposed by Raveh and Avnimelech (1978). They postulated that organic
macromolecules in their natural pedological settings are aggregated by hydrogen bonds. When soils are
dried, the evaporation of water disrupts the hydrogen bonds, and the stability of the organic matter de-
                                             high-Mn soil
                                              16
                                                  18
                                                                  280-
                                                                   120-
                                                                   40-
                                                                             I
                                                                             10
                                                                                    20
                       Air-drying time (weeks)
 Figure 1  Effect of air-drying on the concentration of exchangeable
 manganese (adapted from Fujimoto and Sherman, 1945).
               moisture (%)

Figure 2  Relationship between the
concentration of exchangeable potassium
in the Harpster clay loam and moisture
content (adapted from Luebs et al., 1956).

-------
                      6-
                      4-
                 S
                 o
                      2-
                                                        1	
                                                       10

                                            Oven-drying time (weeks)
15
                   Figure 3  Effect of oven-drying at 105°C on the concentration of water-soluble
                   organic carbon in an Israeli calcareous clay loam (adapted from Raveh and
                   Avnimelech, 1978).


 creases. They also observed that the amount of water-soluble carbon in aqueous extracts increased as
 the length of oven-drying periods at  105°C increased (fig. 3).

 According to Bartlett and James (1980), one of the most noticeable effects of air-drying soils is an in-
 crease in the yellow or amber color of extracts, attributable to the amount of organic matter made soluble
 by drying. They found that the amounts of Al, Fe, and Mn in NH4OAc extracts (pH 4.8) of a soil subjected
 to 40 C for 12 hours were greater than those extracted from field-moist samples of the same soil.

 Drying soil samples has been reported to change the pH of the soil (or soil reaction). Van Lierop and
 MacKenzie (1977) found that oven-drying soil samples at 110°C tended to result in lower pHs relative to
 the pHs of field-moist samples of the same soils. The change in pH varied from 0.3 to 1.1 (table 2) Raveh
 and Aymmelech (1978) suggested that this increase in acidity was caused by the exposure of fresh
 organic surfaces containing acidic groups that were sterically hindered before drying The increase in
 nSfL3?^ ^ 3lSO c°n.sidered by Mortland and Raman (1968) who hypothesized a different mecha-
 nism. As the samples are dried, adsorbed cations more strongly polarize the residual water molecules
 making them more acidic than free water.

 Other studies have demonstrated that drying samples lowers the ability of soils to oxidize chromium
 (Bartlett and James, 1980), and can influence denitrification studies (Patten et al., 1980; Soulides and
 Allison, 1961) and other soil chemical processes that may have an indirect effect on batch-adsorption
 SlUdlGS.
         -         MM              nave also been documented; Ashton and Sheets (1959) found that
the herbicide ethyl-A/,A/-d.-/7-ProPylth,olcarbamate (EPTC) was adsorbed as a vapor to a greater extent by
air-dried soils than soils that were moist. The adsorption of EPTC was suppressed at higher moisture con-
tents because of the competition of the EPTC vapor and water molecules for adsorption sites Dao and
Lavy (1978) observed that the adsorption of atrazine by Nebraskan soils decreased with an increase in
soil moisture. They suggested that competition between the atrazine and water could account for this rela-
 '
Oven-drying may increase the hydrophobicity of a-soil, which in turn would enhance the adsorbent's af-
finity for hydrophobia solutes. Research has established that forest fires can increase the hydrophobicitv
of soil materials near ground surface. Heat-induced hydrophobicity studies by Debano et al (1976) sua-
gested that temperatures as low as 98° to 1 18°C for an exposure time of as little as 5 minutes can in-
crease the hydrophobicity of a sample as measured by water-drop penetration time. It is not certain

-------
Table 1  Effect of drying on exchangeable  Mn in four Hawaiian soils
(from Fujimoto and Sherman, 1945)

                                             Mn concentration (mg/L)
Soil
location
Kemoo
Koko Head
Kahuku
Waimanalo

pH (1:1 )•
4.2
7.1
7.6
8.6
Field-
moist
3.4
0.0
0.0
0.5
Air-
dried
4.5
4.3
0.4
0.4
Oven-
driedf
621.2
29.4
11.7
1.2
Auto-
clavedt
374.8
ND§
367.9
ND
*pH of a 1:1  soihwater suspension.
fOven-dried for 24 hours at 105°C.
^Autoclaved for 3  hours at 15 pounds pressure.
§ND,  not determined.
Table 2  pH of soil-water slurries (1:2 v/v) made with field-moist and
oven-dried (110°C) samples (from van Lierop and MacKenzie, 1977)

                        PH
Soil
1
2
3
4
5
6
7
8
9
10
Field moist
4.0
4.2
5.5
6.2
4.5
4.1
4.2
4.2
6.7
6.3
Oven-dried
3.0
3.7
5.2
5.8
4.0
3.7
4.1
3.9
5.6
5.8
ApH
1.0
0.5
0.3
0.4
0.5
0.4
0.1
0.3
1.1
0.5
Table 3  Effect of sample pretreatment on the Freundlich partition
coefficients (Kf) of acetophenone adsorption (Hassett et al.,  1980)

                                Freundlich partition coefficients

Sample
Sangamon
Crane Island

Fresh
1.07
0.90

Frozen
1.07
0.99
Air-
dried
0.95
1.04
Freeze-
dried
0.95
1.09
Oven-
dried
0.84
1.09
Mean*
0.98
1.03
1.00
1.01
0.97
*Means were not significantly different at the 5% level.

-------
 whether this heat-induced hydrophobicity will influence adsorption results obtained by batch techniques.
 Hassett et at. (1980), for example, found that the adsorption behavior of acetophenone on two alluvial silt
 samples was not significantly affected by various drying techniques (table 3). Oven-drying a Sangamon
 River sample appeared to generate a slightly lower Freundlich constant (Kf) relative to the values for fresh
 (field-moist), frozen, air-dried, or freeze-dried samples, but the difference was not significant at the 5%
 level of probability. As shown in figure 4, the distribution of isotherm points generated from air-dried sam-
 ples tended to be similar to the distribution generated from fresh field-moist samples.

 In contrast, Bartlett and James (1980) found that a soil sample that had been oven-dried at 40°C ad-
 sorbed more phosphate during a 6-hoUr equilibration than samples that were kept moist. Harada and
 Wada (1974) reported that air-drying their soil samples resulted in slight but significant increases in the
 cation exchange capacity and anion exchange capacity. Bar-Yosef et al. (1969) found that oven-drying
 kaolinite at 110°C reduced the amount of phosphate that could be desorbed relative to clay samples that
 were not heat-treated. They thought that possibly during drying the phosphate tetrahedron may have
 changed its stearic configuration to a form more conducive to bonding.

 In summary, drying adsorbent samples to homogenize and store the samples until they are  needed may
 influence the results obtained by batch-adsorption studies. Bartlett and James (1980) concluded that air-
 drying or oven-drying both may be viable methods of sample preparation if the potential changes in ad-
 sorbent properties are understood and confronted. However, understanding and confronting these
 changes may be research projects in themselves.

 Recommendations  Oven-drying of adsorbents is not an advisable technique to accelerate drying even
though air-drying may take several days with large bulk samples. Air-drying samples by the  atmosphere
 minimizes changes that may occur from drying and is the most practical approach at this time. The Ameri-
can Society for Testing and Materials (ASTM) defined air-drying as a process of partial drying (of the sam-
ple) to bring  its moisture content near equilibrium with the atmosphere in the room in which further
reduction and division of the sample is to take place (ASTM,  1979). Air-drying should be kept to the mini-
mum necessary to allow preparation of the sample, and a stable condition should be provided for meas-
urements of the sample, such as weighing. Air-drying anaerobic soils and sediments requires special
                    800-
                                                       1	T
                                           400         600       800
                                  Equilibrium acetophenone concentration (mg/L)
                 Figure 4  Adsorption isotherm of acetophenone by fresh field-moist and
                 air-dried samples of Crane Island alluvium (adapted from Hassett et al., 1980).
                                               8

-------
handling to prevent the relatively reduced materials from oxidizing if exposed to the atmosphere.
Anaerobic materials can be "air-dried" in a glove box or glove bag supported by a continuous supply of
dry oxygen-free nitrogen or argon gas.

-------

-------
                                        CHAPTER 3

                                 EFFECTS OF TEMPERATURE

Adsorption at the solid-liquid interface tends to occur when the attractive forces between the surface and
ionic solutes are greater than those between the solutes and the solvent (Zettlemoyer and Micale, 1971).
The adsorption of an ionic or polar solute is often the result of a thermodynamically favorable change in
the enthalpy (AH) (Hassett et al., 1981) or sometimes of a favorable change in the entropy (S) of the
system in which the -T AS term from the Gibbs-Helmholtz equation compensates for the positive value of
AH (Thomas, 1961), where  T is the temperature of the system. The adsorption of nonpolar organic sol-
utes is thought to result primarily from a thermodynamically favorable change in entropy (AS) involving lit-
tle energy transformation as heat. Thus, the adsorption behavior of ionic or polar solutes will probably
show some temperature dependency, whereas the adsorption of nonpolar solutes may not be greatly influ-
enced by the temperature of the system. The direction and magnitude of temperature dependency will de-
pend on the specific solute-soil system.

An early paper by Jurinak and Bauer (1956) reported that adsorption of zinc by calcite was exothermic;
the amount of zinc adsorbed decreased with increasing temperature. In contrast, Kuo and Mikkelsen
(1979) studied the adsorption behavior of zinc by soils at temperatures ranging from 10° to 35°C and
found that zinc adsorbed endothermically; increased adsorption was associated with higher temperatures.
Kinniburgh and Jackson (1981) reviewed the literature on cation adsorption by soils and concluded that
the effects of temperature were usually small, but in some cases temperature significantly influenced ad-
sorption data.

The adsorption of phosphate by soils and soil materials is often endothermic (Low and Black, 1950; Gard-
ner and Jones, 1973; Griffin and Jurinak, 1973a; Singh and Jones, 1977; Taylor and Ellis, 1978; Roy et
                  400 -I
                                                     35 C
                                                                         15"C
                                  20            40            60
                                   Equilibrium arsenic concentration (mg/L)
           Figure 5  Arsenate adsorption isotherms by Catlin soil at 15°, 25°, and 35°C, and at pH 6.6.
                                              11

-------
 al., 1989). The adsorption of arsenate was also found to be an endothermic reaction (fig. 5). The amount
 of arsenate adsorbed in equilibrium with a solution concentration of 50 mg/L total arsenic at 15°C
 was about 31% (mass basis) less than that observed at 25°C and approximately 51% less than the
 amount adsorbed at 35°C. Hassett et al. (1983) found that in contrast to the adsorption of ionic species,
 the adsorption of the nonpolar solutes phenanthrene and oc-naphthol by soils was largely unaffected by
 temperature variations from 15° to 35°C (table 4). The adsorption of 1 ,2-dichlorobenzene by a soil sample
 studied by Chiou et al. (1979) was not affected much by temperature differences between 3.5° and 20°C,
 but the adsorption of 1 ,1 ,1 -trichloromethane was reduced at the lower temperature.

 Weber et al. (1 983) found that the adsorption of Aroclor 1 254 by a Saginaw River sediment was tempera-
 ture-dependent; adsorption was reduced over a 10-degree temperature range. Moreover, Voice (1986,
 written communication) demonstrated that the adsorption of 2,4,5,2',4',5'-hexadiclorobiphenyl by a Lake
 Michigan sediment decreased with decreasing temperature over a 20-degree temperature range.

 The effect of temperature on adsorption data is ultimately linked to the thermodynamics of the adsorption
 process. This relationship may be approximated by a Clausius-Clapeyron-type equation, integrated over a
 narrow temperature range,
                                                         AH'                               [1]

where Ct and Cz are the equilibrium concentrations of a solute at two different temperatures, 7, and T2,
and AH' is the apparent heat of adsorption.

Apparent heat-of-adsorption values can be used as estimates of the amount of heat energy isothermally
released or absorbed during the course of adsorption, although more correctly, these values probably re-
flect the heat of the overall reaction. Equation 1 can be rearranged as
                              Cfe/C, « exp((1/72-1/7|)AH'/H)
[2]
Equation 2 can be used to estimate the effects of temperature if the AH' of the specific adsorbent-solute
system is known. If AH' is small, such as with the adsorption of some hydrophobic organic solutes, then
the ratio of Cz to Ci will be close to 1. In other words, the solute concentration at 7, will be nearly the
same concentration as at 72l given that all other conditions are the same; such results for some organic
compounds are given by Hassett et al. (1983). In contrast, the adsorption of phosphate is often associ-
ated with relatively large AH' values. Consequently, phosphate adsorption may be sensitive to ambient
temperature fluctuations. Moreover, the larger the temperature fluctuations (the difference between 7,
and 72 in eq. 2), the greater the potential for the equilibrium solute concentrations to be affected. Thus, ad-
sorption experiments are usually conducted with temperature-controlled water baths or constant-tempera-
              Table 4 Effect of temperature on Freundlich adsorption constants (Kf) for
              phenanthrene and a-naphthol (from Hassett et al., 1983)
Solute
Phenanthrene
a-Naphthol
Soil
(EPA number)
5
15
5
15
Freundlich constant (Kf )
15°C
328
117
5.4
19
25°C-
304
151
5.5
25
35°C
340
126
7.7
31
                                               12

-------
ture rooms. If such facilities are not available or are impractical, the laboratory work should be conducted
in rooms where the ambient temperature fluctuates by no more than 6°C (e.g., 22 ± 3°C). This 6-degree
range is based on the assumption that a "typical" heat-of-adsorption value for most solutes of environ-
mental significance is approximately 20 kJ/mol. This suggested range should be acceptable for most situ-
ations, but if adsorption of the solute results in a comparatively large heat of adsorption, more rigorous
temperature control may have to be implemented.

Recommendations Batch-adsorption procedures should be conducted under constant-temperature con-
ditions, if available, or in rooms where the ambient temperature is fairly constant (e.g., ± 3°C). When
batch experiments are performed, the temperature of the room should be recorded and treated as a po-
tential variable that can influence the data, and therefore may be useful in the interpretation of the results.
                                               13

-------

-------
                                         CHAPTER 4

                  STABILITY OF NONIONIC ORGANIC SOLUTES IN SOLUTION

In conducting batch-adsorption procedures, investigators must consider the physicochemical stability of
the solute in solution. Processes such as photodegradation, hydrolysis, and/or microbial degradation can
potentially contribute to a decrease in solute concentration concomitantly with adsorption, and these
changes may even occur before the solution contacts the adsorbent. A standard method is used to deter-
mine hydrolysis rate constants of organic compounds in water (ASTM, 1986a) and to conduct aqueous
photolysis tests (ASTM, 1986b). The following procedures can be used as simple screening tests to iden-
tify potential problems.
• Photolysis Photoreactive solutes that absorb light at wavelengths greater than 290 nm may be subject
to rapid photolysis in glass containers. For example, the half-life of hexachlorocyclopentadiene (C-56)
was found to be less than 5 minutes when exposed to sunlight (Chou and Griffin, 1983). Therefore, pre-
cautions should be taken to ensure that substances such as these are protected from light, not only sun-
light but laboratory lights as well. Appropriate measures include using amber glass, wrapping glassware
in aluminum foil, or adopting any other suitable technique that will eliminate photolysis transformations. A
simple aqueous screening test can help determine the stability of the solute(s) in the presence of light.
This procedure was designed to eliminate volatilization losses and ensure that only reductions in concen-
tration due to photolysis are measured during the test.

        Photolysis test Place the initial stock solution into either a 30- or 50-mL borosilicate glass
        hypovial, and fill the vial to eliminate any head space. Seal the vial with a Teflon-faced
        septum and aluminum crimp cap to prevent volatilization, and place replicate samples in
        sunlight for 2,4, and 6 hours. Analyze duplicate samples of the unexposed solute to de-
        termine the concentration at time 0 and in two freshly opened hyppvials after 2,4, and 6
        hours of exposure. Also determine the concentration of the solute in each of two control
        vials (wrapped with aluminum foil or in amber glass vials) that have also been exposed.
        Select an analytical method that is most applicable to the analysis of the specific solute
        under study. Chromatographic methods are generally recommended because of their
        chemical specificity in analyzing the parent compounds without interference from impuri-
        ties. If the results  indicate the solute is photoreactive, then all subsequent tests and ad-
        sorption studies must be conducted under conditions that prevent exposure to light during
        the reactions and analytical steps.
• Hydrolysis Hydrolysis is an important degradation path for certain classes of nonionic solutes, and an
investigator should know whether the solute under study will be subject to hydrolysis during the period of
the adsorption study. Otherwise, the amount of solute adsorbed by soils or sediments could be overesti-
mated if changes in solution concentration due to hydrolysis are not taken into account. Details of the hy-
drolysis reactions of various types of compounds can be found in many kinetics texts (e.g., Laidler, 1965;
Frost  and Pearson, 1961). Discussions of hydrolysis  from an environmental point of view have also been
published (Mabey and Mill, 1978;  Tinsley, 1979). The temperature of a hydrolysis screening test proce-
dure should be kept constant. The temperature used in the hydrolysis test procedure should be the same
as that used in the adsorption experiments. The pH is also important and the hydrolysis screening test
should be carried out at the same pH range that will be used in the adsorption studies. Measures to pre-
vent photolysis should be implemented as previously discussed. In some cases, the hydrolysis of solutes
may be enhanced by the presence of other substances such as iron, which catalyze the rate of hexachlo-
rocyclopentadiene hydrolysis under conditions of low pH (Chou and Griffin, 1983). Therefore, the composi-
tion of the test solution must be considered.

        Hydrolysis screening test Fill a 30- or 50-mL hypovial completely with the test solution to
        eliminate any head space, then seal the vial  with a Teflon-faced septum and aluminum
        crimp cap. Place  replicate samples in a constant-temperature room or water bath for 6,
        12,24, and 48 hours. Select an analytical  method that is most applicable to the analysis
        of the specific compound under study and analyze duplicate samples for the concentra-
        tion of the chemical substance at time 0 (control), and in two freshly opened hypovials af-
                                               15

-------
        ter 6,12,24, and 48 hours. If significant hydrolysis is indicated by the results of this test,
        this must be considered in the interpretation of results from adsorption studies, and spe-
        cial care should be given to the handling of flasks and to the analytical steps used.

•  Microbial Degradation Microbial degradation can decrease the solution concentration of the solute,
leading to an overestimation of the amount adsorbed by the adsorbent. Therefore, for easily degraded
(labile) compounds, a batch technique will measure "apparent adsorption," which is a combination of ad-
sorption and degradation (and hydrolysis as indicated by the results of the hydrolysis screening test). The
influence of microbial degradation on "apparent adsorption" of phenol by soil was studied by Scott et al.
(1982). They found that Freundlich (Ki) values for the adsorption of phenol by nonsterile soils increased
linearly with time with a Palouse silt loam and increased exponentially with time with a Captina silt loam.
The Freundlich (Kf) values associated with adsorption by sterile soils remained essentially constant after
8 hours. A similar study for p-cresol was also reported by Boyd and King (1984). Their data indicated that
under aerobic conditions, p-cresol degradation began within 10 hours, and complete degradation oc-
curred within 48 hours or less for initial p-cresol concentrations of 5,10,20, and 50 u.g/L The adsorption
of organic compounds, such as phenol or other labile organics that are degraded within the time required
to attain adsorption equilibrium, cannot be evaluated accurately without accounting for or eliminating mi-
crobial degradation losses.

        Biodegradation screening test The most common approach used to screen whether an
        adsorbate undergoes biodegradation is to conduct kinetic studies by using sterile and
        nonsterile soil. Before the kinetic studies, the weighed soil is placed into a reaction bottle
        and then autoclaved three times at 2-day intervals, each time for 2 hours at 120°C and at
        1.4-bar pressure (Scott et al., 1982). (See chapter 2 to help evaluate the possible
        changes in adsorbent characteristics caused by autoclaving.)
          Bulk solutions of the solute are prepared in sterilized distilled water and passed
        through a sterilized 0.22-u.m membrane. Then, known amounts of the solutions are trans-
        ferred to the sterilized and nonsterilized reaction bottles and sealed with sterilized Teflon-
        faced septa and aluminum crimp caps. All samples are equilibrated at constant
        temperature for 4,8,16, 24, and 48 hours. At the end of each equilibration period, the
        solid-phase soil particles are separated from the solution phase by centrifuging duplicate
        reaction bottles at 2,000 rpm for 1 hour. Aliquots of the supernatant solution are taken
        with a syringe through a hole and septum in the caps of the bottles. Select an analytical
        method that is most applicable to the analysis of the specific solute under study. The only
        purpose of the suggested test procedure is to screen for biodegradability.
          If the test indicates the solute would be biodegradable to a significant extent during
        the period of the adsorption test, then the reaction times or temperatures may have to be
        modified to reflect this result. The results of the adsorption study must then be interpreted
        in the context of the solute equilibration time and the environmental significance of the
        biodegradation of the solute relative to its adsorption affinity for soil materials.
                                                16

-------
                                        CHAPTER 5

                                  EFFECTS OF SOLUTION pH

The adsorption behavior of ionic and ionizable inorganic and organic solutes by soils and soil materials is
often influenced by the pH of the soil-water system. In general, the adsorption of inorganic cations in-
creases with increasing pH (Kinniburgh and Jackson, 1981). For example, Griffin and Shimp (1976) re-
ported that the amount of lead adsorbed by kaolinite from a landfill leachate was pH dependent; the
amount of lead removed from solution increased with increasing pH. In their batch-adsorption experi-
ments, as with similar studies, the pH of the soil solutions was periodically adjusted to the indicated pH by
the addition of either dilute acids or bases. A sharp change in slope of the isotherms between pH 4 and 6
was attributed to the precipitation of PbCOs. The reduced adsorption at the lower pH values was  attrib-
uted to the increase in competition for adsorption sites by H+ and by Al   resulting from the dissolution of
the clay. Similar examples for Cd, Cu, and Zn (fig. 6) show that higher pH values have been associated
with greater removal from solution.

The pH of the soil solution has also been shown to have a direct effect on the adsorption of anionic sol-
utes. In contrast to cationic solutes, anion adsorption is generally enhanced in acidic environments, al-
though some anionic solutes are adsorbed to a greater extent in alkaline systems. Parfitt (1978)
generalized that sulfate adsorption by soils becomes essentially insignificant above pH 8, while the ad-
sorption maxima of boric acid and silicic acid appears to correspond to a pH of approximately 9. White
(1980) generalized that phosphate adsorption by goethite decreased uniformally between pH 3 and 12,
while the magnitude of phosphate adsorption by alumina passed through a maximum value between pH 4
and 5. Griffin et al. (1977a) found that the adsorption of chromium (VI) at  low concentrations by kaolinite
passed through a maximum value between pH 4 and 5 (fig. 7). No adsorption occurred above pH 8.5.
                                                              r
The adsorption of arsenic as arsenate (As(V))  is also pH dependent with  lower pHs resulting in greater ad-
sorption (fig. 8). The adsorption of molybdate by soils also appears to exhibit a maximum value at pH 4
(Parfitt, 1978). This trend, characteristic of most inorganic oxyanions, is thought to result from the in-
creased positive charge due to the increased protonation of surface hydroxyls associated with  the edges
of colloidal particles and hydrous metal oxides in acidic environments. The adsorption behavior of arsenic
as arsenite (As(lll)) may (Griffin et al., 1977b) or may not (Pierce and Moore, 1982; fig. 8 of this report) be
strongly dependent on pH.

The adsorption of ionizable organic solutes is also influenced by the pH of the soil solution. For example,
Frissel and Bolt (1962), Weber (1966), and Hance (1969) showed that the adsorption of the triazine in-
creased as the pH decreased (fig. 9). At low pHs, the triazine solutes may have been increasingly proton-
ated, which increased the magnitude of coulombic interaction with negatively charged sites on clay
surfaces.  McGlamery and Slife (1966) found that the adsorption of atrazine by the Drummer clay loam
was influenced more by pH than by temperature. Frissel and Bolt (1962) also  presented data illustrating
the pH dependency of the adsorption of other ionizable organic compounds (herbicides MCPA, 2,4-D,
DNBP, and 2,4,5-T) by clays. The adsorption of DNBP (fig. 10), for example, sharply decreased as the
pH of the  system increased from approximately pH 4.7 to 6. In alkaline solutions (pH > 7), DNBP adsorp-
tion was reduced because of negative adsorption,  i.e., clay repelled DNBP. In this pH range, DNBP oc-
curred largely as neutral molecules, since the pK of the organic solute was 4.35. The adsorption of
benzidine also followed a similar pattern. The ionization constants of benzidine are 4.3 and 3.3 (pKb1 and
pKb2). Consequently, Zierath et al. (1980) found that the amount of benzidine  adsorbed by two soils de-
creased when the solution pH was increased from a pH of 5 to 11. Benzidine  can exist in solution as both
ionized (cationic) species and a neutral species. As the pH of the suspensions was increased, a larger
portion of the total amount of benzidine existed as the neutral  form. Both  species are subject to adsorp-
tion, although the cationic form should be adsorbed to a much greater extent due to coulombic interac-
tions.

The adsorption behavior of neutral, nonpolar hydrophobic organic solutes appears to be largely unaf-
fected by the pH of the soil-water system. Hassett et al. (1980) found no correlation between the adsorp-
                                                17

-------
                   8 -
                             50
                                    100      150      200      250
                                       Equilibrium concentration (mg/L)
                                                                      300
                                                                              350
             Figure 6  Zinc, copper, and cadmium adsorption from a Du Page County landfill leachate
             by kaolinite at 25°C at various pH levels (adapted from Frost and Griffin, 1977).

tfon behavior of polycyclic aromatic hydrocarbons (PAH) and the pHs of 14 soils ranging from pH 4.5 to
8.3. Correlations between the adsorption constants and the actual pH of the solutions were not attempted.

