United States
                        Environmental Protection
   Off ice of Water
   Regulations and Standards
   Criteria and Standards Division
   Washington DC 20460
August  1987
SCD# 12
vvEPA
                        Water
                    SEDIMENT QUALITY CRITERIA FOR METALS: IV.
              SURFACE COMPLEXATION  AND  ACIDITY CONSTANTS FOR
          MODELING  CADMIUM AND ZINC ADSORPTION ON TO IRON OXIDES



Schematic of
Surface Species




Schematic of —
Charge- 2
Potential |
Relationship 5
^^•M^BB^^^H

X-
X-
X-
X-
X-
X-
X-
X-

0
*0
fa
••••••i
I 1
1
-0-H"
•a-o"
•0 - Hj"
-0 • ( )' - Cd2*
-0 • ( f - CdOH'
•0 • H«
-0 - ( f - Na*
•0 • Ha* - Cl"


—

Ci
I"— 1
. — J

Ca
1 	 1
^


•— *Kaa
• *§Kcd Constants in Text
• ••KcdOH ' Corresponding to
Surface Species
^^^^™ Kne
^^^•^ *^Kfy



""i~"l">>r^^
WQ


h
                                                  Diffuse Layer
                                                  of Counter Ions
                                        Distance trom Surface (i)
                          Schematic Representation of Surface Species and Surface Charae-
                          Potentlal Relationship for the Triple-Layer Model. Brackets In
                          the  zero1 plane Indicate deprotonated surface sites. (After
                          Felmy. Glrvln and Jenne 1984)

-------
       SEDIMENT QUALITY CRITERIA FOR METALS:
IV.  SURFACE COMPLEXATION AND ACIDITY CONSTANTS FOR
MODELING CADMIUM AND ZINC ADSORPTION ONTO IRON OXIDES
            Work Assignment 56, Task 4


                    August 1987



                    Prepared by:
                 Everett A. Jenne
       Battelle, Pacific Northwest Laboratory
               Richland,  Washington
                        for:

        U.S.  Environmental  Protection Agency
          Standards and Criteria Division
                 Washington,  D.C.
                   Submitted by:

                      BATTELLE
      Washington Environmental  Program Office
                 Washington, D.C.

-------
                                ABSTRACT

     This report 1s one of a series of reports that collectively describe
an approach for establishing sediment criteria for metals.  The criteria
will be established by using surface complexation constants in conjuction
with estimates of adsorbed metals and the quantities of important
adsorbents to estimate the equilibrium activity of metals in the pore
water of sediments.  This estimate of metal  activity in pore water will
then be used to evaluate the potential toxicity of contaminated sediments
to benthic organisms.
     Surface complexation models quantify the adsorption process and the
effects of the electrostatic potential that result from the reaction of
charged surfaces with aqueous solutions.  However, these adsorption
constants are only valid for the model for which they are calculated.
Thus, to select a set of constants for the adsorption of Cd and Zn onto Fe
oxides, the model must be selected first.  The members of this family of
models differ primarily 1n the way that they describe the electrostatics
of the so!1d:water Interface.  Fortunately,  earlier surface complexation
model formulations have evolved to the point that either a constant
capacitance or triple-layer model 1s commonly used 1n current research.
The triple-layer model appears, most useful for developing sediment
criteria for metals because the sorptlon constants derived for this model
are Independent of solution pH and Ionic strength.  To the extent that
surface complexation constants are available for other competitive metals
and electrolytes, surface complexation constants for metals are largely
Independent of the Inorganic composition of the aqueous phase.
     Adsorption constants for the adsorption of Cd and Zn onto Fe oxides
calculated using the triple-layer model have been compiled.  These
sorptlon data are sparse; however, constants for the (SOCd+), (SOZn+),
(SOCdOH0), and (SOZnOH0) surface complexes were found for both amorpMc Fe
oxide and geothlte (a-FeOOH).  Differences of a factor of 10 or more occur
in the constants derived by different investigators for a given surface
complex on the same adsorbent.  Recalculation of these constants using a
common set of auxiliary constants and a common optimization scheme may
significantly reduce the variability of these compiled constants.
                                     111

-------
                                CONTENTS





ABSTRACT	 111



INTRODUCTION 	 1



SURFACE COMPLEXATION MODELS 	 3



TRIPLE-LAYER MODEL 	 9



SURFACE COMPLEXATION AND ACIDITY CONSTANTS 	  12



OTHER DATA NEEDED	 16



DISCUSSION 	 17



REFERENCES 	 19

-------
                 SEDIMENT QUALITY CRITERIA FOR METALS:
           IV. SURFACE COMPLEXATION AND ACIDITY CONSTANTS FOR
           MODELING CADMIUM AND ZINC ADSORPTION ONTO IRON OXIDES