In the present study, the adsorption of the PCB Aroclor 1242 was not significantly influenced by the pH of
three different soil suspensions (fig. 11). The linear Freundlich constants (Kd) were essentially constant
over the range of pH 3 to approximately 10. However, humic and fulvic acids are considered to be the ma-
jor constituents of organic matter in soil (Stevenson, 1982). Ghosh and Schnitzer (1980) postulated that
the macromolecular structure of humic and fulvic acids may be influenced by solution pH and ionic
strength. They may change from rigid spherocolloids with a limited adsorption capacity to flexible, linear
                                                 18

-------
                      Figure 7  Chromium (VI) adsorption by kaolinite at 25°C at various pH levels.
                      Chromium concentrations shown (4,8, and 17 mg/L) are the initial concentrations
                      added (modified from Griffin et al., 1977a).
   1.6-
   1.4-
   1.2-
I  °-8-
E
i  0.6


I
to
_l
   0.4-
   0.2-
   0.0
                                                                                                    10
                                                     pH
           Figure 8  Langmuir-type maximum (mM/g) for the adsorption of arsenic as As(V) and As(lll) by amorphous
           iron hydroxide (from Pierce and Moore, 1982).


     polyelectrolytes with a greater capacity to retain nonpplar organic compounds. The influence of pH and
     ionic strength on the adsorption of hydrophobic organics is a current area of research.

     Recommendations The potential influence of pH on the results generated by batch-adsorption proce-
     dures depends on the system under study. The equilibrium pH of the soil-solute mixtures should be deter-
     mined before separating the solution from the soil or soil component suspension. The measurement
     should be given along with the adsorption data. For anaerobic adsorbent-solute systems, pH measure-
     ments should be conducted in a glove box or bag so that the suspensions do not oxidize when the con-
     tainers are opened. The failure to measure and report pH data may render the adsorption data impossible
     to interpret in a meaningful way.
                                                    19

-------
   100


    80
 n
 SS  40-

    20-
6

i
a
                    5        6
                        pH
Figure 9  Effect of pH on the adsorption of triazines by    Figure 10  Effect of pH on the adsorption of different
a Ca-montmorillonite sample (adapted from Hance, 1969).    ionizable organic solutes by an illite sample (1) 2,4-D,
                                                       (2) 2,4,5-T, (3) MCPA, and (4) DNBP (adapted from Frissel
                                                       and Bolt, 1962).
200-
"S
.J
e
£-150-
constant,
•5
=5 100-
3
it
50-
0-
••^"^



f EPA-14




Cecil clay

                                                                         10
                                                       pH
                                                                                    12
                 Figure 11  Adsorption of the PCB Aroclor 1242 by a synthetic goethite, a Cecil clay,
                 and EPA-14 soil samples as a function of pH at 24°C.
                                                       20

-------
                                          CHAPTER 6

                                 EFFECTS OF IONIC STRENGTH

The ionic strength of the solution may have several direct and indirect effects on adsorption data. The ex-
tent of these effects depends on the magnitude of the ionic strength and on the concentration, composi-
tion, and charge of the ionic constituents.

Ionic strength may affect adsorption data in two ways: (1) by changing solute activity, and (2) by changing
the thickness (and therefore properties) of the diffuse electrical double layers associated with colloidal par-
ticles. Because of the shielding effect of neighboring ions, the activity of most solutes tends to decrease
as the ionic strength of the solution increases. However, beyond a threshold ionic strength (often in very
concentrated solutions such as brines), the activities of some ionic constituents reverse themselves and
steadily increase, finally yielding activities exceeding those of their original concentrations (fig. 12). Thus,
adsorption data based on solute concentrations rather than ionic activity may produce calculated results
that disagree with observed results because of the departure of concentrations from ideality in nondilute
systems. Discussion of this topic may be found elsewhere (Atkins, 1982; Bohn et al., 1979; Bolt and
Bruggenwert, 1978; Garrels and Christ, 1965; Stumm and Morgan, 1981).

A basic tenet of Diffuse Double Layer Theory states that the physical thickness of the electrical double
layer composed of adsorbed cations around a colloidal particle is inversely proportional to the ionic
strength of the bulk solution. This phenomenon can affect not only exchange and adsorption reactions at
the solid-liquid interface, but may control the physicochemical properties of the material, such as hydraulic
conductivity, at the macroscopic level.
       1.2
       1.0
   I 0.8
   it
   •S  0.6
   %
   c
   o
    E   0.4
      0.2
       0.0
                                                                                       Na"*
                                                                                       Ca2 +
                                                                                   SO4
                               I
                                             j_
                                                          j_
                      0.005   0.01
0.05    0.1     0.2
Ionic strength (M/L)
0.5
1.0    2.0
5.0
           Figure 12  Ratio of concentration to activity (i.e., single ion activity coefficient) versus ionic
           strength for some common ions.
                                                 21

-------
 Some Investigators, attempting to minimize changes in ionic strength in the construction of adsorption iso-
 therms, have added a water-soluble compound to serve as a background electrolyte (sometimes referred
 to as a support medium or background ionic medium) to the solutions containing the solute(s) under
 study. The selection of background electrolytes and concentration has varied considerably, and the ration-
 ale for the choice has rarely been explained (Ryden and Syers, 1975). The addition of a background
 electrolyte  has been observed to have no measurable effect in some soil-solute systems, whereas both
 synergistic and antagonistic effects have been observed in other systems. The effect of ionic strength on
 phosphate  adsorption has received much attention. Helyar et al. (1976) concluded that phosphate adsorp-
 tion by gibbsite was independent of ionic strength in the range of 0.002 to 0.02 M when the ionic strength
 was controlled by NaCI, KCI, and MgCI2. Ryden and Syers (1975) and Ryden et al. (1977) reported that
 phosphate  adsorption by two soils in a 40-hour interval increased as the ionic strength of the solutions
 was increased by the addition of 10~3 to 1 M NaCI. The adsorption of selenite by goethite was reported by
 Hingston et al. (1968) as being insensitive to ionic strength in the range of 0.01 to 1.0 M.

 In many studies, polyvalent cation salts promoted phosphate adsorption (Barrow, 1972; El Mahi and
 Mustafa, 1980; Fox and Searle, 1978; Heylar et al., 1976; White, 1980). Helyar et al. (1976) speculated
 that Ca2+ may act as a potential determining ion; El Mahi and Mustafa (1980) suspected that the solubility
 of solid phosphate compounds was exceeded (see Anderson et al., 1981).

 The relationship between ionic strength and the adsorption of organic solutes has also been examined. In-
 creasing the ionic strength from less than 0.01  to 0.1 N resulted in a significant increase in adsorption of
 2,4,5-T (Koskinen and Cheng, 1983). This trend has been observed with other weakly acidic herbicides,
 such as picloram (4-amino-3,5,6-trichloropicolinic acid) (Farmer and Aochi, 1974) and 2,4-D (2,4-dichloro-
 phenoxyacetic acid) (Moreale and Van Bladel,  1980).  The increase in adsorption of these weakly acidic
 herbicides was attributed to a decrease in pH. A decrease in pH would increase the proportion of the mo-
 lecular species, which could then be adsorbed. Choi and Aomine (1974) found that increasing the ionic
 strength at  constant pH decreased the adsorption of pentachlorophenol (weak acid: pKa = 4.5). The
 amount of decrease was dependent on the anion used to adjust the ionic strength of the solution contain-
 ing the pentachlorophenol. In batch-adsorption studies, Abernathy and Davidson (1971) found that the ad-
 sorption of fluometuron (1,1-dimethyl-3-(a,a,a-trifluoro-/77-tolyl)urea) was decreased and prometryn
 (2,4-i>/s(isopropylamino)-6-(methylthio)-s-triazine) was increased by increasing the CaCI2 concentration
 from 0.01 to 0.5 N.

 In experiments designed to evaluate the effect  of solution ionic strength on 2,4,5,2',4',5' -hexachloro-
 biphenyl (HCBP) adsorption, Horzempa and DiToro (1983) found that the Freundlich constant (Kf) ap-
 peared to be only slightly influenced by increasing NaCI concentration from 10"4 to 10"2 M. But in similar
 experiments, CaCI2 significantly affected the Kf values over the same concentration range.

The use of  background electrolytes may also promote competitive interactions between the ions derived
from the background electrolyte and the solute(s) under study. (Competitive interactions are discussed in
chapter 11.) For example, Griffin and Au (1977) found that the adsorption of Pb by montmorillonite was re-
duced when 0.1 M Ca(CIO4)2 was used as a background electrolyte. The excess Ca2+ in solution was
also adsorbed by the clay, reducing the number of adsorption sites available to Pb. Other "side reactions"
 may take place that can complicate batch-adsorption data; Na-Ca and Na-Mg exchange reactions on ben-
ton'rte were unaffected by CIO4 in a study by Sposito et al. (1983), whereas Cl~ appeared to serve as a re-
actant in the exchange reactions, rather than serving as an "inert" background electrolyte. The formation
of CaCI* and MgCI* complexes may have caused the  observed exchange behavior. The appropriateness
of the use of a background electrolyte depends on three factors: (1) the specific conceptual model of the
 adsorbent-solute system envisioned by the investigator, (2) the chemical nature of the system itself, and
 (3) the overall objectives of the investigation and the intended use of the data.

The position taken in developing the batch-adsorption procedures presented in this document was gov-
 erned by the philosophy that they should be simple and designed primarily for routine use. Thus, the use
                                              22

-------
of a background electrolyte was rejected in anticipation that the inherent ionic strength of the solutions will
be influenced by the chemical constituents occurring in the leachate or extract, and those derived from sol-
uble constituents in the particular clay or soil under investigation.

Recommendation Measure the electrical conductivity of the equilibrated soil-solution so that the ionic
strength of the solution can be calculated by the relationship given by Griffin and Jurinak (1973b),

                                      /=0.0127xEC                                          [3]

where / is ionic strength in moles per liter, and EC is electrical conductivity in decisiemens per meter. For
anaerobic adsorbent-solute systems, EC measurements should be conducted in a glove box or bag so
that the suspensions do not oxidize when the containers are opened. The failure to measure and report
EC data and/or ionic strength can make adsorption data difficult to interpret.
                                               23

-------

-------
                                          CHAPTER 7

                               EFFECTS OF PHASE SEPARATION

A search of the literature indicated that very few researchers have used a filtration technique to separate
the liquid and solid phases before analyzing the liquid phase in batch-adsorption studies. This is probably
due to the potential of the filter membranes to retain significant quantities of the solute, particularly or-
ganic compounds. Luh and Baker (1970) found that a correction factor was necessary to account for re-
tention of   C-labeled materials on the filters used in their study. The factor was reasonably constant, but
the filtration technique was abandoned for a centrifugation technique, which uses gravity to separate the
solids from the liquid phase. In a preliminary test, Yaron and Saltzman (1972) also abandoned the filtra-
tion technique because the filter paper retained parathion. In similar studies, Griffin and Chou (1980)
found that cellulose acetate membranes (0.45- and 0.22-u.m pore size) adsorbed significant amounts of
polybrominated biphenyls (PBBs) or hexachlorobenzene (HCB). The problem could be overcome but re-
quired a tedious presaturation technique. They showed that continuously passing nine 100-mL portions of
HCB-saturated water through the membranes saturated the adsorption sites and yielded constant and re-
producible values for the concentration of the compound passing through the membranes (fig. 13). Pre-
saturating the membranes by soaking in HCB-saturated water yielded results that were not significantly
different from results obtained by passing solution through the membrane.

The effects of centrifugation and filtration on arsenic concentrations were investigated (fig. 14). No signifi-
cant differences were found between filtration and centrifugation with  respect to solute concentrations.
We concluded that laboratories performing adsorption studies could be given the option of either filtration
or centrifugation without impairing the general usefulness of the results, as long as the affinity of the filtra-
tion membrane for the solute was evaluated adequately; failure to do so could lead to erroneous results.

Recommendation Solid and liquid phases should be separated by centrifugation unless the investigator
can clearly demonstrate that the filtration technique does not significantly affect the results.
                                                               A Wlillipore membrane (0.45 jUm)

                                                               • Millipore membrane (0.22 jUm)

                                                               ® Membrane presaturated by soaking
                                                            r
                                        5678
                                       Number of continuative filtrations
T
10
       11
T
12
       Figure 13  Effect of pore size and number of continuative filtrations of 100-mL aliquots of HCB-saturated
       water on the concentration of HCB in filtrates (Griffin and Chou, 1980).
                                                25

-------
         130-
                      110
                                          120
                          Arsenic concentration (mg/L)
                             CENTRIFUGATION
                                                               130
Figure 14  Distribution of arsenic concentrations in solutions that were
either centrifuged or filtered. Values obtained by the two methods were
statistically not significantly different (adapted from Griffin  et al., 1985).
                                 26

-------
                                         CHAPTER 8

                             EFFECTS OF THE METHOD OF MIXING

In theory, the mechanical device used to mix the solid-liquid mixture during the equilibration interval
should have no effect on the equilibrium distribution of solutes and adsorbates. But some studies show
that the mixing method can influence the resulting adsorption data. In a study on phosphate adsorption,
Barrow and Shaw (1979) compared three mixing devices: a reciprocating shaker, a rotating tumbler, and
a roller. They found that the amount of phosphate adsorbed was greatest when a reciprocating shaker
was used and tended to be less with a roller. They concluded that the more vigorous the agitation, the
greater the breakdown of soil particles and the more "new" sites exposed for phosphate adsorption.
Barrow and Shaw also concluded that the efficacies of the three agitation devices to thoroughly mix the
suspensions may have contributed to the differences.

In the development of ASTM procedure D4646 (ASTM, 1987) described by Griffin et al. (1985), a first-gen-
eration procedure was formulated involving the ASTM-A water-shake extraction method (ASTM,  1979). A
round-robin sensitivity analysis of this early procedure was performed by several  laboratories. The mixing
method influenced the amount of cadmium and arsenic adsorbed by a Catlin silt loam sample. When
shaking was more vigorous, greater amounts of solute were adsorbed. Results from the first sensitivity

      Table 5  First ASTM sensitivity analysis for cadmium and arsenic at high and  low initial
      concentrations with shakers and a  paddle stirrer as the mixing devices	

                                    Initial concentration*
Replicate
Lab no.
A


B


C1


C2


D


E


Mean
±SD
%CV
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3



200 |j.g/mL
Cd
16.87
13.24
11.36
83.8
88.2
86.7
1.88
1.77
1.70
26.5
21.5
10.0
3.2
3.2
2.9
2.7
2.9
2.7
21.2
±30.8
145.4
As



186
186
185
128
131
127
162
168
175



130
134
130
153.5
±25.6
16.7
10 jig/mL Shaker rate
Cd
0.080
0.034
0.022
0.166
0.159
0.176
<0.01
<0.01
<0.01
0.064
0.057
0.096
<0.01
<0.01
<0.01
0.007
0.008
0.008
0.073b
±0.064
87.7
As Strokes/min Throw (cm)

59 8

7.92
7.77 70 4
7.91
0.43
0.38 100 3
0.46
5.00
4.78 70 3
5.81

Paddle stirrer used

0.53
0.55 Not known
0.55
3.51
±3.32
94.5
      * Represents postprocedure solute concentrations.
      t Does not include values lower than the detection limit.
                                              27

-------
 analysis are reported in table 5. Large differences in concentrations between the laboratories yielded inter-
 laboratory coefficients of variation in excess of 145%. This first round of interlaboratory study clearly
 shows why a standard adsorption procedure was needed.

 To improve the consistency of interlaboratory results, a National Bureau of Standards (NBS) rotary extrac-
 tor was tested as the mixing system (fig. 15). A second sensitivity analysis was carried out (tables 6 and
 7) in which each of the participating laboratories used an NBS rotating extractor. The coefficient of vari-
 ation between the mean values for each laboratory reflects in part the precision of the mixing method. The
 coefficient of variation of Rvalues based on initial cadmium and arsenic.concentrations of 10 and 200
 mg/L were less than 8% and 12% for cadmium and arsenic, respectively. These results can be compared
 with those from the first round, in which shakers were predominantly used: the coefficients of variation
 were as great as 145% for similar concentrations (table 5). Because all other parts of the procedure were
 the same in both cases, the mixing method was concluded to be a primary contributor to the variation be-
 tween the interlaboratory means. The NBS rotary extractor was adopted as the method of choice because
 of the much lower coefficient of variation between laboratory means.

 Recommendations  For all adsorption  experiments, an NBS rotary extractor or its equivalent should be
 used during each phase of the construction of an adsorption curve, i.e., determining a soil:solution ratio
 (chapter 8), equilibration time (chapter 3), and of course the adsorption curves themselves. Adsorption
 data generated with other mixing devices may be valid; however, these data should not be routinely ac-
 cepted unless the investigator can document that the device used yields data comparable to those from
 an NBS rotary extractor or its equivalent. This documentation will help standardize results among
 laboratories.

                                                       2-Liter plastic or glass bottles
               1 /15-Horsepower electric motor
29RPM
            Figure 15  National Bureau of Standards rotary extractor (Diamondstone et al., 1982).
                                               28

-------
Table 6  Cadmium adsorption data from second. ASTM interlaboratory sensitivity analysis

Replicate
Lab no.
A 1
2
3
B 1
2
3
C 1
2
3
D 1
2
3
Mean
±SD
%CV
Table 7 Arsenic
Initial
cone
(ng/mL)
200
200
200
200
200
200
200
200
200
190
190
190



adsorption
with an NBS rotary extractor

Replicate
Lab no. (
A 1
/» i
2
3
B 1
U >
2
3
C 1
^^ i
3
D 1
2
3
Mean
±SD
%CV
Initial
cone
;^g/mL)
205
205
205
200
200
200
200
200
200
200
200
200


24-h
cone
(ng/mL)
35.7
36.2
34.6
31.8
35.8
36.8
35.6
35.6
35.0
31.0
30.0
31.0
34.1
±2.3
7.1
data from the
as the mixing
24-h
cone
(|4,g/mL)
180.3
180.3
182.0
175.5
178.0
170.7
186.3
177.3
175.0
160.0
180.0
180.0
177.1
±6.65
3.76

(mL%)
92.0
90.5
95.6
105.8
91.7
88.7
92.4
92.4
94.3
102.5
106.6
102.5
96.3
±5.2
6.6
second ASTM
device

"d
(mL/g)
2.74
2.74
2.53
2.79
2.47
3.43
1.47
2.56
2.85
5.0
2.22
2.22
2.75
±0.85
30.9
Initial
cone
(ixg/mL)
10.1
10.1
10.1
10.0
10.0
10.0
10.0
10.0
10.0
9.8
9.8
9.8



24-h
cone
(ng/mL)
0.114
0.126
0.125
0.110
0.135
0.165
0.127
0.127
0.132
0.130
0.110
0.120
0.127
±0.01
7.94
interteboratory sensitivity

Initial
cone
(|ig/mL)
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
12.0
12.0
12.0



24-h
cone
(|ig/mL)
5.76
5.85
5.89
5.52
5.40
5.48
5.57
5.64
5.59
6.80
6.80
6.90
5.93
±0.56
9.47

(mL/g)
1734
1567
1580
1798
1461
1214
1554
1554
1495
1487
1761
1613
1568
±156
9.97
analysis


Rd
(mL/g)
14.72
14.18
13.95
16.23
17.03
16.49
15.90
15.29
15.77
15.29
15.29
14.78
15.42
±0.92
5.99
                                            29

-------

-------
                                         CHAPTER 9

                 SELECTION OF A SOIL:SOLUTION RATIO FOR IONIC SOLUTES

The term "soil to solution ratio" refers to the ratio of the mass of the adsorbent sample to the volume of liq-
uid. For the purposes of these procedures, 1  mL of solution, regardless of its composition, weighs 1 g. To
construct an adsorption isotherm (curve), soihsolution ratios must be determined that will permit enough
solute to be adsorbed to result in measurable, statistically significant differences in solution concentration.
In these procedures, increasing the soihsolution ratio from a "lower" to a "higher" ratio, such as from 1:1
to 1:100, means that the volume of solution increases relative to the weight of the  soil material. If the
soihsolution ratio is too low, i.e., too much adsorbent or too little solution, most of the solute may be ad-
sorbed, forcing the investigator to attempt to  measure small differences in concentration between small
quantities of solute. If the ratio is too high, i.e., not enough adsorbent for a given volume, the changes in
the initial solute concentration may be very small, forcing the investigator to measure small differences in
concentration between large amounts of solute. Unfortunately, with inorganic and  polar organic com-
pounds, a suitable soihsolution ratio cannot be determined a priori. The soihsolution ratio of the ASTM pro-
cedure D4646 is 1:20 (ASTM, 1987). However, a single ratio cannot be used satisfactorily in all cases.

An empirical, systematic procedure to determine a suitable ratio for a given soil-water and concentration
range is given in chapter 17. A value of 10% to about 30% adsorption for the highest solute concentration
used is a useful criterion for selecting a soihsolution ratio. This will give a discernible decrease in solute
concentration that is statistically acceptable with respect to the initial concentration. Justification for this
guideline is given in chapter 12. An example of this type of approach is given in table 8. With a 1:4 soihso-
lution ratio, more than 90% of the cadmium initially added (200 mg/L) was adsorbed by both a Sangamon
Paleosol and Vandalia till sample. If the 1:4 soihsolution ratio was used to generate data at lower concen-
trations than the 200 mg/L used in this example, the equilibrium cadmium concentrations would be below
analytical detection limits. In contrast, when a 1:500 ratio was used at the lower concentration (10 mg/L),
about 60% of the cadmium initially added was adsorbed by the Sangamon sample. However, when the
200 mg/L Cd solution was used, only 9.8% was adsorbed at the same soihsolution ratio. Essentially, the
object is to select a soihsolution ratio that is a compromise. In this case, a ratio of 1:100 was chosen to
generate an adsorption isotherm because the amount of cadmium adsorbed from the high concentration
range of the isotherm (200 mg/L solution) was approximately between 10% and 30%; at the same time

               Table 8  Soiksolution ratio determination for the  Sangamon soil and  Vandalia
               ablation till using cadmium as the adsorbate	__^	
                                           Sangamon
                                          Cd adsorbed
Vandalia (ablation)

  Cd adsorbed
               Soiksolution ratio
Initial concentration, 200 mg/L
1:4
1:10
1:20
1:40
1:60
1:100
1:200
1:500
Initial concentration, 10 mg/L
1:100
1:200
1:500
1:1000
722
1631
2792
4246
5165
6250
7500
9250

957
1736
3178
4325
95.2
86.1
73.7
56.0
45.4
33.0
19.8
9.8

91.1
82.7
60.5
41.2
635
1359
2143
3012
3441
3880
4560
4900

840
1474
2215
—
94.1
76.2
44.3
25.4
19.1
13.0
8.0
4.6

80.0
70.2
42.2
—
                                                31

-------
                       remaining fr°m a low °°ncentrati°n 00 mg/L) solution was also within analytical


This rationale for selecting soilsolution ratios is illustrated in figures 16 and 17. In each figure, the amount
of solute remaining in solution after 24 hours is plotted against the soihsolution ratio. The speckled area
approximates the desired solute concentration after 24 hours of mixing based on 10% to 30% solute ad-
sorption. When the data points or lines connecting the data points fall within this speckled area or "adsorp-
tion target zone," the corresponding soihsolution ratio will usually yield satisfactory results. The adsorption
behavior of seven soil materials with respect to arsenic is shown in figure 16; here, a 1 -10 ratio was cho-
sen to construct adsorption isotherms with six of the seven adsorbents. Figure 17 illustrates the same con-
cept with six samples using cadmium as the solute.  A1:10 ratio was chosen for the Tifton loamy sand
but any ratio between 1:10 and 1:4 probably would have yielded satisfactory results. A 1 -20 ratio was '
                                                                               Vandalia Till
                                                                               (ablation phase)
                                                                               Sangamon Paleosol
          1:100  1:40   1:20
                                   1:10

                                Soihsolution ratio (mass/volume)
1:5

                                              32

-------
                      180-f
                      160-
                                                                 Tifton loamy sand
                        1 :100 1 :40 1:20     1:10               1 :5
                                     Soil:solution ratio (mass/volume)
                 Figure 17  Distribution of cadmium concentrations after 24 hours of contact
                 with different soil materials as a function of soil:solution ratio.

Comparison of the two figures indicates that the adsorption of arsenic was essentially a linear function of
the soihsolution ratio, whereas the adsorption behavior of cadmium appeared to be influenced by the
soihsolution ratio: as the ratio of soil to solution decreases, progressively less cadmium was adsorbed per
gram of adsorbent. The significance of this trend is discussed in the next chapter.

The same type of rationale may be applied to solutions containing more than one solute of interest. A
laboratory extract of a metallic waste sample (see appendix B) will help to illustrate this point. The aque-
ous extract of the waste contained several aqueous constituents of interest, and a suitable soihsolution ra-
tio had to be determined for each solute. If one single ratio could be used for all of the solutes with each
given soil, that would be ideal, but this is not always possible. For example, the concentration of zinc in
the extract (550 mg/L) was much larger than that  of barium (2.26 mg/L).

A1:20 soil-.solution ratio for the Sangamon sample (table 9) resulted  in 32.5% of the zinc being adsorbed,
but the same ratio also  resulted in 96.2% of the lead in solution being adsorbed; the solution concentra-
tion of lead approached the limits of detection. When the "stock" extract was diluted to construct an ad-
                                                 33

-------
sorptfon isotherm, the adsorption behavior of lead could not be described using this soiksolution ratio
(1:20) because most of the equilibrium concentrations of lead would be below analytical detection limits
Thus a 1:20 ratio was selected to construct a zinc adsorption isotherm, and a 1:100 ratio appeared to be
useful for deriving Pb and Ba adsorption data.

Barium was adsorbed by Cecil clay loam but not to a significant extent (table 9). Since a 1:1 ratio did not
result in at least 10% adsorption, no additional experiments were done. A1:20 ratio was selected for lead
adsorption by Cecil clay loam (table 9), although any ratio between 1:20 and 1:60 would probably have
been acceptable.

A1:4 soil:solution was chosen to study zinc adsorption by Cecil clay loam (table 9), although any ratio be-
tween 1:3 and about 1:8 could have been used. In some cases, there  is a range of suitable soil-solution
ratios for a given soil, but even this range of values must be found experimentally. However, as discussed
in chapter 11, there are guidelines for selecting ratios within the acceptable range. Thus, three different
soil:so!ution ratios (1:4,1:20,1:100) were used to construct barium, lead, and zinc adsorption isotherms
with the two soil samples (results shown in appendix B).
         Table 9 Soiksolution ratios for the Sangamon Paleosol and the Cecil clay loam
         sample using an extract of Sandoval zinc slurry
                  Sangamon soil
Cecil clay loam
Solution
cone (mg/L)
Barium
0.84
1.15
1.80
2.00
2.19
2.25
2.26
Lead
0.15
0.55
2.64
4.70
8.18
11.4
14.6
Zinc
269
365
485
494
532
542
541
%Adsorbed

62.9
49.1
20.3
11.5
3.1
0.4


99.0
96.2
81.2
67.8
44.0
21.9


50.3
32.5
10.4
8.9
1.7
0

Soil:solution
ratio

1:10
1:20
1:60
1:100*
1:200
1:500
blank

1:10
1:20
1:60
1:100*
1:200
1:500
blank

1:10
1:20*
1:60
1:100
1:200
1:500
blank
Solution
cone (mg/L)


2.09
2.19
2.24
2.24
2.24
2.27

4.51
6.98
10.8
11.6
12.7
13.0
14.7

262
365
444
486
515
552

%Adsorbed


8.7
4.4
5.0
0.4
0.4


69.3
52.5
26.5
21.1
13.6
11.6


53.5
35.3
21.3
10.0
4.6


Soiksolution
ratio


1:1
1:2
1:4
1:10
1:20
blank

1:10
1:20*
1:60
1:100
1:200
1:500
blank

1:1
1:2
1:4*
1:10
1:20
blank

        * Soiksolution ratio selected for the kinetic experiments and the adsorption isotherms.
                                              34

-------
                                        CHAPTER 10

              SELECTION OF A SOIL:SOLUTION RATIO FOR NONIONIC SOLUTES

Finding a suitable soilisolution ratio for ionic and polar solutes requires laboratory work, but a simple cal-
culation can be used to estimate a suitable ratio for nonionic solutes, particularly hydrophobic organic spe-
cies. This estimation technique requires a value for the organic carbon content of the adsorbent and for
the organic carbon partition coefficient (K^) of the solute (see ASTM, 1988).

A derivation of this estimation technique begins with
  |igssolute/g soil    _ |j.gs/g
u.gwsolute/g solution  ~ u.gw/g
                                                                                             [4]
where Kd is equivalent to the Freundlich constant (Kf) (see chapter 14) in the special case where the iso-
therm is linear (i.e., 1/n is unity), ugs/g is the mass of solute adsorbed per gram of the adsorbent, and
u.gw/g is the mass of solute per gram of solution.

Also, let R = g adsorbent/g aqueous solution. If we assume that the weight of the solution is approxi-
mately equal to its volume (i.e., 1 mL «1 g), then R is the soihsolution ratio. Equation 4 becomes
                                                                                              [5]
 Since u.gs + u.gw should equal the total mass of solute initially added (u,g°), assuming that losses from vola-
 tilization or microbial degradation are negligible, then
                                    K  _
                                      "
                                or
                                     R =
                                              MOs
                                                           [6]



                                                           [7]
 Thus, it is possible to select an appropriate soil:solution ratio (R) based on an estimate of the Kd value of
 the specific solute-adsorbent system. An estimate of Kd can be calculated if the organic carbon content
 (OC) of the adsorbent and the K^0 of the solute are known by

                                                                                              [8]

 The organic carbon partition coefficients (/^c) of many hydrophobic and other organic solutes have been
 compiled and are given elsewhere (Kenaga, 1980; Kenaga and Goring, 1980; Banerjee et al.,  1980;
 Hassett et al., 1983; Griffin and Roy, 1985; Roy and Griffin, 1985; see Nirmalakhandan and Speece,
 1988, for solubility data). Many of the Koc values reported were based on empirical equations that relate
 the solubility (S) of the solute in water to its organic carbon partition coefficient (K^), such as the expres-
 sion given by Hassett et al. (1983),

                               log /(oc = 3.95 - 0.62 log S (mg/L)                               [9]

 A similar linear relationship has been observed between the octanol-water partition coefficient (/Cow) and
 its organic carbon partition coefficient (see Hassett et al., 1983).
                                                35

-------
                                      log Kbc = 0.088 + log KOVI
                                                       [10]
 A compilation of octanol-water partition coefficients was published by Leo et al. (1971). The historical evo-
 lution of these concepts was discussed by Griffin and Roy (1985).

 To illustrate the application of this estimation technique, the adsorption behavior of a ternary-solute mix-
 ture containing dichloroethane, tetrachloroethylene, and o-xylene by a Catlin silt loam sample was stud-
 ied. To construct adsorption isotherms, suitable soihsolution ratios for each solute had to be determined.
 The organic carbon content of this soil sample was 4.04%. An estimate of a Rvalue for each solute was
 based on its water solubility using equation 9.

 The solubility of dichloroethane and tetrachloroethylene is 8,450 and 200 mg/L, respectively, at 25°C
 (Chiou et al., 1979), and the solubility of o-xylene is approximately 175 mg/L at 25°C (McAuliffe, 1966).
 From equations 8 and 9, the calculated /^values of dichloroethane, tetrachloroethylene, and o-xylene
 were approximately 1.3,13.4, and  14.7, respectively. Recall from chapter 9, a soihsolution ratio corre-
 sponding to about 10% to 30% adsorption is a useful criterion for selecting a suitable ratio. Thus, assum-
 ing that a 20% adsorption will fall into the "target zone" for each of the organic solutes, set u.gs/u,g° equal
 to 20. Then the soihsolution ratio for each solute may be calculated by arbitrarily setting ug° equal to 100.
 For example,
                          tetrachloroethylene:  R =
                                                          20
                              1
                                                    (100-20)13.4   53.6
                                  o-oxylene:   R =
                                                          20
                              1
                                                    (100-20)14.7    58.8
 These awkward soihsolution ratios can be simplified to 1:50 and 1:60. If possible (see chapter 9), use a
 single soihsolution ratio to generate an adsorption isotherm for every solute of interest in a multicompo-
 nent mixture. In this example, a single ratio (1:50) was suitable for generating adsorption isotherms for
 each solute (figs. 18 and 19). There may be a range of suitable ratios for some organic solutes. Consult
 chapter 11 for guidelines on selecting suitable ratios.
•5 18
           0.4      0.8      1.2     1.6
              Solution concentration (mg/L)
2.0
Figure 18  Adsorption isotherm of o-xylene by the
Catlin soil at 23'C and at pH 6.1.
                  0         0.02       0.04
                         Solution concentration (mg/L)

            Figure 19  Adsorption isotherms of dichloroethane
            and tetrachloroethylene by the Catlin soil at 23°C and
            at pH 6.1.
                                               36

-------
This estimation technique can be generalized and shown as a relationship between the linear Freundlich
constant (Kd) and the soil:solution ratio (/?) as a function of different amounts of adsorption on a percent-
age basis (figs. 20 and 21). McCall et al. (1981) demonstrated that equation 5 could be rearranged as
                                                                                              [11]
This equation was used to generate figures 20 and 21, which should serve as convenient guides for se-
lecting soihsolution ratios. For example, the solubility of carbon tetrachloride in water is 800 mg/L at
25°C. The K"oc value, estimated using equation 9, was 140. Using a Catlin silt loam sample with an or-
ganic carbon content of 4.04%, a Rvalue was calculated as
                                            140(4.04)
                                               100
[12]
                  1:180-
                                                  60       80      100      120       140
              Figure 20  Relationship between the linear Freundlich constant (Kd) and soil solution
              ratio as a function of percent adsorption (lower range).
                                                 37

-------
From figure 20, a soilrsolution ratio of about 1:10 should yield approximately 30% adsorption.

The organic carbon content of soils and sediments may control the extent of adsorption of nonpolar
organic solutes provided that the organic carbon content is greater than approximately 0.1% (Schwarzen-
bach and Westall, 1981). At lower organic carbon contents, the extent of adsorption may be underesti-
mated by the empirical regressions, equations 9 and 10, because they do not account for the adsorption
properties of mineral surfaces (see MacKay et al., 1986).
                  1:900-
                      0        200      400      600      800      1000      1200
                                      Linear Freundlich constant (Kd)

           Figure 21   Relationship between the linear Freundlich constant (Kd) and soihsolution
           -.itio as a function of percent adsorption (upper range).
                                              38

-------
                                        CHAPTER 11

                           EFFECTS OF THE SOIL:SOLUTION RATIO

The soihsolution ratio may be one of the most important experimental variables to consider when con-
structing an adsorption isotherm and evaluating the adsorption data, particularly when comparing results
from different investigators who used different ratios. As shown in figure 18, increasing the amount of ad-
sorbent while holding the volume of solution constant had the effect of increasing the mass as well as sur-
face area on which the arsenate ions could be adsorbed.  Hence, logic suggests that as the amount of
adsorbent is increased, the amount of arsenic left in solution after exposure should decrease in an essen-
tially uniform manner as shown in figure 16.

Figure 17 demonstrated a nonlinear response; the amount of cadmium left in solution after 24 hours ap-
peared to be approaching a constant value as the amount of adsorbent was increased (i.e., the soihsolu-
tion ratio was decreased). There is no single explanation  for all systems for this nonlinear response, or
what White (1966) called the "soihsolution ratio effect." The soihsolution ratio effect does not negate the
selection of a soihsolution ratio, but the consequences of that selection must be considered. The influence
of this phenomenon on phosphate adsorption has received a great deal of attention, although the reports
on the effects conflict (Barrow and Shaw, 1979). Phosphate adsorption was increased by the use of high
soihsolution ratios in the studies of Fordham (1963), Barrow et al.  (1965), and White (1966). Hope and Sy-
ers (1976) found that high ratios resulted in lower phosphate adsorption. An early paper by Kurtz et al.
(1946) found no soihsolution ratio effect (i.e., a linear response) when studying phosphate adsorption by
Illinois soils.

White (1966) attempted to reconcile his results by arguing that the system was  not at equilibrium. How-
ever, this  line of reasoning contradicted his rationale for selecting  an equilibration interval. Larsen and
Widdowson (1964) had concluded 2 years earlier that the soihsolution ratio effect was due to an increase
in microbial activity as the mass of the soil was increased.

Hope and Syers (1976) argued that different soihsolution ratios affected only the rate at which phosphate
was removed from solution. They found that the change  in solution phosphate concentration was propor-
tional to the reciprocal of time. Thus, when the reciprocal-time scale was extrapolated to zero, i.e., infinite
time, the effects of different soihsolution ratios disappeared; the isotherms merged into a single point.
 They concluded from this analysis that about 2 to 3 months of equilibration would be necessary for soihso-
 lution effects to essentially disappear. Adsorption data would then be essentially independent of the
 soihsolution ratio, approaching the expected linear response.

 This hypothesis was challenged by Barrow and Shaw (1979), who found that the reciprocal-time analysis
 did not explain the soihsolution ratio effects observed in their study. They concluded that such effects
 were related to particle breakdown during shaking. As more soil was used (i.e., as the ratio decreased),
 more particles broke down, exposing new adsorption sites available to phosphate. However, this concept
 does not explain the results shown in figure 17.

 The soihsolution ratio effect often has been attributed to the competitive interactions between a given sol-
 ute and species that are concomitantly desorbed or exchanged during the partitioning of solutes and ad-
 sorbates. As the amount of adsorbent is increased, a larger source of these potentially competing
 constituents becomes available. The net effect is that the magnitude of adsorption decreases (given equal
 initial concentrations). For example, Griffin and Au (1977) found that the adsorption of lead progressively
 decreased as the sample size of a calcium-saturated montmorillonite was increased. As the amount of ad-
 sorbent was increased, the amount of calcium that was  desorbed or exchanged from the clay also in-
 creased and competed with lead for adsorption sites.

 A similar phenomenon was observed in this study. Using CdCIa,  we investigated the adsorption charac-
 teristics of a Sangamon Paleosol. A pronounced soihsolution ratio effect on cadmium adsorption occurred
                                                 39

-------
                                                                              1:200
                              20
                                     40     60     80     100     120
                                     Equilibrium cadmium concentration (mg/L)
140
       160
                 Figure 22  Effect of soihsolution ratio on cadmium adsorption by a Sangamon
                 Paleosol sample at pH 6.1 and at 22°C. The solid dots were derived by usina a
                 1:100 ratio (Roy et al., 1984).


 (fig. 22). The curvilinear distribution of data points was derived by using a 1:100 soihsolution ratio How-
 ever, when different soihsolution ratios were used, the resulting data did not follow the same pattern but
 felkin a nearly straight line that intersected the adsorption curve obtained where 1:100 ratios were used
 Ca  andMg   probably were exchanging with cadmium and thus reducing cadmium adsorption Hence
 the greater the amount of sample, the larger the amount of Ca2+ and Mg2+ capable of competing with cad-
 mium. At any given equilibrium concentration of cadmium, higher soihsolution ratios (i.e., less absorbent
 per volume of liquid) were associated with increased cadmium adsorption.

 The Sangamon sample contained about 50% expandable clays and 40% illite (appendix A) Work by
 Bittel and Miller (1974) indicated that selectivity coefficients for Ca2+ and Cd2+ exchange reactions with
 rnontmorillonite, iilite, and kaolin'rte were between 0.8 and 1.3 (on a concentration basis), suggesting that
 these clay minerals have no strong affinity for one cation versus the other over a pH range of approxi-
 mately 5 to 7 (cf. Bolt and Bruggenwert, 1978). Calcium will readily exchange with cadmium and vice
 versa. If the adsorption data are plotted as cadmium absorbed relative to Cd2+/(Ca2+ + Mg2+) on a molar
 basis (fig. 23), the different soihsolution ratios coalesce into one adsorption curve.

The soihsolution ratio can also influence the chemical composition of the system, which in turn can di-
rectly or indirectly affect adsorption data. It is a well-established practice to generate aqueous extracts of
soil samples to make qualitative assessments for soil management. Reitemeir (1945) reviewed the litera-
ture on effects of dilution on ionic concentration in soil solutions and attempted to generalize the results:

   Nonsaline soils:
   •  Solution potassium increases with dilution,
   •  Calcium and magnesium in solution frequently increase with dilution, while the ratio of
       Ca:Mg changes, and
   •  Phosphorus usually increases proportionally to dilution.
                                               40

-------
               6-
               5-
               4-
               2-
               1 -
                                                                                   1 MOO
                                                          •
                                                         1:60
                                          1:40
                              •    •
                             *^1:20
                         •1:10
                      • 1:4
                       2.0    4.0     6.0    8.0    10.0    12.0   14.0    16.0

                             Ratio of equilibrium molar concentrations of [Cd] / [Ca + Wig]
                                                                          18.0
                                                                                 20.0
           Figure 23  Cadmium adsorption by a Sangamon Paleosol sample. The adsorption curve
           shown is a transformation of figure 22, taking competitive interactions of Ca2+ and Mg2+
           into account (Roy et al., 1984).


    Alkali, calcareous, and gypsiferous soils:
    • In virtually all cases, dilution results in increased amounts of Ca, Mg, Na, K, SO4, P, and Si.

Thus the ionic concentrations in soil solutions and soil extracts are not inversely proportional to the
amount of water present.

The pH of the soil-liquid suspension will also be affected by the soihsolution ratio. The relationship be-
tween pH and adsorption is discussed in chapter 5. The pH of a soil suspension in a batch-adsorption
procedure is controlled by three factors:

    • the "natural" pH of the adsorbent and its buffering capacity to maintain that pH,
    • the pH and composition of the liquid phase, and
    • adsorption reactions that directly or indirectly change the HsO* and/or OHT
      concentration in solution.

The first two factors are illustrated by figures 24 and 25. The equilibrium pH of solutions mixed with eight
soil materials are plotted against the soil:soiution ratio. In figure 24, the soil materials were exposed to a
sodium arsenate solution containing 200 mg/L As with an initial pH of 4.65. Consequently, at progres-
sively higher ratios (i.e., more dilute systems), the pH of the solutions became progressively closer to that
of the arsenate solution. Thus, at ratios of approximately 1:20 or higher, the pH of the arsenate solution
dominated the pH of the suspensions. At lower soihsolution ratios, the equilibrium pH of each solution be-
came more like that of the soil, the relative strength of this tendency depending on the pH-buffering capac-
ity of the soil.

In the second example (fig. 25), the soil materials were exposed to a cadmium chloride solution contain-
ing 200 mg/L Cd with an initial pH of 5.45. A1:20 ratio for a kaolinite clay sample was associated with a
solution pH of 7.05; a 1:4  ratio resulted in a solution pH of 7.45. Thus an isotherm generated with a 1:4 ra-
tio may yield lower amounts of cadmium adsorption than one using a 1:20 ratio simply because the pH of
the former was more basic for reasons discussed in chapter 5.
                                                41

-------
                                                                              !u nattered)
                                                	VandaliaTill (ablation)



                                                      — Solute solution tends to dominate pH of mixture



                                                             Tifton foamy sand
                 4-
                                    Soil tends to dominate pH of mixture	
                    1:100 1:40   1:20
1:10
Soil solution ratio (mass/volume)
—I—
 1:5
                                                                                      1:4
                                        ^t* °f ar1enate SOlutions (earning the same initial
                                         24 hours of contact with different soil materials as a
 1m r£ nftho "   P,    y J °bSfrVed Wlth complex ^''component extracts or leachates. The equilib-
  r2^ ?n 1th       7 extracVappendix B>was P'otted ^ainst the soil:solution ratio using two so s
 ££? u       TS> u°Wer soil:solution ratios tended to be associated with pHs lower than that of the
 extract. However, the pH tended to be constant when a 1:10 or smaller ratio was used

 The soilrsolution ratio will often influence the ionic strength of the solution. This is to be expected  ThP
 ionic strennth nf am/ end if inn io rontrnllpH hu tho ™n^,Tt.",*-~«   -i  u      1 ^  .      expeciea. l ne
                            is controlled by the concentration and charge of both the solute(s) under
                                  and other aqueous ions derived from the dissolution of soluble miner-
                          	er|t- The ionic strength of the solutions in contact with the Tifton
                   Cecil clay loam tended to decrease as the soihsolution ratio decreased (fia 27)  This
                   to two factors: (1) as the ratio decreased, more arsenic or cadmium was removed
 : , u7	Jl W !?  '°wered tne ionic strength, and (2) these two soils contained a low content of watPr
 ffta 27?w°pTr± that,contributked to the ioni« strength upon dissolution. The other t^solatSs
 fig. 27) were shghtly calcareous by comparison, and consequently lower ratios resulted in an increase in
 onic strength  because of the dissolution of slightly soluble minerals. Discernible decreases^S5?to
 removal of cadmium were masked by the dissolution of carbonates. Whether these changes or d^er

wSX^
                                                42

-------
       7-
       6-
                                                            Kaolinite
                                                 — Solute solution tends to dominate pH of mixture
    pH -j- pH of Cd solution
       5-
       4-
                                                                          Tifton loamy sand
                                                                          Cecil clay loam
                         Soil tends to dominate pH of mixture	
i
o.
1—1	1	1—
1:100  1:40    1:20
                                         1:10
                                         Soil isolation ratio (mass/volume)
                                                                         —I—
                                                                          1:5
                                                    1:4
       Figure 25  Distribution of pH values of cadmium solutions (containing the same initial
       cadmium concentration)  after 24 hours of contact with different soil materials as a
       function of soihsolution ratio.
         . pH of zinc slurry extract
                                       > Sangamon Paleosol
    4-
               1	r
       1:100  1:40    1:20
                                 T
1:10                             1:5
   Soihsolution ratio (mass/volume)
                                                                             1:4     1:2
        Figure 26  Distribution of pH values of solutions of the zinc slurry extract after 24
        hours of contact with two soil samples as a function of soiksolution ratio.
                                                  43

-------
                                                                                            s
                                                                                      -na  w
                                     1:10
                                   Soil :solution ratio (mass/volume)
        Figure 27  Distribution of the ionic strength of solution containing either arsenate or cadmium after
        24 hours of contact as a function of soil:solution ratio.

the relationship between ionic strength, soihsolution ratio, and adsorption for any soil-solute(s) system
may be a major project in its own right.

The adsorption of organic solutes can also be influenced by the soil:solution ratio used in batch proce-
dures. Grover and Hance (1970) found that the Freundlich constant (K» decreased significantly by a fac-
tor of 2.6 as the soil:solution ratio was decreased from 1:10 to 1:0.25 in a study concerned with linuron
and atrazine adsorption. They suggested that the differences in adsorption were related to the aggregate
size of the soil. In a comparison of the relative soil particle  sizes at three soihsolution ratios, they placed
into flasks 10 g of soil that had been passed through a no.  10 mesh sieve. Added to the flasks were 2.5,
10, and 100 ml of a 0.1 M CaCI2 solution. The solutions were mixed by shaking the flasks gently end
over end for 30 seconds, and the solutions were then allowed to stand. Grover and Hance found that the
dispersion of soil aggregates was greater at the 1:10 soihsolution ratio than at the 1:0.25 ratio; the 1:1 ra-
tio was intermediate. A similar sedimentation behavior was observed in the absence of 0.1 M CaCI2.
Thus, they concluded that the extent of adsorption of linuron and atrazine is related to the aggregate size
of the soil.
                                                44


-------
Voice et al. (1983) reported that the solids concentration seemed to significantly affect the adsorption of
several hydrophobic pollutants by Lake Michigan sediments. They concluded that the soil solution ratio ef-
fect in this case appeared to result from the presence of soluble microparticles derived from the soil,
which also tended to retain the solutes (see Voice and Weber, 1985). They concluded that soihsolution
effects reported in the literature may have resulted from incomplete phase separation during centrifuga-
tion or from accumulative relative errors in measuring concentrations.

Similar conclusions were also reached by Gschwend and Wu (1985). If precautions were taken to elimi-
nate or account for nonsettling (or nonfilterable) microparticles or organic macromolecules, which re-
mained in the aqueous phase during batch-adsorption procedures, the observed partition coefficients (Kf
or K"oc) were found to remain constant over a wide range of soihsolution ratios. A succession of prewash-
ing treatments of sediments greatly reduced the effects of the nonsettling particles (fig. 28). When pre-
washed sediments were used for batch equilibration experiments, the observed Kf remained
virtually constant over the range of soihsolution ratios tested. This relationship was most dramatically
shown for the partitioning of the hydrophobic compound, 2,3,4,5,6,2',5'-heptachlorobiphenyl, and the dif-
ference in Kf with and without prewashing clearly reflected the great sensitivity of very strongly adsorbed
compounds to small nonsettling particle concentrations in the aqueous phase.

Voice and Weber (1985) concluded that although soluble microparticles could play the major role in the
soihsolution ratio effect with organic solutes, microparticles could not account for all of the data given in
the literature. Voice and Weber hypothesized that the soihsolution effect was the result of a "complexation
phenomenon," whereby organic matter in the solution phase forms complexes with the solute. The  solute
can exist as a complexed and uncomplexed state in solution, and possibly in other solution states.
      10°-
   I
   I
       10'
        ,3-
                                                    800-1
                                                    700-
                 2, 4, 5, 2', 5'-Pentachlorobiphenyl
200-
                                                    100-
                2, 3, 4, 5, 6, 2', 5'-Heptachlorobiphenyl
                            T
                                         T
                                                                              Catlin
                         EPA-14
                                                                            Cecil clay
                                                                       Sangamon Paleosol
                                                              T
                                                                            T
             1:10,000        1:1,000        1:100

                  Soihsolution ratio (mass/volume)
                              —I	
                               1:30
         1:120   1:60     1:40    1:30    1:24

               Soihsolution ratio (mass/volume)
                                                                                                 1:20
 Figure 28  Freundlich constant (Kf) for two PCB isomers   Figure 29  Freundlich constant (Kf) for the adsorption of
 versus sediment concentration with (open symbols) and   Aroclor 1242 by four different soils at 23°C as a function
 without (closed symbols) prewashing to remove nonset-   of soil solution ratio.
 tling particles (adapted from Gschwend and Wu, 1985).
                                                 45

-------
In other organic solute-adsorbent systems, the adsorption behavior of the solute was not influenced by
the soil:solution ratio. Bowman and Sans (1985) reported that the adsorbent concentration (soil solution
ratio) did not appear to affect significantly the partitioning of several pesticides in sediment-water systems
over a fairly wide range of values.

The adsorption of Aroclor 1242 was not influenced by the soihsolution ratio (fig. 29). The Freundlich con-
stant was essentially constant over a wide range of soil:solution ratios. When different soihsolution ratios
were used in the construction of adsorption isotherms, the resulting data tended to plot on the same line
(figs. 30 and 31), and the slopes of the adsorption isotherms were near unity. In some cases, a curvilinear
distribution of data points was derived by using different soil:solution ratios with some adsorbents (fig. 32).
          45-
          40
                                                    1:500
                      0.02
                                 0.04        0.06        0.08        0.10

                                   Equilibrium Aroclor 1242 concentration (mg/L)
0.12
           0.14
       Figure 30  Aroclor 1242 adsorption isotherms by 5 soils at 23°C using various soil:solution ratios.
                                                46
                                                                                                            ^

-------
However, the application of different soihsolution ratios still yielded a single, consistent relationship be-
tween the amount of Aroclor 1242 in solution and the amount retained by the tills at equilibrium (fig. 32).

In summary, the selection of a soihsolution ratio may or may not have a profound effect on adsorption
data. The soihsolution ratio may influence the pH, ionic strength, and chemical composition of the
suspension, which in turn may influence adsorption data. In some cases, such as competitive interac-
tions, the soihsolution ratio effect can be rationalized, but in other systems, the ratio effect presents prob-
lems, particularly for procedures intended for the routine collection of batch-adsorption data. Voice et al.
(1983) commented that some combination of techniques or new methodologies may evolve to handle the
ratio effect, but no simple solutions are readily apparent.
                                                            1:250
                        0.02
                      ~T
0.04        0.06         0.08        0.10
       Equilibrium concentration (mg/U
                                                                              0.12
         Figure 31   Adsorption of dieldrin, tetrachloroethylene, and 1,2-dichloroethane by the Catlin soil
         at 23°C using various soil solution ratios.
                                                47

-------
Recommendations The specific soilrsolution ratios provided in chapter 17, section 8.3, should be
adopted as standard ratios for the construction of adsorption isotherms. For example, if a 1:8 ratio is satis-
factory for the generation of adsorption data, the investigator should use a 1:10 ratio, which is one of the
standard ratios. For many systems, there will be a range of suitable ratios. The user should not arbitrarily
select any ratio within this range but should select the closest standard ratio. These standard ratios range
from 1:4 to 1:10,000 and should accommodate most situations. Adherence to this recommendation will
enable direct comparisons of adsorption data generated by different investigators. Adsorption data based
on ad hoc ratios may provide a basis for limited comparison; however, unless a particular solute-adsorb-
ent system can be shown not to be subject to soil solution ratio effects, there will always be some doubt
that the results are comparable.
                25-
                20-
              215-
               8
               Ti
               Z
               o  10-
                 5-
           1:250


Vandalia till (altered)
                                                        1:100
                        1:6
                             0.02
                                        0.04        0.06       0.08        0.10
                                     Equilibrium Aroclor 1242 concentration (mg/L)
                                                                                   0.12
         Figure 32 Adsorption of Aroclor 1242 by altered Vandalia till and unaltered Vandalia till at 23°C
         using various soihsolution ratios.
                                                 48

-------
                                         CHAPTER 12

                      CONSTANT AND VARIABLE SOIL:SOLUTION RATIOS

Two experimental techniques are used to generate batch-adsorption data:

• Constant soil:solution ratio method: mixing a batch of aqueous solutions—each solution containing pro-
gressively decreasing solute concentrations—with absorbent, keeping the amount of adsorbent (by
weight) constant in all solutions.

• Variable soihsolution ratio method: mixing a batch of solutions, all containing the same solute concentra-
tion, with progressively increasing amounts of absorbent.

The first technique presumably is based on a standard ratio selected from the procedures given in chap-
ters 9,10, and 17. The second is similar to the technique used to select a soihsolution ratio for ionic sol-
utes (chapter 9). Although one might expect that both techniques would yield the same results and thus
could be used interchangeably, this does not hold true for all systems.

In the first technique, the initial (stock) solute solution—a solution prepared in the laboratory or a leachate
taken from the field—is progressively diluted, forming a batch of diluted solutions that are added to con-
tainers holding the same amounts of soil material. As discussed in chapter 11, the soihsolution ratio may
affect the adsorption data. Figure 22 showed that using soihsolution ratios ranging from 1:200 to 1:4
yielded adsorption data that were in poor agreement with the isotherm generated from a 1:100 ratio. In
this case, the phenomenon was attributed to competitive interactions between Cd2+ and desorbed Ca2+
and Mg2+. Where the data were replotted (fig. 23) with these competitive interactions taken into account,
the adsorption data coalesced into a single consistent relationship. This replotting technique does not
work in all cases. Techniques for modeling competitive adsorption are emerging (see Murali and Aylmore,
1983a,b,c; Roy et al., 1989) and are currently too complicated to use in routine batch procedures. Also,
not all "soihsolution ratio effects" can be attributed to competition (see chapter 11). This effect was ob-
served during the development of these procedures; applying the variable soihsolution ratio technique
yielded results (amounts of cadmium and lead adsorbed) similar to or lower than the amounts obtained by
the constant soihsolution procedure.

For this reason, an isotherm produced by variable soihsolution ratios was considered the more environ-
mentally conservative isotherm. In this document an isotherm produced in this way is called an environ-
mentally conservative isotherm (ECI). The ECI  has two major advantages over an isotherm produced by
the use of a fixed soihsolution ratio: (1) if the solute-adsorbent system reaches equilibrium in 24 hours—
or, more correctly, satisfies the conditions of the operational definition of equilibrium (chapter 13)—then
the data generated in selecting a soihsolution ratio can be used to construct an isotherm; and (2) the
effects of competition and other processes are implicitly accounted for, although their exact nature is
unknown.

Figure 33 provides further evidence that using a variable soihsolution ratio yields environmentally conser-
vative estimates of adsorption. The adsorption data were modeled with the Freundlich equation (chapter
14), yielding the isotherm constants shown. The isotherms associated with the constant soihsolution ratio
technique are called constant soihsolution ratio isotherms (CSI). The results shown in figure 33 may be
generalized as
                              and
                                         1 //?ECI ^
This type of analysis indicates that an isotherm generated by using different adsorbent masses generally
yields lower predictions of solute adsorption and is thus viewed as being environmentally conservative.
The ECI is recommended as the method of choice for routine use.
                                              49

-------
                 1   1-
                 o
                 UI
                        0 0.02
                                              Kf (CSI), L/mg
                 B
                     1.0-1
                     0.8-
                 O   0.6-
                 ui
                     0.4-
                     0.2-
                     0.0-
• Cadmium

O Arsenic

A Lead

A PCS
                       0.0
                              0.2
                                     0.4

                                     1/n(CSI)
                                            0.6
                                                   0.8
                                                         1.0
                  Figure 33  Distribution of (A) Freundlich constants (Kf) and (B) exponents
                  (1/n) associated with arsenic, cadmium, lead, and PCS (Aroclor 1242)
                  adsorption isotherms.

Experimental data produced during the process of selecting a soihsolution ratio may be used to construct
an isotherm if the system equilibrated within 24 hours (equilibration time is discussed in the next chapter).
However, some of the data points inevitably will be associated with situations in which less than 10% of
the solute was adsorbed. In chapter 9, we recommended that a soihsolution ratio be chosen from which at
least 10% or greater adsorption occurred. To understand the importance of this recommendation, con-
sider a situation in which an investigator conducted experiments for selecting a soihsolution ratio, then
used data points from the entire concentration range and attempted to construct adsorption isotherms
(figs. 34 and 35). As shown in figures 34 and 35, the data points associated with less than 10% adsorp-
tion did not conform to the general pattern established by data associated with greater than 10% adsorp-
tion. Eliminating these data points yielded better results, i.e., more reasonable revalues (chapter 14).
Rgure 35 shows an extreme case; nearly all data were associated with less than 10% adsorption. There
was Itttle point in fitting this data set with an isotherm equation.

Although the ECI  is useful for many  situations, it cannot be universally applied to every situation. The ECI
may be limited to cases in which (1) the adsorbent has a relatively high affinity for the solute, and (2) the
initial solute concentration is relatively low. The ECI technique is often used with sparingly-soluble organic
solutes in which the initial solute concentration is low.

The ECI technique was used to derive arsenic adsorption isotherms with the soil adsorbents used to de-
velop these procedures. Soihsolution ratios of 1:4  and higher were used, and the initial concentration of ar-
senic was 200 mg/L. Varying the amount of adsorbent over this range of soihsolution ratios did not
                                               50

-------
     0.0
                    0.4
                1	T
0.8             1.2             1.6
Equilibrium cadmium concentration (mg/L)
 r
2.0
1
2.4
Figure 34   Cadmium adsorption isotherm at 22°C for a Vandalia till sample (unaltered) with the amount
adsorbed associated with each isotherm data point shown. The mean pH of the soil-solute suspensions
was 6.8.
240-


200-


-a 160-
1
T3
J3
I 120-
(O
c
=1
<
80-


40-
0
0

5.9%
2.7
\_ 10.2%
x^^f *
24.6% adsorbed ^"^
X
X
X
r2 = 0.04
X
/
'
,
,
' c
/ £
/ I
/ s
/ 8
' 1
1 i II
40 80 120 160


0
.7%
••^



0.32%







I
200
                                 Equilibrium cadmium concentration (mg/L)

                 Figure  35  Distribution of cadmium adsorption data at 22°C for a
                 Tifton sandy loam with the  amount adsorbed associated with each
                 isotherm data point shown. The mean pH of the soil-solute suspensions
                 was 6.8.
                                                51

-------
change the equilibrium arsenic concentration substantially (fig. 36). The relatively small changes in arse-
nic equilibrium concentrations caused the data points to be somewhat clustered, leaving an area between
the on'gin of the isotherm and the lowermost arsenic equilibrium concentration without data points. There-
fore, regression of these data sets using isotherm equations could potentially lead to large errors. More-
over, lower soil:solution ratios cannot be used to fill in the gaps; the use of ratios much lower than 1:4
would eventually produce a very thick suspension or paste that could not be efficiently mixed, separated,
or analyzed. This "ratio gap" problem is accentuated as the initial solute concentration increases; the
"cluster" simply migrates to the right side of the isotherm. For these reasons, the constant soihsolution ra-
tio isotherm (CSI) is also recommended for application as an alternate procedure, given that the ECI tech-
nique does not produce useful or applicable results, as figure 36 illustrates.
              0.6-
              0.5-
              0.4-
         8
         n
              0.2-
              0.1-
              o.o-
                                                                            1:20 • (13% ads rbed)
                                                                      1:10
                                                                   Catlin silt loam
                                                   1:4
                                                                                  1:18 (11%)
                                                  /     Till (Vandalia ablation)
                                                                             1:10(18%)
                                                                   1:5
                            Inaccessible region
            Kaolinite clay
                                                                         1:5
                                                                   Till (altered Vandalia)
                                                                                   1:10 (11%)
                               40
                                            —T~
                                             80
—1—
 120
—T~
 160
               200
                                       Equilibrium arsenic concentration (mg/L)
           Figure 36  Distribution of arsenate adsorption data at 23°C for different soil samples using
           different soihsolution ratios. The pH values of each soil-solute system were similar to those
           given for each soil in appendix A.
                                                  52

-------
                                         CHAPTER 13

                         DETERMINATION OF THE EQUILIBRATION TIME

 Adsorption at the solid-liquid interface is a thermodynamic process, and adsorption measurements are
 taken when the system has equilibrated. The equilibration time in batch-adsorption experiments is the
 time interval in which the system reaches chemical equilibrium and the concentrations of the products
 and reactants cease to change with respect to time,


                                           ft = °                                           [13]

 Past studies have used many different equilibration times. Lawrence and Tosine (1976) used 30 minutes
 to equilibrate PCBs with soil, whereas Jones et al. (1979) allowed a soil-phosphate mixture to equilibrate
 for 6 days before separating the liquid from the soil. The equilibration times used in most studies were
 probably based on preliminary kinetic studies. However, there is a clear danger in assuming that the
 equilibration time reported by one investigator is valid for another system even when the adsorbent-solute
 systems are similar. Equilibration time is an experimental  variable that must be determined for any sys-
 tem before an adsorption isotherm (curve) is constructed.

 The ASTM procedure D4646 (ASTM, 1987) can be used to characterize the affinity of a soil or clay for sol-
 utes after 24 hours, but 24 hours may or may not be long enough for chemical equilibrium to develop (as
 indicated earlier, some investigators have used equilibration times of days or even weeks)  Adsorption is
 generally regarded as a fast reaction, and subsequent removal of solute from solution may be attributed
 to other processes. Adsorption processes at solid-liquid interfaces are often initially rapid; further reduc-
 tion in solute concentration continues at a decreasing rate, asymptotically approaching a constant concen-
 tration. In some cases, equilibrium is never clearly attained. The ambiguity in the definition  and
 measurement of equilibration times has been acknowledged as a major problem in adsorption studies
 (Anderson et al., 1981). For most systems involving complicated adsorbents such as soils,  it is very diffi-
 cult to determine when adsorption processes dominate and when they become less important as other
 processes, such as ion penetration or precipitation, become significant. The EPA (EPA, 1982) suggested
 that the equilibration time should be the minimum amount  of time needed to establish a rate of change of
 the solute concentration in solution equal to or less than 5% per 24-hour interval. This definition is an op-
 erational definition of equilibrium and is equivalent to a steady state. Cast in a form similar to that of equa-
 tion 13 it may be written as
                               AC
                               — < 0.05 per 24-hour interval
[14]
The efficacy of this operational definition for equilibrium was evaluated using seven soil materials  Each of
the adsorbents was exposed to arsenic and cadmium solutions, initially containing 200 mg/L for periods
of up to 72 hours. The solutions were analyzed, and the rate of removal of the solute was determined All
of these particular soil-solute systems were in equilibrium after 24 hours as defined by this operational
definition. Figure 37 presents representative data for cadmium adsorption, and figure 38 shows the ad-
sorption behavior of arsenate for 11 different soil materials. In this example, it is not obvious in some
cases when the rate of change of the solute concentration is equal to or less than 5% per 24-hour inter-
val. It may be more convenient to analyze kinetic data as shown in table 10: an equilibration time of 24
hours was selected for three of the samples, and a period of 48 hours was used to equilibrate arsenate
with a Vandalia till sample (ablation phase). In each example, the calculated %AC represents the change
in concentration during the preceding 24 hours. During the next 24-hour interval, the arsenic concentra-
tion continued to decrease, but only by 1.14%. It is this slow and relatively small decrease in solute con-
centration that follows the more rapid and pronounced decrease that is frequently a problem. The applica-
tion of this operational definition of equilibrium assumes that this additional 1.14% decrease is negligible
                                              53

-------
idmium adsorbed per sram of soil (mg/g
> ro *> at m
» - i
Kaolinite
L ^-*— -~^_ Sangamon paleosol
1
T _-___*_^j
f " ]
f
Vandalia till (ablation) f
( Vandalia till (altered)
_ Vandalia till (unaltered) _

6 "6 10 20 30 40 50
Time (hours)

60 70
Figure 37  Adsorption of cadmium by 5 soil materials at 22°C as a function of contact time.
      200
      190
                                       30       40
                                        Time (hours)
                                                          50
                                                                    60
                                                                             70
     Figure 38  Adsorption of arsenic by 11 soil materials as a function of contact time.
                                           54

-------
                  Table 10  Equilibration times for adsorption of arsenate by soil materials
Time
(h)

0
1
4
8
16
24f
36
48
72

0
1
4
8
16
24
36
48t
72
%AC *
Cecil
—
—
—
—
—
8.17
—
0.56
1.08
Vandalia
—
—
—
—
—
18.97
—
5.51
3.15
Solution
cone (mg/L)
clay loam
193.4
182.9
181.5
180.0
177.6
177.6
177.1
176.6
174.7
till (ablation)
199.3
179.5
173.2
169.4
164.0
161.5
157.3
152.6
147.8
%AC
Solution
cone (mg/L)
Kaolinite
—
—
—
—
—
16.36
—
1.14
0.55
Sangamon
—
—
—
—
—
33.13
—
2.20
2.25
199.3
171.4
168.6
168.1
166.2
166.7
164.8
164.8
163.9
Paleosol
199.3
165.3
158.2
158.1
152.6
149.7
148.3
146.4
143.1
                  * %AC = (C, - CgJ/C,, where C,  is solution concentration at time *
                  and C2 is the concentration at time t + 24 hours.
                  t Equilibration time selected for the adsorption isotherms.
and may be attributable to processes other than adsorption. Therefore, this solute-adsorbent system is de-
fined as being at steady state after 24 hours of contact.

The application of this definition with multicomponent solutions is exemplified by the metallic waste slurry
(appendix B). To determine the equilibration time, we conducted preliminary kinetic experiments using the
soil:solution ratios previously determined (chapter 9). Barium was adsorbed by the Sangamon Paleosol
sample and this system appeared to reach equilibrium within 24 hours. The rate of change in solute con-
centration for the first 24-hour period was 12.2% (table 11). After 24 hours, the rate of change was less
than 5% for each subsequent 24-hour interval. Similarly, the solution concentrations of lead and zinc were
also constant after 24 hours (fig. 39)—constant in the sense that the rate of change in solution concentra-
tion of these two solutes per 24-hour interval was less than 5% (table 11).

The zinc-Cecil clay system appeared to reach the operational equilibrium within 24 hours (fig 39)  The
rate of change in zinc concentration during the 24- to 48-hour interval was 0.2% (table 11). Lead did not
equilibrate with the Cecil clay until about 48 hours; the rate of change in lead concentration during the 48-
to 72-hour interval was 1.5% (table 11). Thus, an equilibrium interval of 24 hours was used to construct
adsorption isotherms except that for the adsorption of lead by the Cecil clay (a 48-hour interval was used).

Solution concentrations of o-xylene, dichloroethane, and tetrachloroethylene tended to change by
amounts less than about 5% after 24 hours when in contact with Catlin (fig. 40). The rate of adsorption of
the PCB Aroclor 1242 by a Catlin sample was nil after the initial 24 hours of contact (table 12)
                                               55

-------

560-
1
520-
480-
3
J 440-
c
| 400-
e
g 360-
u
c
•2 A
3
s
10-



O-J


in
3
O
JZ











	 I I


w
CO









1 *

' 1 1



|
S

Zn-Cecll clay
Zn-Sangamon




Pb-Sangamon
\ Pb-Cecil clay
"— \
Ba-Sangamon
i — ,-
I I 1













i
I

                           10
                                    20
                                             30
         40

   Time (hours)
                                                               50
                                                                        60
                                                                                 70
              Figure 39  Equilibration times of Ba, Pb, and Zn from a laboratory extract of the
              Sandoval zinc slurry with the Sangamon Paleosol and the Cecil clay sample.
                  500-
                  400-
                  300-
                .0
                s.
                   12-
                    8-
                     4-
                                                            Dichloroethane
                                                              J3-xy)ene
                                 10
20         30
  Time (hours)
                                                                   40
                                                                              50
               Figure 40  Adsorption of o-xylene, dichloroethane, and tetrachloroethylene at
               23°C by Catlin soil as a function of contact time.

Recommendation  The equilibration time should be the minimum time needed to establish a rate of
change of the solute concentration in solution that is equal to or less than 5% for a 24-hour interval. The
equilibration time should be determined for each solute-adsorbent system before adsorption isotherms
are constructed. The typical equilibration time is 24 hours.
                                                  56

-------
 Table 11  Equilibration times for adsorption of Ba,  Pb, and Zn from
 a Sandoval zinc slurry extract by the Sangamon  Paleosol and Cecil clay
Time
(h)
Barium
0
1
8
24*
31
48
72
Lead
0
1
8
24*
31
48
72
Zinc
0
1
8
24*
31
48
72
%AC,

—
—
—
12.2
—
-1.0
-1.5

—
—
—
66.8
—
8.9
5.1

—
—
—
33.4
—
1.1
0.3
Solution
cone (mg/L)

2.30
2.00
2.02
2.02
2.02
2.04
2.07

15.5
5.68
5.01
5.14
4.81
4.68
4.92

563
387
375
375
371
371
370
Time
(h)
Lead
0
1
8
24
31
48*
72
Zinc
0
1
8
24*
31
48
72








Solution
%AC^ cone (mg/L)

— 15.4
— 7.72
— 7.27
54.5 7.01
— 6.72
6.3 6.57
1.5 6.47

— 549
— 421
— 430
20.9 434
— 430
0.2 433
0.2 432








* %AC = (C, - C2)/C,, where  G, is solution concentration at time f,
   and C2 is concentration at time  f + 24 hours.
t Equilibration time selected for the adsorption isotherms.
         Table 12  Equilibration times for adsorption of the PCB
         Aroclor 1242 by Catlin soil
        Time (h)
%AC
Solution cone (mg/L)
0
2
4
6
8
24*
48
_
—
—
—
—
94.31
0.00
0.220
0.020
0.018
0.017
0.017
0.013
0.013
        *Equilibration time selected for an adsorption isotherm.
                                  57

-------

-------
                                        CHAPTER 14

                    CONSTRUCTION OF ADSORPTION ISOTHERMS (CURVES)

 An adsorption isotherm or curve is a graphic representation showing the amount of solute adsorbed by an
 adsorbent as a function of the equilibrium concentration of the solute. This relationship is quantitatively de-
 fined by some type of partition function or adsorption isotherm equation that is statistically applied to the
 adsorption data to generalize the adsorption data.

 In studies concerned with the adsorption of gases by solids, more than 40 equations have been used to
 describe the data. Historically, only a few of the equations have been found to be applicable to solid-liquid
 systems. Only the two most commonly used and simplest of these adsorption equations will be discussed
 here—the Freundlich and Langmuir isotherms. Neither may be appropriate for a given system. The
 reader may  wish to consult a paper by Kinniburgh (1986)  on the applicability of other adsorption equa-
 tions.

 The Freundlich Equation
 Probably the oldest, most widely used adsorption equation for solid-liquid systems is the Freundlich ad-
 sorption equation, named after H. Freundlich (Freundlich, 1909),
                                        x _ isrA/n
                                                                                           f15]

where x is the amount or concentration of the solute adsorbed, m is the mass of the adsorbent, C is the
equilibrium concentration of the solute, and K>and 1/n are constants.

The Freundlich equation was originally proposed as an empirical expression without a theoretical founda-
tion. However, some investigators have referred to the Freundlich constant (Kf) as being related to the ca-
pacity or affinity of the adsorbent; the exponential term may be an indicator of the intensity of adsorption
or how the capacity of the adsorbent varies with the equilibrium solute concentration (see Suffet and
McGuire, 1980).

Other investigators attempted to show that the Freundlich equation has a theoretical basis. A number of
derivations of the Freundlich equation were based on the Gibbs adsorption equation (Chakravarti and
Dhar, 1927; Rideal, 1930; Freundlich, 1930; Halsey and Taylor, 1947; see Hayward and Trapnell, 1964;
Kipling,  1965). Zeldowitsch (1935) demonstrated that the Freundlich equation could be explained in terms
of a nonhomogeneous surface. Sips (1948) established in a rigorous fashion a general relationship be-
tween surface heterogeneity and the Freundlich equation, a derivation Sposito (1980) partially adapted to
his system to derive a Freundlich-type expression for trace-level exchange reactions.

The Freundlich equation is frequently used, probably because it is simple. It contains two constants; both
are positive-value numbers that can be solved statistically when expressed in logarithmic form:
                               log(x/m) = \ogKf+-[/n\og C
[16]
By taking the logarithms of both sides of equation 15, the constants K>and 1/n may be solved, via equa-
tion 16, as a simple linear regression,
                                       y, = a + bx,
[17]
                                              59

-------
where log(x//77)/  »  //
          logK>  »  a
             11 n  -  b
          log C/  »  x/

The technique for solving a linear regression can be found in any introductory statistics textbook and is
also a common feature of most moderately priced electronic calculators. (Note: linear regressions are
sometimes referred to as the line of best fit or method of least squares.) For the sake of completeness,
the constants may be solved (with n* as the number of pairs of data points) using

                                                                 X//77/)
                       1     n*(£log C/x log x/mf) - (£log
                                                                                             [18]
The following example is given to illustrate the application of the Freundlich equation. Previous work
showed that the adsorption of arsenate by kaolinite could be characterized by using a 1:10 soil: solution
ratio (chapter 9) and that the system reached a steady state after 24 hours. Under these experimental
conditions, 17 dilutions of a stock KH2AsO4 solution were mixed with an NBS rotary extractor with kao-
linite for 24 hours. Table 13 contains all the data needed to construct an isotherm and also includes the
            Table 13  Data  reduction for arsenic adsorption at 25°C by a kaolinite clay
            sample (volume  of solution, 200 ml_)
Initial
cone
(mg/L)
4.89
10.0
15.2
19.9
19.9
19.9
29.9
40.3
49.4
80.5
80.5
80.5
98.8
121.0
137.7
160.3
160.3
Equilibrium
cone
(mg/L)
1.20
3.56
6.78
10.1
10.1
10.3
17.6
25.0
33.4
58.4
59.5
58.9
76.3
92.6
109.4
128.3
129.7
Adsorbent
wife)
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
20.42
Amount
adsorbed
O^g/g)
36*
64
84
98
98
96
123
153
160
221
210
216
225
284
283
320
306
PH
8.30
8.26
8.26
8.19
8.23
8.25
8.16
8.03
8.02
7.77
7.80
7.83
7.69
7.56
7.50
7.27
7.26
EC (dS/m)
160
168
170
185
185
185
205
221
240
305
313
305
350
385
413
434
430
             * Sample calculation:
                 _x_ _ (Initial cone. - equil. cone.) x volume of solution
                 m ~              weight of adsorbent
                       (4.89 mg/L-
                                      20 mg/L)x 0.200 L =
                                                                 mg/g =
                                                60

-------
 linite for 24 hours. Table 13 contains all the data needed to construct an isotherm and also includes the
 pH and electrical conductivity (EC) of each solution, determined as recommended at the ends of chapters
 5 and 6.
 In this example (table 13),
and thus
         log Kf = 1.536       1//7 = 0.452



             •%• = 34.328  (As)0'452
                                                                                            [19]
where As is the equilibrium concentration of arsenic in solution (mg/L). The units mg/L are equivalent to
u.g/mL, and therefore the units of Kf are mL1/"u.g(1"1/n)/g~1 from
x/m(u.g/g) =
                                                        As (u.g1/n/mL1/n)
The 1/nterm has no units. The selection of the units for x/m and the equilibrium solute concentration will
determine the units of Kf in a given situation. When Mn < 1 , the units used must be considered when ad-
sorption constants are compared from different sources (see Bowman, 1981 ; Hassett et al., 1983).

Thus, equation 19 becomes a predictive equation capable of describing the adsorption data. The reader
may wish to use the data given in table 13 to verify equation 19. For example, equation 19 should not be
used to predict x/m at equilibrium concentrations greater than 130 mg/L; to do so requires the collection of
data in this higher concentration  range. The validity of this cautionary note becomes apparent when one
considers that the Freundlich equation predicts infinite adsorption at infinite concentrations; hence, any
soil or clay would have an unlimited capacity to retain chemicals dissolved in water. Not only would an infi-
nite capacity be thermodynamically inconsistent, but experience has shown that the extent of adsorption
is ultimately limited by the surface area (or some portion of the surface) of the adsorbent. Thus, there are
two drawbacks in using the Freundlich equation: (1) it cannot be extrapolated with confidence beyond the
experimental range used in its construction, and (2) it will not yield a maximum capacity term, which in
many cases is a convenient single-value number that estimates the maximum amount of adsorption be-
yond which the soil or clay is saturated and no further net adsorption can be expected.

The Langmuir Equation
The Langmuir equation has given rise to a number of Langmuir-type expressions that have been widely
used to describe adsorption data for solid-liquid systems.  The most commonly used expression may be
generalized as
                                             KLMC
                                       m
                                                                                           [20]
where x is the amount or concentration of the solute adsorbed, m is the mass of the adsorbent, C is the
equilibrium concentration of the solute, and KL and Mare constants.

Langmuir (1918) derived an expression similar to equation 20 to describe the adsorption of gases on sol-
ids (flat surfaces of glass, mica, and platinum). He generalized that the Freundlich equation was unable to
describe the adsorption of gases when the range of pressures was large. Langmuir's original derivation
was based on the premise that during the adsorption of gases, a dynamic equilibrium is established in
which the rate of condensation (adsorption) is equal to the rate of evaporation (desorption). Derivations of
the Langmuir and Langmuir-type equations for gas-solid interactions are given elsewhere (Langmuir,
1918; Hayward and Trapnell, 1964; Ponec et al., 1974). Langmuir-type expressions for ion exchange reac-
tions in soils have also been derived (Sposito,  1979; Elprince and Sposito, 1981).
                                              61

-------
The applicability of Langmuir-type equations to solid-liquid systems has been a controversial topic in re-
cent years (see Harter and Baker, 1977; Veith and Sposito, 1977; Barrow, 1978; Sposito, 1982). How-
ever, this controversy is concerned not with the ability of the equation to simply describe the adsorption
data, but with interpretations of adsorption mechanisms and energetics that are based on the results of
applying Langmuir-type expressions.

Some investigators have concluded that the Langmuir constant (KL) is somehow related to the bonding
energy between the adsorbed ion and the adsorbent, but that specific functional relationship is uncertain.
The constant Min equation 20 is generally accepted as the adsorption maximum of the adsorbent with re-
spect to the specific solute and is interpreted as the maximum amount or concentration that an adsorbent
can retain.

Langmuir-type equations are frequently used because of their ease of application. Like the Freundlich
equation, such equations contain only two constants, both of which are positive-value numbers that can
be statistically solved when equation 20 is cast in a  linear form. Two  linearized expressions are possible:
                                        x/m
                                        x/m
    1    .  C
  KLM    M
    1    ,  J_
  KLMC    M
                                                                                            [21]
                                                                                            [22]
The linearized form of equation 21 is sometimes referred to as the "traditional linear Langmuir equation,"
and equation 22 is called the "double-reciprocal Langmuir equation." The latter is more suitable for situ-
ations in which the distribution of equilibrium concentrations tends to be skewed towards the lower end of
the range of the equilibrium concentrations. As indicated above, linearized Langmuir-type expressions
such as equations 21 and 22 are equivalent to a simple linear regression,
                                           y,= a + bx,
 where the traditional linear Langmuir equation is
                                          y\
                                          a
                                          b
 and the double-reciprocal form is
                                          y\
                                          a
                                          b
                                          X,
=  (C/x/m),
=  1/Ki.M
=  1/M
=  C,
   (1/X//77);
   MM
   VKLM
   1/C,
 The techniques for solving either equations 21 or 22 are the same as those used to solve the linear form
 of the Freundlich equation (eq. 16). From the data set given in table 13, application of the linear Langmuir-
 type equations yields:

 Traditional Linear Langmuir
                                     a =
                                          KLM
                                               = 0.0792
                                               [23]
                                               62

-------
 and thus
 Double-Reciprocal Langmuir
                                        = -   = 0.0028
                                   _x _ 3.568 x10"2(353.856)C
                                   m ~   1 + 3.568x10~2(C)
a = -J; =  0.0050
                                                       [24]
                                                       [25]
                                                                                          [26]
                                    b =
        ;•= 0.0297
                                                                                          [27]
and thus
                                   _x  = 0.1702(198.098) C
                                   /n  ~   1+0.1702(0)
In this example, the units for the adsorption maximum are the same as for x/m (ug/g), and the units for K,
are liters per milligram:
                                               M(ng/g) C(mg/L)
                                         i+/C(L/mg)C(mg/L)
                                                                                          [28]
The selection of units for x/m and the equilibrium solute concentration determines the units for M and KL.

Equations 25 and 28 are predictive expressions that can describe the adsorption of arsenic by kaolinite.
The reader should work through these examples to verify the results. In the previous examples, the iso-
therm constants were derived by linear regression. Kinniburgh (1986) recommended that isotherm con-
stants be solved by nonlinear regression (nonlinear least squares) to obtain more accurate values than
those derived by linear regression. A short BASIC program using a nonlinear least-squares method for de-
termining Langmuir constants was written by Persoff and Thomas (1988).
                                            63

-------

-------
                                          CHAPTER 15

                            SELECTION OF ADSORPTION EQUATIONS

 Three isotherm regressions were used to describe the example data set given in table 12. Given the se-
 lection of different models, one equation usually will describe the results with the greatest accuracy No
 clear consensus has been reached on which equation (Freundlich or Langmuir-type) is the most reliable
 for simply fitting data. Barrow (1978) objected to the application of Langmuir-type expressions but his ob-
 jection was based on theoretical considerations. Singh (1984) compared five adsorption equations and
 found that the Freundlich equation was the most accurate in describing the adsorption of SO42' by soils
 Polyzopoulos et al. (1984) compared four adsorption equations in a study concerned with phosphate ad-
 sorption by soil. They found that Langmuir-type or Freundlich expressions described the data with compa-
 rable success.                                                                               K

 Generally the choice of an equation is based on the coefficient of determination (r2) obtained in  a given
 case and the equation's simplicity (Polyzopoulos et al., 1984). The Freundlich and Langmuir equations
 each contain only two constants and are easily solved.

 The coefficient of determination (sometimes called the goodness of fit) is a measure of how closely the re-
 gression line fits the data, and may be calculated using equation 29:

                                       ,    £(y/-y)2

                                            I(y/-y)2                                         [29]
 where y,  is the value of the dependent variable predicted by the regression, y, is the value actually meas-
 ured, and y is the arithmetic mean of all y,. The value of r2 will always be between 0 and 1 inclusive  If all
 of the points are close to the regression line or, in this  example, if all of the adsorption data plot closely to
 the statistically constructed adsorption isotherm, the corresponding r2 will be close to 1  The application
 of equations 16, 21, and 22 to the data set in table 12 yielded dissimilar r2 values:

                         Freundlich                        0.996
                         traditional linear Langmuir          0.954
                         double-reciprocal Langmuir        0.916

 When the coefficient of determination is used as a criterion, the Freundlich equation best described the
 adsorption data, although the traditional linear Langmuir expression would also yield satisfactory  results
 Figure 41 clearly shows that the double-reciprocal linear Langmuir equation did not fit the adsorption data
 well and that the traditional linear form tended to overpredict adsorption in the upper part of the isotherm
 Obviously the high r value associated with the Freundlich equation is reflected by the closeness of fit of
 the isotherm with the data.

 Obtaining a reliable fit of adsorption data with the chosen equation (so that r2 values are close to 1) is a
 major concern in the construction of adsorption isotherms. However, in some cases, a low r2 value will be
 obtained regardless of the equation used, raising concerns that the adsorption constants actually have lit-
tle meaning. Probably the simplest statistical test for such situations  is to use f-statistics to examine
whether the sample correlation coefficient (r) is significantly different from a population correlation coeffi-
cient (p) where p = 0. This test appears in most introductory statistics textbooks and will not be discussed
 here.
                                                65

-------
            320-
            280-
            240-
        _  200-
         "H

         I
            160-
            120-
                                 Traditional linear
                                 Langmuir equation
                                   (r2 - 0.954)
     Freundlich equation
        (r2 = 0.996)
  Double-reciprocal
near Langmuir equation
     r2 =0.916)
                                               60
                                                         SO
                                                                   100
                                                                             120
                               Equilibrium arsenic concentration (mg/L)
Figure 41  Adsorption of arsenic by a kaolinite clay sample at 25°C, described by
the traditional linear Langmuir, double-reciprocal Langmuir, and Freundlich equation.
The mean pH of the soil-solute suspensions was 7.8.
                                                 66

-------
                                        CHAPTER 16

                         APPLICATION OF BATCH-ADSORPTION DATA

Adsorption data are used in describing the partitioning of chemicals between soils and water, and have
been used successfully as input parameters in many models describing the movement of chemicals in
soil (Dragun, 1988). Batch-adsorption data have also been applied successfully to groundwater systems.
For example, Curtis et al. (1986) found that the rates of movement of halogenated organic solutes in a
sandy aquifer in Canada were in good agreement with those predicted from adsorption data. In a study
described by CH2M Hill, Inc. (1986), data on the distribution and concentration of organic solutes at a
field site in Indiana were in  good agreement with data predicted from laboratory adsorption studies.

Miller et al. (1989) found that isotherms generated with a batch technique were very similar to those de-
rived from flow-through column experiments for the adsorption of anions by soils. Adsorption tended to be
greater in the flow systems, possibly because of precipitation or reduced competition between the solutes
and desorbed antecedent species.

This chapter is a brief introduction to the application of batch-adsorption data in calculations of solute
movement through compacted landfill liners. These calculations are used particularly for estimating the
minimum thickness of liner required to prevent pollutant movement beyond a certain depth of the liner for
a specified period of time. As leachate moves through a liner, the movement of chemical solutes in the
leachate may be retarded if they are adsorbed by the liner. We may define R as the ratio of the velocity of
the leachate to that of the solute,

                                     R  « leachate/ Solute                                     [30]

The flterm is called the retardation function or factor. When the solute is not retained by the liner, /?
equals 1 : the solute moves at the same velocity as the leachate. Increasing degrees of adsorption yield
larger values for R. The retardation factor may also be defined by an empirical relationship (Freeze and
Cherry, 1979, and references cited therein) as
                                                 8

where pb is the dry bulk density of the liner, Kd is a distribution coefficient, and  6 is the volumetric water
content of the liner. The distribution coefficient is a parameter that describes the partitioning of solutes be-
tween the leachate and liner soil materials at equilibrium. The distribution coefficient may be defined as


                                         f - **

where S is equal to x/m (the amount adsorbed per mass of adsorbent), and C is the equilibrium concentra-
tion of the solute. In other words, equation 32 is the slope of an adsorption isotherm.

Before equation 31 can be used, a functional relationship for dS/dC must be determined. The possible so-
lutions range from simple assumptions to complex numerical solutions. The simplest case is one in which
the adsorption of the solute conforms to a Freundlich equation (chapter 14) isotherm where the 1/n term
is unity,


                                   — = S = KfC""  = K,C                                   [33]
                                                67

-------
Such an isotherm is termed linear; a plot of S versus C is a straight line. The slope of this type of plot
yields Kd,

                                       -= = KforKd
                                                                                           [34]
hence,
                                                                                            [35]
The retardation factor is unitless; if Kd is in milliliters per gram, then the units of the term

                                p6 (g/cm3) K"d(mL/g)/0 (cmVcm3)

cancel because 1 cm3 = 1 ml.

When a linear isotherm is used, the Freundlich constant (Kf} reduces to the simple partition constant
(Kd), a single-value number used to calculate solute-adsorbate partitioning at any equilibrium concentra-
tion of the solute. Because of its mathematical simplicity, this approach (the linear isotherm assumption)
has been widely used and may be valid for many dilute systems. When the adsorption isotherm of a sol-
ute is a nonlinear function (\ln *=  1), the retardation factor is concentration-dependent:
hence,
                                   _d_
                                    dC
                                  R(C) = 1 +
                                                   n
                                                        -1
                                                   en
                                                                                            [36]
                                                                                            [37]
Equation 37 is complicated by the fact that the numerical value of R depends on the concentration of the
solute. Solute movement may be seriously underestimated if, when dealing with nonlinear isotherms, in-
vestigators assume that a constant retardation factor is valid for a given system.

Rao (1974) developed an empirical technique to estimate a weighted-mean adsorption partition coeffi-
cient (Kd) for the Freundlich equation. In this technique, the rate of adsorption with respect to concentra-
tion (3S/0C) is normalized by the total amount of solute in a given concentration range,
                 Kri =
  fdS
  ~*r{
1 "v/
 i»Cn
 \dC
                                          n
                                                dC
                                                         A/n
                                        f'
                                           dC
                                                        Co
                                                                                            [38]
The solute concentration C0 isjhe highest concentration (before contact with the adsorbent). If Kf is in
units of mL1/nu,g(1"1/n)/g, then K^may be expressed in milliliters per gram, since
                                        mi 1
                                                                   mL
                                                                    9
A weighted-mean retardation factor (R) may be calculated as
                                               68

-------
                                   R =  1 +
                                                8
                                               [39]
In a study concerned with pesticide adsorption by a soil sample, Davidson et al. (1980) found that the er-
ror introduced by assuming linear adsorption isotherms was not serious at low concentrations
(<10 mg/L) but became significant at higher concentrations. Van Genuchten et al. (1977) proposed
an alternative method for isotherm linearization that the reader may wish to examine.

To demonstrate possible applications of these concepts, the following examples are presented to illus-
trate how batch-adsorption data are used to estimate clay liner thickness.

In this hypothetical example, the metallic waste described in appendix B is to be placed into a disposal ba-
sin lined with Cecil clay loam (see appendix A). The soil, which was graded, blended, and compacted,
has a saturated hydraulic conductivity of 10"7 cm/sec. The major concern of the company operating the
disposal facility is the possible uncontrolled movement of a leachate plume containing high concentra-
tions of lead in solution. In a preliminary analysis, this company conducted batch-adsorption experiments
using a Pb(NO3)2 salt and samples of the Cecil soil (table 14). The question posed is, what must the mini-
mum thickness of the liner be to attenuate the lead from solution over a 5-year operating life and a 30-
year post-closure period?
Several approaches can be used to answer this question. For each approach, the mean pore velocity of
the leachate through the liner must be calculated by using Darcy's law as
                                        V = Ksati/n
                                               [40]
where /Csat is the saturated hydraulic conductivity of the liner, /is the hydraulic gradient (dH/dZ), and neis
the effective (water-conducting) porosity of the liner.

If we assume saturated conditions, subject to steady-state flow through an isotropic liner over time t, and
neglect the effects of dispersion and diffusion, equation 40 can be combined with equation 31 to yield

                                      Z=  tKsali/Rng                                        [41]

where Zis the estimated vertical distance of migration of the solute in centimeters, and t is time in sec-
onds.

Equation 41 treats solute movement as a piston-flow problem: a chemically uniform slug of leachate mov-
ing downward. This equation is simple and may readily be used to estimate the minimum thickness of a
liner. The application of the equation is simplified by assuming that the isotherm is linear. In this example
(table 14 and figure 42), a linear regression of the data through the origin (Steel and Torrie, 1960) yielded


                                       •£ = S = 342 (Pb)
Moreover, the liner is assumed to have the following properties:
                                     ne  =
                                       r  =
                                     Pb  =
                                    "sat  =
                                       /  =
                        1.1038 x109 sec  =
0.09 cm3/cm3
0.36 cm3/cm3
1.7g/cm3
1 x10"7 cm/sec
dH/dZ= 1 cm/cm, and that
35 years
                                                69

-------
 With these assumptions, the retardation factor becomes
                                    R = 1+-

 and solving equation 40 becomes

                       Z = (1.1038 x 109) (1 x 10~7) (1)71619(0.09) = 0.8 cm

 On the basis of this approach, the compacted liner would have to be at least 1 cm thick to adsorb lead
 over a 35-year period. But although the application of a linear isotherm yields a reasonable coefficient of
 determination (r  = 0.95), inspection of figure 42 indicates that this approach overestimates lead adsorp-
 tion at high lead concentrations and underestimates adsorption at lower concentrations. The adsorption of
 lead (table 14) is more accurately described by a Freundlich equation,
                                      -   = S = 291(Pb)
                                                       °-492
As a second level of refinement, the nonlinearity of the isotherm is considered using equation 38 to esti-
mate a weighted-mean retardation factor (Davidson et al., 1980). An appropriate value for CQ was deter-
mined from a laboratory extract of the metallic waste sample (appendix B), which suggests that the
maximum amount of lead that initially will come in contact with the liner is approximately 15 mg/L Pb A re-
vised retardation factor is derived from equation 38:
and the minimum thickness, based on the weighted-mean retardation factor, is

                       Z = (1.1038 x 109) (1 x 10~7) (1)7348(0.09) = 3.5 cm ,

Thus, when the nohlinearity of the isotherm is considered, the minimum thickness of the liner is estimated
to be about 4 cm. As a third level of refinement, the chemical composition of the leachate was consid-
ered. The first two estimates were based on lead adsorption from a pure Pb(NO3)2 solution. Laboratory ex-


             Table 14 Lead adsorption data for a Pb(NO3)2 salt and the Cecil clay
             (volume of solution, 200 mL; adsorbent weight, 10.18 g)
Initial
cone
(mg/L)
2.07
5.11
5.11
6.22
7.28
10.2
10.2
12.4
14.6
14.6
Equilibrium
cone
(mg/L)
0.05
0.11
0.11
0.16
0.22
0.41
0.43
0.65
0.94
0.94
Amount
adsorbed
(x/m) as \ig/g
61
100
100
121
141
196
195
235
273
273
PH
4.79
4.74
4.75
4.74
4.73
4.68
4.67
4.66
4.62
4.62
EC
(dS/m)
27
33
35
34
33
39
40
45
45
43
                                              70

-------
                               280
                              240
                               200
                               160
                           o
                           +->
                           c
                           3
                           o
120
                                80
                                40
                                  0.0   0.2    0.4    0.6     0.8     1.0

                                     Equilibrium lead concentration (mg/l)


                   Figure 42  Lead adsorption by Cecil clay loam at pH 4.5 and at 25°C,
                   described by a linear Freundlich equation forced through the origin.



tracts, of the waste also contained large concentrations of zinc (appendix B). The adsorption of lead from

the extracts was significantly less than that from the pure Pb(NO3)2 solution, presumably because of com-

petitive interactions between Zn2+ and Pb2+ for adsorption sites. The net effect is that lead could be more

mobile in the presence of zinc. The adsorption of lead by Cecil from the laboratory extract of the waste

can be described by
                                           = s = 70 (Pb)
                                                         °-481

                                                  71

-------
 If the minimum liner thickness is recalculated using these isotherm constants and equations 38 and 40,
 the thickness is estimated to be about 15 cm, again assuming that the initial lead concentration in the
 leachate is 15 mg/L Clearly, migration distance estimates based on adsorption data from pure, single-sol-
 ute tests may underestimate the minimum thickness of liners because these estimates fail to account for
 competitive interactions that may significantly reduce adsorption. At the next level in refining the esti-
 mated liner thickness, the effects of dispersion and diffusion are considered. In saturated homogeneous
 materials that are subjected to steady-state flow conditions along a flow path z, the change in solute con-
 centration as a function of time may be generalized (Ogata, 1970; Bear, 1972; Boast, 1973; Freeze and
 Cherry, 1979) as
                                        ac2
                                        'az2'
 ac   pjb as
'• dz ~ e  ar
                                                                                             [42]
where
           Pb
            6
            S
            t
                  concentration of the solute,
                  effective diffusion-dispersion coefficient (distance/time) along the flow path z,
                  mean convective flow velocity (distance/time) along the flow path z,
                  bulk density (wt/vol) of the material,
                  volumetric water content (vol/vol),
                  amount of solute adsorbed per mass of adsorbent (x/m), and
                  time
Equation 41 can be rearranged as
                                                                                             [43]
                                     «*       l/^.      w«-

where R is the retardation factor

The analytical solution to this second-order differential equation (Ogata, 1970) is given by

                        C   1 I"  ,  ( z- Vt*  }      (Vz}  .  ( z+Vt*
                       -pr = — erfc  ———^ +exp \-=r erfc  	^
                       Co   21     l2(D2/*)0-5l      \Dz\     2(Dzif*)a5
                                                                                             [44]
where  CICQ  =  ratio of the solute concentration at time t and distance z to the intial solute
                  concentration C0,
         erfc  =  complementary error function,
            V -  average linear pore water velocity (cm/sec),
          DZ  s  vertical dispersion coefficient (cm2/sec),
          t *  =  retarded time (actual time divided by the retardation factor of R or R), and
            z  =  vertical distance of migration (cm).

Furthermore, Dz - aV+ D*, where a is the dispersivity (cm) and D* is the effective diffusion coefficient in
porous media (cm2/sec).

In the following examples, the three previous liner thickness estimations were recalculated using equation
43. The only additional information needed to conduct this analysis was a dispersivity value. The disper-
sivity has been found to be scale-dependent and is estimated to be about 10% of the distance measure-
ment of the analysis (Gelhar and Axness, 1981). A diffusion coefficient of Pb2+ in soil of 1 x I0"6cm2/sec
was used in this analysis  (Daniel et al., 1988). In figure 43, the relative concentration (C/C0) is shown as a
function of distance of migration after 35 years.  Case A represents the first situation, in which the adsorp-
tion of lead,  a Pb(NO3)2 salt, was assumed to be depicted by a linear isotherm. Case B corresponds to
the second calculation, in which a weighted-mean retardation factor was used with the Pb(NO3)2 solute-
                                               72

-------
   c
   o
   01
   EC
   8  0.4. -
      0.2 _
      0.0
                                    12        16        20

                                    Distance of migration (cm)
24
28
                    32
  Figure 43   Predicted distance of lead migration in Cecil clay loam after 35 years, based on three approaches:
  case A (linear isotherm assumption, Pb(NO3)2 salt); case B (weighted-mean retardation factor, Pb(NO3)2 salt);
  and case C (weighted-mean retardation factor, multicomponent waste extract).


soil system.  Case C represents the adsorption of lead from the multicomponent-waste extract, coupled
with the corresponding weighted-mean retardation factor. Case C, which takes into account dispersion,
indicates that lead may move farther than predicted by an elementary piston-flow model (eq. 40). The ef-
fects of diffusion on the predicted migration distances were negligible (not shown).

An element of interpretation is involved in evaluating graphs (see fig. 43) for the purpose of estimating
liner thickness. A judgment must be made as to which C/C0 ratio, for practical considerations, translates
into the minimum significant concentration. In this hypothetical example, the regulatory agency decided
that a lead concentration of <0.05 mg/L (the U.S. drinking water standard for lead) would be an opera-
tional definition of the compliance concentration.

If the initial lead concentration is 15 mg/L, the lead concentration of <0.05 mg/L is predicted to occur at a
depth of 5 cm in  case A and at 10 cm in case B. The results for case C represent the fourth level of refine-
ment in this analysis, yielding the most accurate liner thickness estimate. After 35 years, the concentra-
tion of lead in solution would be reduced to <0.05 mg/L at a depth of 35 cm on the basis of these calcu-
lations. Consequently, the minimum liner thickness would be 35 cm. The actual thickness  necessary in a
field application must be somewhat greater to allow for nonequilibrium conditions and the normal engi-
neering safety factors. The application of batch-adsorption data provides an estimation of boundary condi-
tions, i.e., the minimum thickness.

In summary, the minimum liner thickness for a hypothetical liner varied from 1 to 35 cm, depending on the
approach (table  15).  Liner thickness estimates can be refined further if the adsorption data can be inte-
grated with other information, such as the solubility of solid phases, oxidation-reduction equilibria, co-
solvent effects, and the design and performance of on-site earthen liners. This information would include
                                                73

-------
seepage rate through the cover, fraction of seepage that will pass through the liner, and other water flux
information that would allow calculation of the distribution of a pollutant in soil as a function of time and
space.

        Table 15 Approaches for estimating minimum liner thicknesses on the basis of adsorption
Flow model
Piston flow*
Piston flow
Advection dipersionf
Advection dispersion
Piston flow
Advection dispersion
Isotherm
treatment
linear
nonlinear
linear
nonlinear
nonlinear
nonlinear
Solute system
single solute
single solute
single solute
single solute
mixture:}:
mixture
Minimum liner
thickness (cm)
1
4
5
10
15
35
        * Represented by equation 41.
        t Represented by equation 44.
        t Laboratory extract.
                                               74

-------
                                        CHAPTER 17

             LABORATORY PROCEDURES FOR GENERATING ADSORPTION DATA

Procedures for the determination of the soil:solution ratio, equilibration time, and other parameters neces-
sary for the construction of adsorption isotherms are contained in this chapter:
                                                                          page
       1        Scope of Application                                           69
      2        Summary of Methods                                           69
      3        Interferences                                                  69
      4        Terminology and Definitions                                    70
      5        Equipment and Procedural Requirements                         71
      6        Volatile Organic Solutes: Experimental Considerations             73
      7        Preparation of Adsorbents                                      74
      8        Determination of Soil:Solution Ratios for Ionic Solutes              75
      9        Determination of Soil:Solution Ratios for Nonionic Solutes          77
     10        Determination of Equilibration Time                              77
     11        Construction of the Environmentally Conservative Isotherm         78
     12        Construction of the Constant Soil:SoIution Ratio Isotherm           79

The rationale for these procedures, presented in previous chapters, should be studied before attempting
to use them. This chapter refers to parts of the TRD that elucidate topics relevant to a specific procedural
step. The flow diagram (fig. 44) summarizes the procedures and their interrelationships.

1   Scope of Application
    1.1 The extent of adsorption of a chemical (solute) from solution by an adsorbent (sediment, soil,
    clay) at equilibrium can be estimated using these procedures.

    1.2  These methods apply to the generation of adsorption isotherms or curves for inorganic and or-
    ganic (volatile and nonvolatile) compounds; these isotherms indicate how the extent of adsorption var-
    ies with the  equilibrium concentration of the solute.

    1.3  Contingencies within these methods allow for the construction of adsorption isotherms at various
    solute concentration ranges.
    1.4  These methods can be used for constructing adsorption isotherms to study the adsorption behav-
    ior of solutes in synthetic waste solutions, laboratory extracts, or field leachates including aerobic and
    anaerobic solid-liquid systems.

2   Summary of Methods
The experimental design of these methods is based on a batch technique as opposed to a column ap-
proach. Two general techniques for obtaining adsorption data are incorporated in these methods. The first
technique involves mixing a batch of solutions, each with the same volume but containing serial dilutions
of the initial solute concentrations, with a fixed mass of adsorbent in each reaction vessel. The second
technique involves mixing a batch of solutions, each with the same volume and initial concentration of the
solute, with different amounts of the adsorbent. In either case, the change in solute concentrations after
contact with the adsorbent provides the basis for the construction of adsorption isotherms (chapter 12).
The appropriate soihsolution ratios and equilibration times are determined to maximize the accuracy of
the adsorption isotherm and to complement analytical capabilities.

3   Interferences
Solutes of unknown stability must be handled with care to determine whether precipitation, hydrolysis,
photodegradation, microbial degradation, oxidation-reduction (e.g., Cr3* to Cr  ), or other physicochemical
processes are operating at a significant rate within the time frame of the procedure. The instability and
hence loss of the solute from solution may affect the outcome of this procedure (see chapter 4). The com-
                                               75

-------
                           Generation or collection
                            of solution containing
                                test solute
                            Determination of solute
                               solution stability
                         (hydrolysis, photodegradation,
                            microbial degradation,
                                and volatility)
                         Determination of Interactions
                         between solute solution and
                            laboratory equipment
                       Determination of equilibration time
                           Construction of constant
                         soil to solution isotherm (CSI)
Collection of adsorbent
                                                                          Determination of
                                                                          percent moisture
                            Determination of soil to
                               solution ratios
Preparation of adsorbent
       Air drying



Reduction of aggregates
                                                                      Splitting and subsampling
                                                                Construction of environmentally
                                                                 conservative isotherm (ECI)
                 Figure 44   Flow diagram of the procedures for generating batch adsorption data.


patibility of the method and the solute of interest may be assessed by determining the differences be-
tween the initial solute concentration and the final blank concentration of the solute. If this difference is
greater than 3%, then the adsorption data generated must be carefully evaluated (see 8.5.11).

4  Terminology and Definitions

    4.1  Solute—chemical species (e.g., ion, molecule) in solution

    4.2 Solute solutions include:
                                                       76

-------
   4.2.1 A solution of reagent water containing a known amount of a solute derived from laboratory
   reagents.
   4.2.2 A solution containing a variety of solutes extracted from a material in a laboratory setting by
   the use of methods such as the ASTM-A or ASTM-B extraction procedures. (Note: neither the
   EPA Extraction Procedure nor the proposed Toxicity Characteristic Leaching Procedure is recom-
   mended. These procedures were designed for waste classification and were not intended to pro-
   duce solutions that mimic in-situ leachates.)
   4.2.3 A solution containing a variety of solutes collected in a field situation representing a
   leachate or waste effluent.
4.3 Adsorption—a physicochemical process whereby solutes are retained by an adsorbent and
concentrated at solid-liquid interfaces (chapter 1).
4.4 Adsorbate—chemical species adsorbed by an adsorbent.
4.5 Adsorbent—substance that adsorbs the solute from solution.

Equipment and Procedural Requirements
5.1 Laboratory equipment
    5.1.1 Agitation equipment: the National Bureau of Standards extractor (rotating tumbler) or
    equivalent will be used exclusively as the agitation apparatus (chapter 8).
    5.1.2  Rotation rate: with procedures involving inorganic, volatile, and nonvolatile organic com-
    pounds, the rotary extractor will be operated at 29 ± 2 rpm.
    5.1.3  Glove box or bags: when anaerobic adsorbent-solute systems are being handled, these
    procedures may have to be conducted in air-tight enclosures filled with an oxygen-free inert gas
    (e.g., N2, Ar) to prevent or retard oxidation.
5.2 Phase-separation equipment
    5.2.1  Inorganic compounds: a filtration apparatus made of materials compatible with the solu-
    tions being filtered and equipped with a 0.45-u.m pore-size membrane filter or a constant-tempera-
    ture centrifuge capable of separating >0.1-u.m particles will be used to separate the solid phase
    from the solid-liquid suspensions.
    5.2.2 Filtration membrane: if filtration is used, the affinity of the filtration membrane for the solute
    must be evaluated to prevent errors in the results.
    5.2.3 Organic compounds: a constant-temperature centrifuge, compatible with the reaction con-
    tainers and capable of separating >0.1-u.m particles, should be used for organic solutes. The
    transfer of the organic solute solutions from the reaction containers to centrifuge containers is not
    an acceptable procedure because of adsorption, volatilization, and other losses. The reaction con-
    tainer should be used as the centrifugation container. Filtration of organic solutions is .rjoj a recom-
    mended practice (chapter 7).
    5.2.4 Calculation of centrifugation time may be calculated by using equation 1,
                                  t =
9riln(/Vf?f)
2o>2r2(pp-p)
[1]
                                            77

-------
        where
                  t  »  time(min),
                  tl  =  viscosity of water (8.95 x 1 0"3 g/sec-cm at 25°C),
                  r  -  partical radius (cm),
                 pp  =  partical density (g/cm3)
                  p  -  density of solution (g/cm3),
               rpm  m  revolutions per minute,
                 Rt  =  distance (cm) from the center of the centrifuge rotor to
                        the top of solution  in centrifuge tube, and
                Rb  =  distance (cm) from the center of the centrifuge rotor to bottom of
                        the centrifuge tube.

Removal of particles that are as small as 0.1  urn in radius and have a particle density of 2 65 g/cm3 from
a solution with a density of 1 g/cm3 may be estimated using equation 2,

                                        3 71 x 108
   5.3 Reaction containers

       5.3.1  Inorganic solutes: containers compatible with the rotary extractor should be used with inor-
       ganic solutes. The containers shall be composed of materials that adsorb negligible amounts of
       the solute. They must have a watertight closure made of chemically inert materials (polypro-pvl-
       ene, Teflon, or similar material). The size of the container should be such that the solid and liquid
       phases will fill about 80% to 90% of the container.

       5.3.2  Nonvolatile organic solutes: amber glass serum bottles and stainless steel centrifuge tubes
       or bottles compatible with the rotary extractor and centrifuge are suggested to be used in conjunc-
       tion with nonvolatile organic solutes. The container must have a watertight closure made of chemi-
       cally inert materials (Teflon, plastic, or similar material). The size  of the container must be com-
       patible with the centrifuge, and be such that the volume of the solid and liquid phases should fill
       80% to 90% of the container.

       5.3.3  Volatile organic solutes: amber glass, 125-mL serum bottles (Wheaton no  223787 or
       equivalent) fitted with Teflon septa (Pierce no. 12813 Tuf-Bond Discs or equivalent) will be used
       with volatile organic solutes. The size of this serum bottle (125 ml) is compatible with several
       types and brands of centrifuges. This size provides sufficient volume such that the volume of the
       solid and liquid should occupy 1QQ% of the container (i.e., there should be no head space).

       5.3.4 Note that the commonly available materials for containers can be ranked starting with the
       material most inert with respect to the adsorption of hydrophobic solutes (T.C. Voice written com-
       munication,  1986):

          Corex,
          Pyrex (not much different from Corex),
          silanized serum bottles,
          other glasses, and
          stainless steel, Teflon, and plastic.

  5.4 Reagents

      5.4.1  Reagent-grade chemicals will be used in all experiments and must conform to the specifica-
      tions of the American Chemical Society. Other grades may be used, provided that the reaqent is
      pure enough to be used without lessening the accuracy of the determination
                                              78

-------
    5.4.2 Unless otherwise indicated, references to water mean type IV reagent water, as defined in
    the Handbook for Analytical Quality Control in Water and Wastewater (EPA-600/4-79-019).

5.5 Solute Solution and Adsorbent Requirements
    5.5.1 To construct adsorption isotherms for inorganic solutes, a minimum of 5 liters of solute solu-
    tion would be required, based on the use of 200-mL samples of the solute solution with 250-mL
    reaction containers. Investigators using reaction containers that are a different size should adjust
    the estimated total volume of solution proportionately.
    5.5.2 To construct adsorption isotherms for organic solutes, approximately 9 liters of solute solu-
    tion would be required, based on the use of 100-mL samples of the solute solution with 125-mL
    reaction containers. Investigators using different-sized reaction containers should adjust the esti-
    mated total volume of solution proportionately.
    5.5.3 The mass of adsorbent required to complete this procedure will vary depending on the vol-
    ume of reaction containers, soihsolution ratios, and related factors. Based on 250-mL reaction
    containers and the minimum soihsolution ratio of 1:4 (50 g adsorbent per 200 mL of solute solu-
    tion), about 2 kg of adsorbent would be required.
    5.5.4 This procedure should take about 5 to 9 days, excluding time for analysis.


Volatile Organic Solutes: Experimental Considerations
6.1 Stock solutions could either be purchased as certified solutions or prepared from pure standard
materials (liquid or gaseous phases). Solutions should be prepared in methanol. The use of pipettes
to transfer solutions cannot be recommended; glass syringes should be used to prevent losses due to
volatization. Because of the toxicity of some volatile organic compounds, solutions should be pre-
pared and transferred in a fume hood, and a NIOSH/MESA-approved toxic-gas respirator should be
used by the analyst.
6.2 Preparation of stock volatile solute solutions
    6.2.1  Place approximately 9 mL of methanol into  a 10-mL ground glass stoppered volumetric
    flask. Allow the flask to stand open until all methanol-wetted surfaces have dried. Weigh the flask
    with the remaining methanol to the nearest 0.01 mg. Using a syringe, immediately add the test
    solute, until the change in weight of the flask corresponds to the desired concentration of the test
    solute in the methanol. Be sure that the drops of solute fall directly into the methanol without con-
    tacting the neck or sides of the flask. Dilute to volume with methanol, put the stopper on the flask,
    and mix by inverting it several times.
    6.2.2 Transfer the stock solution into a Teflon-sealed screw-cap vial. Store, with little or no head
    space, at approximately 4°C. All stock solutions must be replaced after 1 month, or sooner if a
    comparison with quality-control standards indicates  a loss of accuracy.

    6.2.3 Stabilize the temperature of the stock solution at 20°C before preparing secondary solu-
    tions.
    6.2.4 Store all solutions so that head space within the container is zero or minimal.

 6.3 Preparation of volatile organic compound  solutions

    6.3.1 Place 990 mL of type IV water that has been boiled and cooled to 20°C into each of a se-
     ries of clean 1-L amber-glass bottles. (Generally eight solute concentrations are required for com-
     pletion of the adsorption procedures.) Seal the bottles with open-top screw caps fitted with
    Teflon-lined septa.
     6.3.2  Inject measured amounts of the stock solution (prepared as described in section 6.2) into
     each of the bottles. Mix by inverting the bottles several times but avoid excessive shaking, which
     may result in partial loss of the solute.
                                             79

-------
        6.3.3 Solutions stored in containers with head space are not stable and should be discarded 1
        hour after preparation if not used in an experiment.

     6.4 Filling of reaction containers

        6.4.1 Immediately upon completing the steps described in section 6.3, pour each solute solution
        carefully— to minimize agitation— into preweighed reaction containers or other containers that
        have a predetermined volume (section 6.5) and contain specific amounts of adsorbent  Fill the
        containers completely; allow no head space. Shake gently to remove trapped air from the adsorb-
        ent. Place the Teflon-faced septum and aluminum seal on the container and invert to be sure no
        head space remains.

     6.5 Determination of Reaction Container Volume

        6.5.1  When transferring the solutions prepared as described in section 6.3 into the reaction con-
        tainers, pour them quickly but gently into containers of predetermined weight or volume.

        6.5.2 Because the volume of solute solution is not measured during transfer into the reaction con-
        tainers, this volume is determined indirectly.

        6.5.3  Reaction vessels containing the same amount of adsorbent as those described in section
        6.4 to determine the container volume for each soihsolution ratio. Pipette type IV water is pipetted
        into each container until there is no head space. With a calibrated syringe, measure the water
        added to each of the containers. The amount of solute solution referred to in section 6 4 should
        be the volume as determined in this section (6.5).

       6.5.4 Alternatively, determine the volume of solution added (described in section 6 4) by weigh-
       ing | the container with the adsorbent before and after adding the solution. The weight can be con-
       verted to a volume if the density of the added solution is known.

    6.6 Throughout all experiments, use blanks to determine effects of adsorption/desorption from con-
    tainers as well as losses due to volatilization. Refer to section 8.5.1 1 for discussion of blanks.

    6.7 For more information on preparing solutions for volatile constituents or the analyses of these con-

    fnSs^^^

7   Preparation of Adsorbents

    7.1 Spread samples of adsorbents such as soils, clays or sediments out on a flat surface in a laver
    no more than 2 to 3 cm deep. Then allow them to air dry, out of direct sunlight, until they are in
    equilibrium with the moisture content of the room atmosphere. The sample should be dried enough to
    facilitate processing and subsampling. Do not oven-dry samples (chapter 2). Process anaerobic sam-
    ples in a similar manner for these and subsequent steps, but these operations should be conducted in
    a glove box or glove bag filled with an oxygen-free inert gas (e.g., N2 or Ar) to prevent oxidation

   7.2 Weigh the entire sample after it has been air-dried. Pass the sample through a 2-mm-screen
   sieve. Using a clean mortar and rubber-tipped pestle, crush large aggregates without grinding the
   sample. Aggregates such as pebbles and stones that cannot be crushed should be removed col-
   lected, and weighed.

   7.3 Mix the sieved material until the sample is homogeneous. To obtain subsamples of size  use a rif-
   rSno Der7 • °m|- °SeJ unbiased sP|ittin9 Procedure (Annual Book ofASTM Standards: method

   So£&™^             to Testing Size' in part 1 4: or method D2013-72- p
   7.4 Determine the moisture content of the air-dried sample by using method D221 6, Laboratory De-
   termination of Moisture Content of Soils, from Annual Book ofASTM Standards, part 1 9.

   7.5 Determine the mass of the sample, corrected for moisture content, required for study.
                                            80

-------
       7.5.1  Determine the air-dry soil (adsorbent) mass equivalent to the desired mass of oven-dried
       soil:
                  A  =  Ms[1 +(M/100)]

       where     A  =  air dry soil mass (g),
                 Ms  =  mass of oven-dried soil desired (g), and
                  M  =  percent moisture

8   Determination of SoihSolution Ratios for Ionic Solutes
    8.1 A series of soihsolution ratios ranging from 1:4 to 1:500 should be tested and evaluated for the
    construction of adsorption isotherms (chapters 9 and 11).
    8.2 The soil:solution ratio is defined as the oven-dry equivalent mass of adsorbent in grams (section
    7.5) per volume in milliliters of solution.
    8.3 Recommended soihsolution ratios are 1:4,1:10,1:20,1:40,1:60,1:100,1:200,1:500. Ratios
    greater than 1:500 are rarely needed for most ionic solutes. In circumstances requiring soihsolution ra-
    tios greater than 1:500 that meet the criteria outlined in section 8.5.14, use the ratios 1:1000,1:2000,
    1:5000 and 1:10,000. The determination of a soiksolution ratio  may be an iterative  process, whereby
    the eight ratios between 1:4 and 1:500 are tested before attempting the extremely "dilute" systems
    (1:1000, and higher). Using an iterative process will reduce the amount of solute solution used, and
    will help ensure that enough  solution will exist to complete the entire procedure. Ratios less than 1:4
    should not be used because of limitations in mixing.
    8.4 An example of how different soihsolution ratios are made is given below for an air-dried sample
    with a moisture content of 3%:
             Soihsolution
              ratio (g/mL)

                 1:4
                 1:10
                 1:20
                 1:40
                 1:60
                 1:100
                 1:200
                 1:500
 Air-dry
weight (g)

  51.5
  20.6
  10.3
   5.15
   3.43
   2.06
   1.03
   0.412
Oven-dry equivalent
  of adsorbent (g)

        50.0
        20.0
        10.0
         5.00
         3.33
         2.00
         1.00
         0.400
  Volume of solution
containing solute (ml_)

        200
        200
        200
        200
        200
        200
        200
        200
     8.5 Soil solution procedure
        8.5.1  Calculate the masses of adsorbent samples for the various soihsolution ratios based on an
        oven-dried equivalent weight (section 7.5), such that for nonvolatile solutes, the volume of adsorb-
        ent plus solution occupies 80% to 90% of the container and for volatile solutes 100%.
        8.5.2 Weigh the samples of adsorbent to be used in the soihsolution series. If handling anaerobic
        adsorbent-solute systems, conduct steps 8.5.2 to 8.5.7 in a glove box or bag before placing the
        containers on the rotary extractor.
        8.5.3 Place the weighed samples into clean, labeled containers.
        8.5.4 Pipette the solution containing the solutes (stock solution) into each container holding the
        adsorbent. The volume of solution should be identical in all containers.
                                                81

-------
  8.5.5  Pipette the stock solution into a container holding no adsorbent. This sample will be the
  "blank" and designated as CB. For each set of tests a minimum of one blank, and preferably
  three blanks, should be tested simultaneously and under identical conditions as the samples.
  8.5.6  Close the bottles, ensuring a watertight seal, and place on a rotary tumbler for mixing.

  8.5.7  Collect, preserve, and analyze an aliquot of the stock solution to determine the initial con-
  centration of the solute(s) before contact with reaction containers, adsorbent, phase separation
  materials, and other surfaces. This sample will be designated as C0. The volume and preserva-
  tion techniques of the aliquot will vary depending on the solute and analytical method.

  8.5.8 Continuously agitate samples at 29 ± 2 rpm for 24 ± 0.5 hours at room temperature (22 ±


  8.5.9 After 24 hours of agitation, open containers. If the suspensions are anaerobic, return the
  containers to a glove box or bags prior to opening the containers and make all  measurements in
  the inert atmosphere of the glove box or bag. Observe and record the solution temperature  oH
  and any changes in the adsorbent or solution.                                             '

  8.5.10 Separate the solid and liquid phases of each sample, using either centrifugation or filtra-
 tion (section 5.2). Determine the electrical conductivity of an aliquot of each supernate (see chap-
 ter 6). Collect and preserve aliquots of each supernate of sufficient volume to determine the
 solute concentration.

 8.5.11  After analysis of all the solutions generated by the soil:solution procedure, compare the in-
 itial solute concentration(s) and blank samples to determine whether there was adsorption or
 desorption of the solute onto or from surfaces other than the adsorbent. If the difference between
 the blank and initial solute concentrations is greater than 3%, the adsorption data must be cor-
 rected. To make this assessment, determine the percent difference between the initial concentra-
 tion and the blank solute concentration:
                               %P=(CO:CB)XIOO
 where
%D
  C0
 CB
                   percent difference,
                   initial solute concentration (mg/L, u,g/L), and
                   solute concentration (mg/L, u.g/L) in blank solution.
 If %D is a negative value, the solute concentration in the blank was greater than the initial solute
 concentration. Subtract the difference in concentration from all adsorption data, excludina the
 stock or initial concentration value.

 8.5.12 Using the analyzed initial solute concentration and the final solute concentration for the
 vanous soil:solution ratios tested, calculate the percent of solute adsorbed:
where
%A
 Co
                 =  percent adsorbed,
                 =  initial solute concentration (mg/L, u/L), and
                 =  solute concentration after contact with the adsorbent.
A negative value implies a contamination problem. Examine the laboratory technique and/or
cleaning procedures. The adsorbent may contain previously adsorbed constituents that are
desorbing into solution.
                                        82

-------
       8.5.13 Select a soil solution ratio indicating between 10% and 30% adsorption of the highest sol-
       ute concentration. Use this ratio to determine the equilibration time (chapter 10) and to generate
       data for construction of a constant soihsolution isotherm (CSI). Often, several soihsolution ratios
       will generate between 10% and 30% solute adsorption. Selection of a specific soil-to-solution ra-
       tio is the investigator's prerogative, with the limitation that it be one listed in section 8.3 (chapters
       9 and 11).
9   Determination of Soil:Solution Ratios for Nonionic Solutes
    9.1 For ionic solutes, a suitable soihsolution ratio must be determined empirically, but a useful soil:so-
    lution ratio for nonionic solutes (hydrophobic organics) can be calculated if the organic carbon content
    of the adsorbent and the water solubility of the solute are known. The equations and their derivations
  i  for determining the soil:solution ratios for nonionic solutes are given in chapter 10.
    9.2 The soil solution ratios listed in section 8.3 most closely matching the calculated soil:solution ratio
    shall be used throughout this procedure. If the calculated ratio is in the middle of two ratios listed in
    section 8.3, the lower ratio (greatest mass of absorbent per milliliter of solute) is recommended to ob-
    tain the highest precision and accuracy.

10  Determination of Equilibration Time
    10.1  To determine equilibration time, use the soil:solution ratio as determined in section 8.5.13 for in-
    organic solutes and in section 9 for hydrophobic organic solutes.
    10.2 Use a minimum of four agitation intervals to determine the equilibration time. Recommended in-
    tervals are 1,24,48, and 72 hours, and represent the amount of time the solution and adsorbent are
    in contact.
    10.3 Weigh the adsorbent on an oven-dry basis (section 7.5) and place into clean, labeled contain-
    ers. If handling anaerobic systems, perform steps 10.3 and 10.8 in a glove box or bag before placing
    the containers on the rotary extractor.
    10.4 Pipette the solute solution into the various containers at the times designated in step 10.2. Im-
    mediately cap the container and place on  the rotary extractor and agitate at 29 ± 2 rpm at room tem-
    perature (22 ± 3°C).
    10.5 Pipette the solute solution into a container containing no adsorbent. Agitate this blank sample
    for 72 hours.
    10.6 Collect, preserve, and analyze an aliquot of the stock solute solution.
    10.7 Remove the containers at the designated times from the rotary extractor and record the solution
    temperature, pH, and any changes in the  adsorbent or solution. If handling anaerobic suspensions, re-
    turn the containers to a glove box or bag before opening the containers.
    10.8 Separate solid and liquid phases using centrifugation or filtration (section 5.2). Determine the
    electrical conductivity of an aliquot of each supernate. Collect and preserve aliquots of each super-
    nate in sufficient quantity for determining the solute concentration.
    10.9 Determine the rate of change in the solute concentrations at the various times by
                                  %AC =
                                                  - C2)
                                                       -,x100
where    %AC

             C2
                         percent change
                         concentration of the solute at time t, and
                         concentration of the solute after 1,24,48, or 72 hours.
                                                 83

-------
    10.10 The equilibrium time is defined as the minimum amount of time needed to establish a rate of
    change of the solute concentration equal to or less than 5% per a 24-hour interval (see chapter 13).

11  Construction of the Environmentally Conservative Isotherm (ECI)

    10) prSedures* be         ^"''68 ** ^ Soil:solution (sections 8 and 9> and equilibrium (section
   112 If the equilibrium time as determined in section 10.9 is equal to or less than 24 hours, use the
   data obtained from the soil:solution procedure to construct an ECI. However, if the equilibrium time is
   greater than 24 hours, the soil:solution ratio determination procedure must be repeated at the eauilib-

               ECI  6d ^ SeCti°n 1°'9' (Refert° °haPter 12 f°r diSCUSSi°n °f the advanta9es       °
   ilf hBT » !! SeCti0? 9 yjeld? a Single soil:solution ^io for nonionic solutes, select additional ratios
   that bracket the cacuated rat.o. Use a minimum of eight soilrsolution ratios selected from those listed
   in section 83. Evaluate these ratios as outlined in section 8.5. When volatile solutes are under study
   refer to section 6 for experimental considerations.                                            y'
      omtn  IT™?1 ,°f ^ Sta PuintS !° Construct an ECL Solution ratios resulting in 'ess than 1 0%
   adsorption of the solute should not be used to construct the ECI (refer to chapter 1 2 for justification)
   tiona? datT         ^^ "** °btained' vary the recommended soihsolution ratios to generate addi-

   1 1 .5  Using the data generated by the soihsolution procedure, calculate the amount of solute ad-
   sorbed per mass of adsorbent.

      1 1 .5.1  Determine the amount of solute adsorbed per mass of adsorbent by
      where  x/tn
               m
               C0

               C
               V
                        m

amount of solute adsorbed per unit mass of adsorbent,
mass of adsorbent (oven-dried basis) in grams added to reaction container
initial solute concentration (determinedanalytically) before exposure to
adsorbent,
solute concentration after exposure to adsorbent at equilibrium, and
volume of solute solution added to reaction container.
  alt iCi°SStrU?2; * a" ECI reqUJ-eS (1) an Vm value for each soil:solution ratio that meets the crite-
  ria in 11.2, and (2) the corresponding equilibrium concentration value (C) of the solute.
  11.7 Construction of an ECI

      1 1 .7.1 Using linear graph paper, plot the equilibrium concentration (C), log C, or C/x/m on the co-
      ordinate (x axis) and the corresponding x/m, log x/m, or C value as the
      (y axis). Refer to chapter 1 2 for an example.
                                11*1' Th9
                                                                           or a La^muir-type
     11.7.3 The linear expression of the Freundlich equation is


                                 log (x/m) = Kf+-\/n\ogC
                                            84

-------
      where   x/m  =   amount of solute adsorbed per unit mass of adsorbent,
               Kf  =   a constant,
              1/n  =   a constant (sometimes written as A/), and
                C  =   equilibrium concentration of solute after contact with adsorbent.

      A linear regression can be used to fit a curve through the data plotted in 11.5.1, where the inter-
      cept equals log K/and the slope equals 1/n. An example in which the Freundlich equation is used
      is given in chapter 14.
      11.7.4 A linear expression of the Langmuir-type equation is:
                                                1
                              C
                            •MM
      where  xlm
               KL
                M
                C
               x/m    KLM  M

amount of solute adsorbed per unit mass of adsorbent,
a constant,
a constant, and
equilibrium concentration of the solute after exposure to adsorbent.
      A linear regression can be used to fit a curve through the data plotted in 11.7.1, where the inter-
      cept equals 1/Ki.M and the slope equals 1/M. Examples using Langmuir-type equations are given
      in chapter 14.
      11.7.5 Calculate the coefficient of determination (r2) of the regression. Examples are given in
      chapter 15.
      11.7.6 Use the equation that results in the coefficient of  determination value closest to 1.0 to gen-
      erate a curve through the data plotted in 11.7.1.
      11.7.7 The data plotted in 11.7.1 and the curve of best fit from 11.7.6 represent an ECI.

      11.7.8 Report the following information with the ECI:
        • temperature at which the tests were conducted,
        • pH and electrical conductivity (EC) of all solute solutions,
        • concentrations of stock (Co) and blank (Ce) solute solutions and the factor, if any, used to cor-
          rect data,
        • soil:solution ratios, corresponding quantity of solute  solution and mass of adsorbent, initial
          (Co) and final (C) solute concentration, and the percent of solute adsorbed,
        • %AC for each equilibration time,
        • equation for the line of best fit and corresponding r  value, and
        • complete description of the adsorbent

12 Construction of the Constant SoikSolution Ratio Isotherm (CSI)
   12.1 The CSI requires the initial solute concentration to vary and the mass of adsorbent to remain
   constant—unlike the ECI, which holds the initial concentration of the solute constant and varies the
   mass of adsorbent in each containers. (Chapter 12 presents advantages and limitations of both tech-
   niques.)
   12.2  Use the recommended soihsolution ratio, %A between 10% and 30% (section 8), and equilib-
   rium time, %AC < 5% per 24-hour interval (section 10), in the construction of a CSI.
   12.3  Weigh the adsorbent (mass prescribed by the soiksolution ratio) into clean, labeled containers.
   If handling anaerobic adsorbent-solute systems, conduct steps 12.3 to 12.5 in a glove box or glove
   bag.
                                               85

-------
  12.4 Make a series of approximately eight dilutions of the stock solute solution so that the series
  shows a progress.ve decrease in solute concentration. In the most dilute solution, the solute concen-
  Si10"-,? £U  !?8 s"fficje"iso that tne amount remaining in solution is above detection limits after con-
  tact with the adsorbent. The volume of each diluted solution necessary for construction of the CSI will
  depend upon the size of the reaction container used.

     12.4.1  The dilution of complex solutions may cause changes in pH and/or redox potential with
     the subsequent precipitation of the solute(s) (see chapter 11). Try to limit such reactions, or if not
     possible, use the procedures in section 11 to create adsorption isotherms.

  12.5 Immediately afterthe dilutions of the stock solute solution, pipette the diluted solutions into the
 containers holding the adsorbent. Each solution should have a corresponding container and the vol-
 ume of solution in all containers should be equal.

 12.6 Place the containers on the rotary extractor at 29 ± 2 rpm at room temperature (22 ± 3°C)  Aai-
   ^ n^ t'me determined in section 1 °- Collect and preserve aliquots of the stock solute solution
 and all dilutions using accepted techniques (e.g., Standard Methods for the Examination of Water'
 and Wastewater, American Public Health Association, 1985,16th ed., Washington,  DC,
 p. 38-45).

 12.7 After the agitation period, remove the containers from the rotary extractor and open If the sus-
 pensions are anaerobic, return the containers to a glove box or bag, then open the containers Ob-
 serve and record the solution temperature, pH, and any changes in the adsorbent or solution.

 12.8 Separate the solid and liquid phases using either centrifugation or filtration (section 5 2)  Deter-
 mine the electrical  conductivity of an aliquot of each supernate. Collect and preserve aliquots of each
 supernate of sufficient volume for the solute concentration determinations.

i!2'9, ??!(m'1? tlle solute concentration'" the stock solution, the dilute solutions before (C0 in equa-
w  I   , ,- and,aftei: (C!n equation 11 -5-1) exposure to the adsorbent. If significant  differences in the
blank solutions (section 8.5.11) are ascertained, the adsorption data must be corrected.

12.10 Using the data generated where the various solute concentrations were exposed to the same
mass of adsorbent, calculate the amount of solute adsorbed per mass of adsorbent (x/m)  Refer to
equation 11.5.1 for calculation of x/m.

12.11 Construction of the CSI requires (1) an x/m value for each solute concentration  and (2) the cor-
responding equilibrium concentration value  (C) of the solute.
12.12 For constructing the CSI, follow the same procedure and reporting requirements as the ECI
Refer to section 11.5 for directions on construction of the ECI/CSI.
                                           86

-------
                                       REFERENCES

Abernathy, J.R., and J.M. Davidson, 1971, Effect of calcium chloride on prometryne and fluometuron ad-
       sorption in soil: Weed Science, v. 19, p. 517-522.
American Society for Testing and Materials, 1979, Proposed methods for leaching of waste materials: An-
       nual Book of ASTM-A Standards, Part 31, Water, Pennsylvania, p. 1258-1261.
American Society for Testing and Materials, 1986a, Determination of hydrolysis rate constants of organic
       chemicals in aqueous solutions: Annual Book of ASTM Standards, Water and Environmental
       Technology, v. 11.04, Pennsylvania, p. 572-577.
American Society for Testing and Materials, 1986b, Conducting aqueous photolysis tests: Annual Book of
       ASTM Standards, Water and Environmental Technology, v. 11.04, Pennsylvania, p. 578-587.
American Society for Testing and Materials, 1987,24-Hour batch-type measurement of contaminant sorp-
       tion by soils and sediments: Annual Book of ASTM Standards, Water and Environmental Technol-
       ogy, v. 11.04, Pennsylvania, p. 163-167.
American Society for Testing and Materials, 1988, Determining a sorption constant (Koc) for an organic
       chemical in soil and sediments: Annual Book of ASTM Standards, Water and Environmental
       Technology, v. 11.04, Pennsylvania, p. 731-737.
Anderson, M.A., C. Bauer, D. Hansmann, N. Loux, and R. Stanforth, 1981, Expectations and limitations
       for aqueous adsorption chemistry, in M.A. Anderson and A.J. Rubin [eds.], Adsorption of Inorgan-
       ics at Solid-Liquid Interfaces, Ann Arbor Science Publishers, Michigan, p. 327-347.
Ashton, P.M., and T.J. Sheets, 1959, The relationship of soil adsorption of EPTC to oats injury in various
       soil types: Weeds, v. 7, p. 88-90.
Atkins, P.W., 1982, Physical Chemistry: W.H. Freeman and Company, New York (pages 1012 to 1030
       briefly review adsorption processes, and various sections discuss the significance of ionic activity
        in solution).
Bailey, G.W., and J.L. White, 1970, Factors influencing the adsorption, desorption, and movement of pes-
       ticides in soil: Residue Reviews, v. 32, p. 29.
Banerjee, S., S.H. Yalkowsky, and S.C. Valvani, 1980, Water solubility and octanol/water partition coeffi-
        cients of organics. Limitations of the solubility-partition coefficient correlation: Environmental Sci-
        ence and Technology, v. 14, p. 1227-1229.
Barrow, N.J., 1972, Influence of solution concentration of calcium on the adsorption of phosphate, sul-
       ' phat'e, and molybdate by soils: Soil Science, v. 113, p. 175-180.
Barrow, N.J.,  1978, The description of phosphate adsorption curves: Journal of Soil Science, v. 29, p. 447-
       ' 462.
 Barrow, N.J.,  P.G. Ozanne, and T.C. Shaw, 1965, Nutrient potential and capacity. I. The concepts of nutri-
        ent potential and capacity and their application to soil potassium and phosphorus: Australian Jour-
        nal of Agricultural Research, v. 16, p. 61-76.
 Barrow, N.J.,  and T.C. Shaw, 1975, The slow reactions between soil and anions. 2. Effect of time and
       'temperature on the decrease in phosphate concentration in the soil solution: Soil Science, v. 119,
        p. 167-177.
 Barrow, N.J.,  and T.C. Shaw, 1979, Effects of soihsolution ratio and vigour of shaking on the rate of phos-
        phate adsorption by soil: Journal of Soil Science, v. 30, p. 67-76.
 Bartlett, R., and B. James, 1980, Studying dried, stored soil samples—some pitfalls: Soil Society of Amer-
        ica Journal, v. 44, p. 721-724.
 Bar-Yosef, B., V. Kafkaki, and N. Lahav, 1969, Relationships among adsorbed phosphate, silica, and hy-
        droxyl during drying and rewetting of kaolinite suspension: Soil Science Society of America
        Proceedings, v. 33, p. 672-676.
 Bear, J., 1972, Dynamics of Fluids in Porous Media: American Elsevier, New York.
 Birch, H.F., 1958, The effect of soil drying on humus decomposition and nitrogen availability: Plant
        and Soil, v. 10, p. 9-31.
 Birch, H.F., 1959, Further observations on humus decomposition and nitrification: Plant and Soil, v. 10, p.
        9-31.
 Bittel, J.E., and R.J. Miller, 1974, Lead cadmium and calcium selectivity coefficients on montmorillonite, il-
         lite, and kaolinite: Journal of Environmental Quality, v. 3, p. 250-253.
                                                87

-------
          iW D1,9?/ M odeling ^e m°vement of chemicals in soils by water: Soil Science, v. 1 1 5, p. 224-230

         fcSh^S1' •    G^' £C°nn0r' 19?9' Soil Chemistry: John Wiley and Sons, New York
         (pages 28 to 32 give a brief discussion on ionic activity in solution)
  Bolt, G.H and M.G.M. Bruggenwert, 1978, Soil Chemistry. A. Basic Elements (2nd ed.): Elsevier Scien-
  R™m   £ Polishing Company, Amsterdam (chapter 2 rigorously treats ionic activity in solution)
  Bowman, B.T  1981 .Anomalies in the log Freundlich equation resulting in deviations in adsorption K val-
         ues of pesticides and other organic compounds when the system of units is changed: Journal of
         Environmental Science and Health, v. B1 6, p. 1 1 3-1 23
  Bowman, B.T., and W.W. Sans, 1985, Partitioning behavior of insecticide-soil-water systems- 1  Adsorb-
         ent concentration effects: Journal of Environmental Quality, v 14 p 265-269
         AdR' Kin9- 1984, Adsorption of labile organic compounds by soil: Soil Science, v.  137, p.
  ChakravaiU M N  and N I.R D. Dhar ,1927, Die Ableitung einer Adsorptions-gleichung aus Langmuirs
         TheonederRestvalenzen:Kolloid-Zeitschrift,v. 43, p. 377-386
  Ch.ou, C.T., LJ. Peters and V.H. Freed, 1979, A physical concept of soil-water equilibria for nonionic or-
         ganic compounds: Science, v. 206, p. 831 -832
  Chiou, C.T, P.E. Porter, and D.W. Schmeddling, 1983, Partition equilibria of nonionic organic compounds
         between iso.l organic matter and water: Environmental Science and Technology  v  17  p 227231

         ^^^^S^!^ °f Pentachloropheno1 adsorPtion by soils: Soil' Science and  '


  CH2M ^Swft^S^ffifffi vsTmour Recycling Corporation' u-a EPA Contract No-
  Chou, S F.J and R A Griffin, 1983, Soil, cla^, and caustic soda effects on solubility, sorption, and mobil-

        ?04 54epa    °Cy  °Pentadiene: "lin°iS ^^ Ge°logical Survey Environmental Geology No?es

       '                "80"' 198°' Advanced ln°rganic Che™stry (4th ed.): John Wiley and Sons,

                                                                       on solute transport in a

                                                                             ReSOurces
                            M' Reinhard' 1986' A natural gradient
        n'^'oH S;Sha? • lfDdl anlW'P- ,Liaa 1988' Transport of inorganic ^pounds through com-
        pacted clay soil, in Proceedings of the Fourteenth Annual Research Symposium- US Environ-
        mental Protection Agency, May 9-11,1 988, EPA/600/9-88,021 , p  1 1 4-1 25
        te'n?nd It1', Lavy/ J978' Atrazine adsorption on soil as influenced by temperature, moisture con-
        tent, and electrolyte concentration: Weed Science, v. 26  p 303-308
 Dav,dson J.M P.S.C Rao LT. Ou W.B. Wheeler, and D.F. Rothwell, 1980, Adsorption, movement,
        and biotogical degradation of large concentrations of selected pesticides in soils: U S Environ-
        mental Protection Agency, EPA-600/2-80-124 (NTIS:PB 81-1 1 1-056)
 Debano  L.F., S M. Savage and D.A. Hamilton, 1976, The transfer of heat and hydrophobic substances
       dunng burning: Soil Science Society of America Journal, v. 40 p 779-782        SUDStances

    Un'65 1988' The fate °f hazardous materials in soil (Part 2): Hazardous Materials Control, v. 1 , p. 41-


                                           adSOrpti°n °f P°'ycyclic aramatic Mrocarbons on


                                                        «ssr and sodium adsorption ra-
             and G' ?POoSi-?o1 ?81 ' T2ermodvnamic derivation of equations of the Langmuir type for ion
                in soils: Soil Science Society of America Journal v 45 p 277-282
Farmer, W.J., anc i Y Aochl, 1974, Picloram sorption by soils: Soil Science Society of America Proceed-
       ings, v. oo, p. 41o-4
-------
Fox, R.L., and P.G.E. Searle, 1978, Phosphate adsorption by soils of the tropics, in Diversity of Soils in
       the Tropics (chapter 7): American Society of Agronomy.
Freeze, R.A., and J.A. Cherry, 1979, Groundwater: Prentice-Hall, New Jersey (chapter 9 discusses
       groundwater contamination).
Freundlich, H., 1909, Kapillarchemie: eine Darstellung der Chemie der Kolloide und Verwandter Gebiete,
       Leipzig, Akademische Verlagsgesellschaft, 591 p.
Freundlich, H., 1930, Kapillarchemie (Band II), Leipzig, Akademische Verlagsgesellschaft, 955 p.
Frissel, M.J., and G.H. Bolt, 1962, Interaction between certain ionizable organic compounds (herbicides)
      ' and clay minerals: Soil Science, v. 94, p. 284-291.
Frost, A.A., and R.G. Pearson, 1961, Kinetics and Mechanism: John Wiley and Sons, New York.
Frost', R.R.', and R.A. Griffin, 1977, Effect of pH on adsorption of copper, zinc, and cadmium from landfill
       lea'chate by clay minerals: Journal of Environmental Science and Health, A12 (4 and 5),  p. 139-
       156.
Fujimoto, O.K., and D. Sherman, 1945, The effect of drying, heating, and wetting on the level of exchange-
       able manganese in Hawaiian soils: Soil Science Society of America Proceedings, v. 10,  p.
       107-112.
Gardner, B.R., and J.P. Jones, 1973, Effects of temperature on phosphate sorption isotherms and phos-
       phate desorption: Communications in Soil Science and Plant Analysis, v. 4, p. 83-93.
Garrels, R.M., and C.L. Christ, 1965, Solutions, Minerals, and Equilibria: Freeman, Cooper, and  Com-
       pany, California (chapter 2 is an excellent treatment of activity-concentration relations).
Gelhar, L.W., and G.J. Axness, 1981, Stochastic analysis of macrodispersion in 3-dimensionally heteroge-
       ' neou's aquifers:  Hydrologic Research Program, New Mexico Institute of Mining and Technology
        Report no. 8, Soccorrco, New Mexico.
Ghosh, K., and M. Schnitzer, 1980, Macromolecular structures of humic substances: Soil Science, v. 129,
       ' p. 266-276.
Gibb, J.P., and K. Cartwright, 1982, Retention of zinc, cadmium, copper, and lead by geologic materials:
        Illinois State Water Survey and Illinois  State Geological Survey Cooperative Groundwater Report
        9,113 p.
Griffin, R.A., and A.K. Au, 1977, Lead adsorption by montmorillonite using a competitive Langmuir equa-
        tion: Soil Science Society of America Journal, v. 41, p. 880-882.
Griffin, R.A., and S.F.J.  Chou, 1980, Attenuation of polybrominated biphenyls and hexachlorobenzene by
        earth materials: Illinois State Geological Survey Environmental Geology Notes 87, p. 53.
Griffin, R.A., A.K. Au, and R.R. Frost, 1977a, Effect of pH on adsorption of chromium from landfill-
        leachate by clay minerals: Journal of Environmental Science and Health, v; A12(8), p. 431-449.
Griffin, R.A., R.R. Frost, A.K. Au, G.D. Robinson, and N.F. Shimp, 1977b, Attenuation of pollutants in mu-
        nicipal landfill leachate by clay minerals: Part 2—Heavy-metal adsorption: Illinois State Geologi-
        cal Survey Environmental Geology Notes 79,47 p.
 Griffin, R.A., and J.J. Jurinak, 1973a, The interaction of phosphate with calcite: Soil Science Society of
        America Proceedings, v. 37, p. 847-850.
 Griffin, R.A., and J.J. Jurinak, 1973b, Estimation of activity coefficients from the electrical conductivity of
        natural aquatic  systems and soil extracts: Soil Science, v. 116, p. 26-30.
 Griffin, R.A., and W.R. Roy, 1985, Interaction of organic solvents with saturated soil-water systems: Envi-
        ronmental Institute for Waste Management Studies, Open File Report 3, University of Alabama,
        86 p.
 Griffin, R.A., W.A. Sack, W.R. Roy, C.C. Ainsworth, and I.G. Krapac, 1985, Batch-type 24-hour distribu-
        tion ratio for contaminant adsorption by soil materials, in D. Lorenzen et al. [eds.], Hazardous and
        Industrial Solid Waste Testing and Disposal, American Society for Testing and Materials, STP
        933, v.  6, p. 390-408.
 Griffin, R.A., and N.F. Shimp, 1976, Effect of pH on exchange-adsorption or precipitation of lead from
        landfill leachates by clay minerals: Environmental Science and Technology, v. 10, p. 1256-1261.
 Graver, R., and R.J. Hance, 1970, Effect of ratio of soil to water on adsorption of linuron and atrazine: Soil
       ' Science, v. 109, p. 136-138.
 Gschwend, P.M., and S. Wu, 1985, On the constancy of sediment-water partition coefficients of hydropho-
         bic organic pollutants: Environmental Science and Technology, v. 19, p. 70-96.
                                                 89

-------
  Halsey, G.D., and H.S Taylor, 1947, The adsorption of hydrogen on tungsten powders: Journal of Chemi-
         cai Hnysics, v. 15, p. 624.
  Hance, R.J., 1969, Influence of pH, exchangeable cation and the presence of organic matter on the ad-
         sorption of some herbicides by montmorillonite: Canadian Journal of Soil Science, v. 49  p.
         357-364.                                                                '
  Harada, Y., and K. Wada, 1974, Effects of previous drying on the measured cation and anion exchanqe
         capacities of Andosols: Tenth International Congress of Soil Science Transactions (Moscow)^.
         b^~
  Harris, C.I., and G.F. Warren, 1964, Adsorption and desorption of herbicides by soil- Weeds v 12 p  120
  Harter, R.D., and D.E. Baker, 1977, Applications and misapplications of the Langmuir equation to soil ad- '
         sorption phenomena: Soil Science Society of America Journal, v. 41 p 1 077-1 080
  Hassett, J.J., JC. Means, W.L Banwart, and S.G. Wood, 1980, Sorption properties of sediments and en-
         ergy-related pollutants: U.S. Environmental Protection Agency, EPA-600/3-80-041 , 147 p.
         (NTIS:PB 80-189-57A).
  Hassett, J.J., W.L Banwart, S.G. Wood, and J.C. Means, 1981 , Sorption of onaphthol: implications con-
         cerning the limits of hydrophobic sorption: Soil Science Society of America Journal, v. 45,
         p. 38-42.
  Hassett, J.J., W.L. Banwart, and R.A. Griffin, 1983, Correlation of compound properties with sorption char-
        acteristics of nonpolar compounds by soils and sediments: Concepts and limitations in C W
        Francis and S.K Auerback [eds.], Characterization, Treatment, and Disposal, Environment and
        Solid Wastes, Butterworth Publishers, Massachusetts, chapter 15 p 161-178
 Hayward, D.O., and B.M.W. Trapnell, 1964, Chemisorption (2nd ed): Butterworth Publishers, London
        (chapter 5 is devoted to adsorption isotherm equations)
 Helyar, K.R. D.N Munns, and R.G. Burau, 1976, Adsorption of phosphate by gibbsite. I. Effects of neu-
          307 31   Sa     calcium' ma9nesiurn, sodium, and potassium: Journal of Soil Science, v. 27,

 Hingston FJ A.M. Posner, and J.P. Quirk, 1968, Adsorption of selenite on goethite, in W.J. Weber and
        E. Matijevic [eds.], Adsorption from Aqueous Solution, Advances in Chemistry Series 79 Ameri-
        can Chemical Society, p. 82-90.
 Hope, G D  and J.K. Syers, 1976, Effects of solution:soil ratio on phosphate sorption by soils: Journal of
        ooii ocience, v. 27, p. 301 -306.
 Horvath, C., W  Melander, and I. Molnar, 1976, Solvophobic interactions in liquid chromatography with
        nonpolar stationary phases: Journal of Chromatography, v. 1 25, p. 1 29-1 56
 Horzempa, L.M., and D.M. DiToro, 1983, PCB partitioning in sediment-water system: The effect of sedi-
        ment concentration: Journal of Environmental Quality, v. 12  p  373-380
 Huheey, J.E., 1978, Inorganic Chemistry: Principles of Structure and Reactivity. Harper and Row, New
        York, 889 p.
 Jones, J.P  B.B. Singh, M.A. Fosberg, and A.L. Falen, 1979, Physical, chemical, and mineralogical char-
        acteristics of soils from volcanic ash in Northern Idaho: II. Phosphate sorption: Soil Science Soci-
        ety of America Journal, v. 43, p. 547-552.
 Jurinak, J J., and N. Bauer,  1956, Thermodynamics of zinc adsorption on calcite, dolomite and maane-
       srte-type minerals: Soil Science Society of America Proceedings, v. 20 p 466-471
 Kenaga, E E., 1980, Predicted bioconcentration factors and soil sorption coefficients of pesticides and
       other chemicals: Ecotoxicology and Environmental Safety, v. 4, p. 26-38
 Kenaga, E.E., and A.I. Goring, 1980, Relationship between water solubility, soil sorption, octanol-water
       partitioning, and concentration of chemicals in biota: Aquatic Toxicology ASTM STP 707 J G  Ea
       ton' P;1?- Parrish- and A>C' Hendricks teds-]' American Society for Testing and Materials, '  '
       p* /o~il&.
Kinniburgh  D.G., 1986, General purpose adsorption isotherms: Environmental Science and Technology,
       v, £.U| p. oyo^yi/T1.
Kinniburgh  D.G., and M.L. Jackson, 1981 , Cation adsorption by hydrous metal oxides and clays in M A
       Anderson and A.J. Rubin [eds.], Adsorption of Inorganics at Solid-Liquid Interfaces, Ann Arbor  '
       Science Publishers, Michigan, p. 91 -1 60 (this article reviews 337 publications on this topic)
Kipling, J.J., 1965, Adsorption from solutions of non-electrolytes: Academic Press London p 215-216
                                              90

-------
Koskinen, W.C., and H.H. Cheng, 1983, Effect of experimental variables on 2,4,5-T adsorption-desorption
       in soil: Journal of Environmental Quality, v. 12, p. 325-330.
Kuo  S., and D.S. Mikkelsen, 1979, Zinc adsorption by two alkaline soils: Soil Science, v. 128,  p. 274-279.
Kurtz, T., E.E. DeTurk, and R.H. Bray, 1946, Phosphate adsorption by Illinois soils: Soil Science, v. 61, p.
       111-124.
Laidler, K.J., 1965, Chemical Kinetics, McGraw-Hill, New York.
Langmuir, I., 1918, The adsorption of gases on plane surfaces of glass, mica, and platinum: Journal of the
       American Chemical Society, v. 40, p. 136-1403.
Larsen, S., and A.E. Widdowson,  1964, Effect of soil/solution ratio on determining chemical potentials of
       ' phosphate ions in soil solutions: Nature, v. 203, p. 942.
Lawrence, J., and H.M. Tosine, 1976, Adsorption of polychlorinated biphenyls from aqueous solution and
       sewage: Environmental Science and Technology, v. 10, p. 381-383.
Leo, A., C. Hansch, and D. Elkins, 1971, Partition coefficients and their uses: Chemical Reviews,  v. 71,
       p. 525.
Low, P.P., and C.A. Black, 1950, Reactions of phosphate with kaolinite: Soil Science, v. 70, p. 273-290.
Lueb's, R.E., G. Stanford, and A.D. Scott, 1956, Relation of available potassium to soil moisture: Soil Sci-
        ence Society of America Proceedings, v. 20, p. 45-50.
Luh, M.D., and R.A. Baker, 1970, Organic sorption from aqueous solution by two clays, in Proceedings of
        the 20th Industrial Waste Conference, Purdue University, Extension Series, v. 137, p.  534-542.
Mabey, W., and T. Mill, 1978, Critical review of hydrolysis of organic compounds in water under environ-
       ' mental conditions: Journal of Physical Chemistry, reference data, v. 7, p. 383-415.
McAuliffe, C., 1966, Solubility in water of paraffin, cycloparaffin, olefin, acetylene cycloolefin, and aromatic
        hydrocarbons: Journal  of Physical Chemistry, v. 70, p. 1267-1275.
McCall, P.J., 1981, Standard practice for determination of sorption constants in soil and sediments: Draft
       ' no. 8 submitted to ASTM committee E35.21, Environmental Chemistry Fate-Modeling (Sorption
        Task Force), 32 p.
McGlamery, M.D., and F.W. Slife, 1966, The adsorption and desorption of atrazine as affected by pH,
        temperature, and concentration: Weeds, v. 14, p. 237-239.
Miller, D.M., M.E. Sumner, and W.P. Miller, 1989, A comparison of batch and flow-generated anion iso-
        therms: Soil Science Society of America Journal, v. 53, p. 373-380.
Moreale, A., and R. Van Bladel, 1980. Behavior of 2,4-D in Belgian soils: Journal of Environmental Qual-
        ity, v. 9, p. 627-633.
Mortland, M.M., and K.V. Raman, 1968, Surface acidity of smectites in relation to hydration, exchange-
        able cation, and structure: Clays and Clay Minerals, v. 16, p. 393-398.
Murali, V., and L.A.G. Aylmore, 1983a, Competitive adsorption during solute transport in soils: 1. Mathe-
        matical models: Soil Science, v. 135, p. 143-150.
Murali, V., and L.A.G. Aylmore, 1983b, Competitive adsorption during solute transport in soils: 2. Simula-
        tions of competitive adsorption: Soil Science, v. 135, p. 203-213.
 Murali, V., and L.A.G. Aylmore, 1983c, Competitive adsorption during solute transport in soils: 3. A review
        of experimental evidence of competitive adsorption and an evaluation of simple competition mod-
         els: Soil Science, v. 136, p. 279-290.
 Nirmalakhandan, N.N., and R.E. Speece, 1988, Prediction of aqueous solubility of organic chemicals
         based on molecular structure: Environmental Science and Technology, v. 22, p. 328-338.
 Ogata, A., 1970, Theory of dispersion in a granular medium: U.S. Geological Survey Professional Paper
         411-1.
 Parfitt, R.L., 1978, Anion adsorption by soils and soil materials: Advances in Agronomy, v. 30, 50 p.
 Patten, O.K., J.M. Bremner, and A.M. Blackmer, 1980, Effects of drying and air-dry storage of soils on
         their capacity for denitrification of nitrate: Soil Science Society of America Journal, v.  44, p. 67-70.
 Persoff, P., and J. F. Thomas, 1988. Estimating Michaelis-Menten or Langmuir isotherm constants by
         weighted nonlinear least squares: Soil Science Society of America Journal, v. 52, p. 886-889.
 Pierce, M.L., and C.B. Moore, 1980, Adsorption of arsenite on amorphous iron hydroxide from dilute aque-
       '  ous'solution: Environmental Science and Technology, v. 14, p. 213-216.
                                                 91

-------
  Polyzopoulos, N.A., V.Z. Keramidas, and H. Kiosse, 1985, Phosphate sorption by some alfisols of Greece
         as described by commonly used isotherms: Soil Science Society of America Journal, v. 49 p.
          81-84.                                                                     '
  Ponec, V. 2. Knor, and S. Cerny, 1974, Adsorption on Solids: Chemical Rubber Company, Cleveland
         Ohio, 693 p.                                                                        '
  Rao, P.S.C., 1974. Pore-geometry effects on solute dispersion in aggregated soils and evaluation of a pre-
         dictive model: unpublished Ph.D. Dissertation, University of Hawaii.
  Raveh, A., and Y. Avnimelech, 1978, The effect of drying on the colloidal properties and stability of humic
         compounds: Plant and Soil, v. 50, p. 545-552.
  Reinbold, K.A., JJ. Hassett, J.C. Means, and W.L. Banwart, 1979, Adsorption of energy-related organic
         ESSSe3 llterature review: U'S- Environmental Protection Agency, Athens, GA, EPA-600/3-79-
         086 (NTIS:PB 80-105-117).
  Reftemeier, R.F, 1945, Effect of moisture content on the dissolved and exchangeable ions of soils in arid
         regions: Soil Science, v. 61, p. 195-214.
  Rideal, E.K., 1930, Surface Chemistry: Cambridge, Cambridge University Press.
  Roy, W.R., and R.A. Griffin, 1985, Mobility of organic solvents in water-saturated soil materials- Environ-
         mental Geology and Water Science, v. 7, p. 241 -247.
  Roy, W.R., J.J. Hassett, and R.A. Griffin, 1986, Competitive interactions of phosphate and molvbdate on
         arsenate adsorption: Soil Science, v. 142, p. 203-210.
  Roy, W.R., C.C. Ainsworth, R.A. Griffin, and I.G. Krapac, 1984, Development and application of batch ad-
         sorption procedures for designing earthen landfill liners, in Seventh Annual Madison Waste Con-
        ference, University of Wisconsin, Madison, Sept. 11-12,1984 p 390-398
 Roy, W.R R.A. Griffin, S.F.J. Chou, C.C. Ainsworth, and I.G. Krapac, 1985, Development of stand-
        ardized batch adsorption procedures: experimental considerations, in Proceedings of the Elev-
        enth Annual Research Symposium on Land Disposal of Hazardous Waste, Cincinnati OH Aoril
        29 - May 1,1985, U.S. EPA-600/9-85-013 (NTISPB 85-196-376).                      '
 Roy, W.R., J.J. Hassett, and R.A. Griffin, 1989, Quasi-thermodynamic basis of competitive-adsorption in-
        teractions for anionic mixtures in soils: Journal of Soil Science, v. 40, p. 9-15.
 Ryden, J.C., and J.K. Syers, 1975, Rationalization of ionic strength and cation effects on phosphate sora-
        tion by soils: Journal of Soil Science, v. 26, p. 395-406.
 Ryden, J.C  J.K. Syers, and J.R. Mclaughlin, 1977, Effects of ionic strength on chemisorption and poten-
        tial-determining sorption of phosphate by soils: Journal of Soil Science v 28  p 62-71
 Schwarzenbach, R.P., and J. Westall, 1981, Transport of nonpolar organic compounds from surface
        water to groundwater: Laboratory studies: Environmental Science and Technology, v. 15, p.
        i ooO~i 367.
 Scott, H.D., D.C. Wolf, and T.L. Lavy, 1982, Apparent adsorption and microbial degradation of phenol bv
        soil: Journal of  Environmental Quality, v. 11, p. 107-111.
 Sinanoglu, p., and S. Abdulnur, 1965, Effect of water and other solvents on the structure of biopolymers-
        Federation Proceedings, v. 24, part III, p. 512-523.
 Singh, B.R 1984, Sulfate adsorption by acid forest soils: 1. Sulfate adsorption isotherms and comparison
    t,  D  different adsorption equations in describing sulfate adsorption: Soil Science, v 138 p  189-197
 Singh, B.B and J.P Jones, 1977, Phosphorus sorption isotherm for evaluating phosphorus requirements
        of lettuce at five temperature regimes: Plant and Soil, v. 46, p. 31-44.
 Sips, R., 1948, On the structure of a catalyst surface: Journal of Chemical Physics v  16 p 490-495
 Souhdes, DA, and F.E. Allison, 1961, Effect of drying and freezing soils on carbon dioxide production
        available mineral nutnents, aggregation, and bacterial population: Soil Science, v 91 p 291-298
 Sposito, G, 1979, Derivation of the Langmuir equation for ion exchange reactions in soils- Soil Science
        Society of America Journal, v. 43, p. 197-198.
 Sposito, G. 1980, Derivation of the Freundlich equation for ion exchange reactions in soils- Soil Science
        Society of America Journal, v. 44, p. 652-654.
Sposito, G. 1982, On the use of the Langmuir equation in the interpretation of "adsorption" phenomena-
        II. The "two-surface" Langmuir equation: Soil Science Society of America Journal, v. 46, p.
        I IT"/"" I I Ob«
                                              92

-------
Sposito, G., K.M. Holtzclaw, L. Charlet, C. Jouany, and A.L. Page, 1983, Sodium-calcium and sodium-
       magnesium exchange on Wyoming bentonite in perchlorate and chloride background ionic me-
       dia: Soil Science Society of America Journal, v. 47, p. 51-56.
Steel, R.G.D., and J.H. Torrie, 1960, Principles of Procedures of Statistics: McGraw-Hill, New York.
Stevenson, F.J., 1982, Humus Chemistry: John Wiley and Sons, New York, 443 p.
Stevenson, I.L., 1956, Some observations on the microbia! activity in remoistened air-dried soils: Plant
       and Soil, v. 8, p. 170-182.
Stumm, W., and J.J. Morgan, 1981, Aquatic Chemistry, John Wiley and Sons (2nd ed.), New York (vari-
       ous sections discuss the significance of ionic activity in solution).
Suffet, I.H., and M.J. McGuire [eds.], 1980, Activated Carbon Adsorption of Organics from the Aqueous
       Phase, v. 1, Ann Arbor Science Publishers, Michigan, 508 p.
Taylor, R.W., and E.G. Ellis, 1978, A mechanism of phosphate adsorption on soil and anion exchange
       resin surfaces: Soil Science Society of American Journal, v. 42, p. 432-436.
Thomas, J.M., 1961, The existence of endothermic adsorption: Journal of Chemical Education, v. 38, p.
       138-139.
Tinsley, I.J.,  1979, Chemical Concepts in Pollutant Behavior: John Wiley and Sons,  New York. U.S. Envi-
       ronmental Protection Agency, 1982, Chemical fate test guidelines, EPA-560/6-82-003.
Van Genuchten,  M.T., P.J. Wierenga, and G.A. O'Connor, 1977, Mass transfer studies in sorbing porous
       media: III. Experimental evaluation with 2,4,5-T: Soil Science Society of America Journal, v. 41,
       p. 278-285.
Van Lierop, W., and A.F. MacKenzie, 1977, Soil pH measurement and its application to organic soils: Ca-
        nadian Journal of Soil Science, v. 57, p. 55-64.
Veith, J.A., and G. Sposito, 1977, On the use of the Langmuir equation in the interpretation of "adsorp-
       tion" phenomena: Soil Science Society of America Journal, v. 41, p. 697-702.
Voice, T.C.,  and W.J. Weber, 1983, Sorption of hydrophobic compounds by sediments, soils and sus-
        pended solids. I. Theory and background: Water Research, v. 17, p. 1433-1441.
Voice, T.C.,  and W.J. Weber, 1985, Sorbent concentration effects in liquid/solid partitioning: Environ-
        mental Science and Technology, v. 19, p. 789-796.
Voice, T.C.,  C.P. Rice, and W.J. Weber, 1983, Effect of solids concentration on the sorptive partition of hy-
        drophobic pollutants in aquatic systems: Environmental Science and Technology, v. 17, p.
        513-518.
Weber, J.B,  1966, Molecular structure and pH effects on the adsorption of 13 s-triazine compounds on
        montmorillonite clay: American Mineralogists, v. 51, p. 1657-1670.
Weber, W.J., T.C. Voice, M. Pirbazari, G.E. Hunt, and D.M. Ulanoff, 1983, Sorption of hydrophobic com-
        pounds by sediments, soils, and suspended solids. II. Sorbent evaluation studies: Water Re-
        search, v. 17, p. 1443-1452.
White, R.E., 1966, Studies of the phosphate potentials of soils. IV. The mechanisms of the "soil/solution
        ratio effect": Australian Journal of Soil Research, v. 4, p. 77-85.
White, R.E., 1980, Retention and release of phosphate by soil and soil constituents: Soils and Agriculture,
        v.2, p.71-114.
Yaron, B., and S. Saltzman, 1972, Influence of water and temperature on adsorption of parathion by soils:
        Soil Science Society of America Proceedings, v. 36, p. 583-856.
Zeldowitsh,  J., 1935, On the theory of the Freundlich adsorption isotherm: Acta Physicochimica U.R.S.S.,
        v. 1, p. 961-974.
Zettlemoyer, A.C., and F.J. Micale, 1971, Solution adsorption thermodynamics for organics on surfaces,
        in S.D. Faust and J.V. Hunter [eds.], Organic Compounds in Aquatic Environments, Marcel
        Dekker, New York, p. 165-185.
 Zierath, D.L, J.J. Hassett, W.L. Banwart, S.G. Wood, and J.C. Means, 1980, Sorption of benzidine by
        sediments and soils: Soil Science, v. 129, p. 277-281.
                                               93

-------

-------
                                        APPENDIX A

            SUMMARY AND CHEMICAL COMPOSITION OF THE ADSORBENT SOILS
                              AND CLAYS USED IN THIS STUDY

Eleven different soil materials were used as adsorbents during the development of the batch-adsorption
procedures. Adsorbents and sample locations are summarized in table A1; relevant physicochemical char-
acteristics and mineralogy are summarized in table A2. Also characterized is the chemical composition of
the materials: major elements in table A3 and trace constituents in table A4.

The eleven clays and soils represented a wide range in physicochemical properties and characteristics.
   •  The Catlin soil is a dark prairie soil (Mollisol) with a relatively high organic matter content in the sur-
     face horizon. The clay-size fraction is predominantly illite. The Catlin, an important soil for agricul-
     ture, is one  of the Mollisols that dominate in the Great Plains.
   •  The two Cecil soils, Tifton and soil EPA-14, are Ultisols—highly weathered and acidic soils that are
     dominated by kaolinite, and iron and aluminum hydroxides. Most of the soils in the southeastern part
     of the United States are Ultisols.
   •  The Vandalia till, an Illinoian-age deposit, is fairly representative of midwestern glacial tills.  It is a
     sandy till, gray, and calcareous where unweathered; the predominate clay is illite.
   •  At the sampling site (table A1), the Sangamon Paleosol was a buried soil that had formed in the Van-
     dalia till and was overlain by glacial loess. The Sangamon Paleosol, Vandalia (ablation phase), al-
     tered (oxidized) Vandalia, and unaltered (unoxidized) Vandalia tills are a common stratigraphic
     sequence in Illinois. This sequence is also present at the Wilsonville hazardous-waste site at Wilson-
     ville, Illinois.

The soil sample designated as EPA-14 was used by Hassett et al. (1980,1981) and Zierath et al. (1980)
in studies concerned with the adsorption of hydrophobic solutes. The Cecil clay sample from South Caro-
lina was used by Roy et al. (1986) in a study concerned with the adsorption of anionic mixtures. The kao-
linite and illite clay samples have also been used in previous studies (Griffin and Shimp, 1976; Frost and
Griffin, 1977).

        Table A1  Summary of adsorbents        	
        Adsorbent
Sample location
Soil horizon
Classification
Catlin silt loam
Cecil clay
Cecil clay loam
EPA-14
Illite
Kaolinite
Sangamon Paleosol

Tifton loamy sand
Vandalia till member
ablation

altered
unaltered
Champaign, Illinois
Spartanburg, South Carolina
Cecil, Georgia
Ceredo, West Virginia
Elizabeth, Illinois
Pike County, Illinois
Macoupin County, Illinois
(near Sawyerville)
Tifton, Georgia
Glasford Formation
Macoupin County, Illinois
(near Sawyerville)
(near Sawyerville)
(near Eagerville)
AI
B2,
Ap
A
—
—
Bt

Ap

B3

C2
C4
Typic Argiudoll
Typic Hapludult
Typic Hapludult
unknown
—
—
unknown

Plinthic Paleudult

—

—
—
                                              95

-------
 Table A2  Selected physicochemical characteristics of clays and soils used in developing this TRD
Adsorbent
Catlin silt
loam
Cecil clay


Cecil clay
Innm
loam
EPA-14
Illito clay

Kaolinite clay
Sangamon
Paleosol
Tifton loamy
sand
Vandalia till
altered
unaltered
ablation
phase
PH*
6.1

4.5


4.6

4.5
7.9

8.1
6.1

4.7
7.4
7.5

6.4
%
Sand
11

31


32

2
0

0
45

85
45
45

56
%
Silt
69

12


17

63
0

0
25

9
38
40

21
% %Organic
Clay carbon
21

58


51

34
100

100
30

5
17
15

23
4.04

0.34


ND§

0.48
1.81

0.51
0.10

ND
0.18
0.34

0.10
CECf
(meq/
100 g)
18.1

3.7


3.8

18.9
20.5

15.1
16.7

1.9
6.6
4.9

10.5
Surface
area$
(m2/g)
14.8

36.9


29.7

145fl
ND

34.2
22.9

1.7
7.3
5.6

10.6
Clay analysis (%)
Illite
55-67

<5


5-6

13
70

8
33-36

0
71-77
75-82

32-58
Kaol-
inite
5-15

68-92


79-92

37
0

87
7-14

73-96
3-10
4-19

2-6
Expand-
ables
24-30

3-32


2-16

14
0

5
50-60

4-27
18-19
6-9

32-29
Other
clay-
sized
minerals
chlorite

gibbsite,
goethite,
hematite
goethite,
hematite
gibbsite
30% mixed
layer
quartz


goethite



goethite
* pH of a 1:1 soikwater suspension.
t Catkin exchange capacity.
i Surface area by N2 adsorption using BET method.
§ No data available.
fl Surface area by ethylene glycol (from Hassett et al., 1981).
 Table A3  Major element composition  (in oxide form) of clay and soils used in developing this TRD
 Adsorbent
Si02    Ti02    AI203   Fe203  CaO   MgO   Na.,,0   K2O    P2O5
Catlin silt loam
Cecil clay
Cecil clay loam
EPA-14*
Illite clayf
Kaolinite clayf
Sangamon Paleosol
Tifton loamy sand
Vandalia
altered
unaltered
ablation phase
72.5
44.8
66.2
ND
48.5
46.6
82.7
96.4

61.3
59.1
83.5
0.73
1.15
0.94
ND
0.67
2.45
0.43
0.27

0.33
0.33
0.35
10.8
30.0
20.4
ND
24.6
41.9
10.2
1.3

6.7
6.5
7.9
4.0
10.4
6.8
6.99
4.11
0.94
2.9
0.5

2.1
2.4
2.5
0.9
<0.1
<0.1
0.71
3.27
0.57
0.50
<0.1

9.4
9.7
0.6
0.71
0.19
0.19
ND
1.73
0.30
0.65
0.03

4.66
4.95
0.54
0.84
0.07
0.04
0.21
0.14
0.13
0.45
0.01

0.52
0.49
0.56
2.14
0.54
0.65
2.94
10.23
1.49
1.49
0.05

2.03
2.08
1.98
0.1
0.1

ND
ND
ND
<0.1
<0.1

<0.1
<0.1
<0.1
 * Data from Hassett et al., 1981.
 t Data from Griffin and Shimp, 1976.
                                                 96

-------
Table A4  Trace element concentrations in the clays and soils used in developing this TRD

As
B
Ba
Be
Br
Cd
Ce
Cr
Co
Cu
Cs
Eu
Ga
Hf
La
Li
Lu
Mn
Ni
Pb
Rb
Sb
Sc
Se
Sm
Sr
Ta
Tb
Th
U
W
Yb
Zn
Catlin
S.I.
10
250
721
3
8
<1
62
73
14
20
4
1
10
12
36
29
0.6
834
<8
20
82
1
9
<2
6
90
1
1
8
5
2
3
88
Cecil
c.
40
30
117
1
19
<1
123
206
6
45
10
1
37
8
63
29
0.4
—
70
44
59
1
25
3
8
<5
2
1
24
7
5
3
40
Cecil
C.I.
4
25
166
2
3
<1
81
73
3
17
5
1
26
18
47
18
0.6
93
<8
14
74
0.4
13
<2
8
<5
2
1
16
7
2
3
37
EPA- Illite Kaol-
14* c. inite c.
10 — t —
— 44 46
450 — —
— — —
_ _ _
— 19 <3
87 — —
— — —
11 — —
— — —
8 — —
•j _ __
23 — —
8 — —
46 — —
— — —
— — —
216 <390 29
— — —
— 94 46
200 — —
6 — —
16 — —
2 — —
— — —
<80 — —
1 	 ^_
1 — —
•]g 	 	
— — —
— — —
3 — —
— 38 20
Sang-
amon
6
230
500
2
<1
<1.3
38
52
15
11
3
1
12
—
28
27
0.5
970
<9
19
79
0.6
8
<1
5
55
—
—
5
<3
—
2
71
Tifton .
S.I.
1
170
44
<0.5
2
<1.3
50
20
1
<4
2
0.4
2
26
19
4
0.4
90
<9
<10
18
0.3
4
<2
3
<5
1
1
6
1
<1
2
<2

altered
6
172
359
2
<7
<1.3
25
39
8
15
3
1
9
—
19
23
0.3
388
<9
F14
68
0.4
6
<2
3
75
—
—
4
<3
—
2
42
Vandalia till
unaltered
7
150
347
2
3
<1.3
24
41
9
19
3
1
7
—
19
25
0.3
<400
<9
13
68
0.4
6
<1
3
75
—
—
4
<2
—
2
73

ablation
5
250
460
2
<2
<1.3
29
43
8
12
3
1
8
—
24
22
0.4
352
<9
<9
88
0.3
7
<1
4
62
—
—
4
<2
—
2
44
* Data from Hassett et al., 1981.
t No data available.
                                              97

-------

-------
                                       APPENDIX B

                     COMPOSITION OF THE METALLIC WASTE EXTRACT
              USED IN THIS STUDY AND ASSOCIATED ADSORPTION ISOTHERMS

The basic batch-adsorption procedure for ionic solutes was tested and refined through the use of a metal-
lic waste sample collected November 1,1984, from the Sandoval Zinc Company near Sandoval, Illinois.
Grab samples were taken from a dry slurry lagoon that was used to store metallic scrubber sludges (Gibb
and Cartwright, 1982). Samples were taken from the surface and at a depth of about 1 m, then compos-
ited and air-dried. The relatively fine-grained material was then mixed and poured through  a 2-mm sieve.

The laboratory work began by making 20 L of an extract of the metal-rich waste using the ASTM-A water-
shake extraction procedure (ASTM, 1979). The aqueous extract contained about 0.05% Zn (table B1) and
lesser quantities of Ba, Ca, K, and Pb. The extract was slightly acidic (pH 6.27) and was used as the
stock solution for all of the adsorption experiments. The Sangamon Paleosol sample and the Cecil clay
were selected for study because these two soils represented widely different physicochemical materials.
Adsorption isotherms (figs. B1 to B3) were generated using the procedures described in this volume.


             Table B1  Chemical constituent concentrations (mg/L) obtained by ASTM-A
             water shake extraction performed on the Sandoval zinc slurry
pH 6.27
EC(dS/m) 0.17
Al <0.05
As <0.08
B <0.08
Ba 2.25
Be <0.01
Ca 17.7
Cd 0.45
Co
Cr
Cu
Fe
K
Mg
Mn
Mo
Na
<0.45
<0.08
<0.01
<0.05
6.57
0.89
0.64
<0.02
<0.75
Ni
P
Pb
Sb
Se
Si
Sn
V
Zn
0.12
<0.05
15.0
<0.05
<0.04
<0.20
<0.03
<0.08
550
                  20-
                  10-
                       ./
                                 0.5
                                              1
                                             1.0
 r
1.5
 I
2.0
                                 Equilibrium barium concentration (mg/L)
             Figure B1  Barium adsorption isotherm at 21 °C with the Sangamon Paleosol from
             the metallic waste extract. The average pH of the soil-solute suspensions was 5.6.
                                             99

-------
      1.2-
      1.0-
  „  0.8-
   8  0.6-
      0.4-
      0.2-
                                              Sangatnon Paleosol
                                                          Cecil clay
                  ~I	1	1	1	T
                   12345
                        Equilibrium lead concentration (mg/L)
Figure B2  Lead adsorption isotherms at 24°C of two soils using the metallic waste
extract. The average pH of the Sangamon Paleosol suspensions was 5.6, and pH
4.3 for the Cecil clay.
 0              100             200            300             400            500
                          Equilibrium zinc concentration (mg/L)

 Figure B3  Zinc adsorption isotherms at 24°C of two soils using the metallic waste
 extract. The average pH of the Sangamon Paieosol suspensions was 5.9, and pH
 4.3 for the Cecil Clay.
                                     100
•U.S. Government Printing Office: 1992— 648-003/41820

-------

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati, OH 45268
      BULK RATE
POSTAGE & FEES PAID
          EPA
   PERMIT No. G-35
Official Business
Penalty for Private Use, $300
                                                                   Please make all necessary changes on the above label,
                                                                   detach or copy, and return to the address In the upper
                                                                   left-hand comer.

                                                                   II you do not wish to receive these reports CHECK HERE Q ;
                                                                   detach, or copy this cover, and return to the address In the
                                                                   upper left-hand comer.
                                                               EPA/530/SW-87/006-F

-------