                              INTRODUCTION

     The Criteria and Standards  Division of the U.S.  Environmental
Protection Agency 1s developing  sediment quality criteria for both
nonpolar organic contaminants  and metals.  These criteria will  be used in
conjunction with water quality criteria to protect aquatic organisms and
the food chain to man 1n both  freshwater and saltwater bodies.   This new
approach to developing sediment  quality criteria for  metals 1s  based on
calculating the thermodynamlc  activity of the uncomplexed metal  1n  the
sediment pore water and relating metal  activity to metal  concentrations
that produce toxic responses (Jenne et al. 1986).   The latter are Inferred
from the water quality criteria  for the metal.  The proposed approach
assumes that the activity of metals 1n pore water 1s  supported by the
quantity of sorbed metal and requires only that the quantity of each
Important absorbent (sorptlon  "sink") and the quantity of sorbed metals be
determined.  These data, plus  surface adsorption constants for the
absorbents, are used along with  an appropriate algorithm (model) to
estimate the activities of metals 1n pore water.  This approach separates
the problems of determining metal availability and evaluating toxldty,
and relates sediment quality criteria to the water quality criteria.
     This report 1s one 1n a series of five reports to review the overall
rationale for the approach, the extraction methods for estimating
quantities of each of the three major adsorbents in sediments and the
quantity of sorbed metals^3',  the available sorption  data for reactive
(a)  Jenne, E. A.  1987.  Sediment Quality Criteria for Metals: II. Review
of Methods for Quantitative Determination of Important Adsorbents and
Sorbed Metals in Sediments.  Submitted by Battelle, Washington
Environmental Program Office, Washington, D.C. to the U.S. Environmental
Protection Agency, Criteria and Standards Division.

-------
partlculate organic carbon', and to recalculate selected adsorption data
for cryptocrystalline Mn dioxide^  .  In this report, selected acidity
constants and Cd and Zn adsorption constants for iron oxides are given.
(a)  Allen, H. E., and J. M. Mazzacone.  1987.  Sediment Quality Criteria
for Metals; III. Review of Data on the Complexation of Trace Metals by
Particulate Organic Carbon.  Drexel University.  Submitted by Battelle,
Washington Environmental Program Office, Washington, D.C. to the U.S.
Environmental Protection Agency, Criteria and Standards Division.


(b)  D1 Toro, D. M., and B. Wu.  1987.  Sediment Quality Criteria for
Metals; III. Review of Data for Determining the Intrinsic Adsorption
Constants for Manganese Dioxide.  HydroQual Co., Inc.  Submitted by
Battelle, Washington Environmental Program Office, Washington, D.C. to the
U.S. Environmental Protection Agency, Criteria and Standards Division.

-------
                       SURFACE CQMPLEXATIQN MODELS

     The use of surface complexatlon models (SCMs) and the determination
of associated adsorption constants are active areas of research.  Current
SCMs combine a conceptualization of adsorption as a complexatlon reaction
with modern versions of the electrical double layer (Stern 1924; James and
Healy 1972; Yates, Levine and Healy 1974; Schindler et al. 1976; Stumrn,
Hohl and Dalang 1976; Davis and Leckle 19785; and Stumm, Kummert and Sigg
1980).  In these models, solids are visualized as possessing an array of
chemically reactive oxygen moieties bound to peripheral metal or carbon
atoms of an oxide, silicate, or organic solid phase (Figure 1).  The
formalism [SOH0] 1s used to represent uncharged hydroxyl sites associated
with a structural metal 1on (S) located at the surface of sol Ids.  The
related protonated and deprotonated surface species are

          SOH2+(s) = SOHfl(s) + H+(e)                               (1)

          SOHB(s) = S0"(s) + H+(e)     -                            (2)

where (s) designates the solid surface and (e) Indicates that the ion is
located within the electrically charged region at the sol id:solution
interface.  The electrical double layer represents the plane between the
solid surface and the bulk solution (Figure 1).  The hydroxyl sites have
add/base properties that are characteristic of the solid and form
complexes with metal ions and with swamping electrolytes such as 0.1N
KN03, which are used to maintain a given Ionic strength in experimental
studies.
     These heterogenous adsorption reactions for hydrogen and metals are
treated in a manner analogous to aqueous homogenous reactions.  An
association (i.e., adsorption) constant can be calculated to describe the
adsorption process.  For reactions 1 and 2, the conditional equilibrium
acidity constants are, respectively,

-------






Schematic of
Surface Species









^^
Schematic of —
Charge- .2
Potential g
Relationship o
OL




X-
X-
X-
X-
X-
X-
X-
X-

i
1
-0-H°
- 0 - { f
-0-H2*
-0 - ( )" - Cd2*
-0 - ( )" - CdOH*
-0-H°
-0 - ( )" - Na*
-0-H2*-Cf
1 1
1 1
1 1
A 1 1
I



00


0b
ti/ei


1 Ci 1 C2

1 1
1 1

I \ 1
1 \J




s ^
s
« 	 »3Kcd
^ 	 "8KcdOH '

•*— *sKNa
••— *SKCI












Illlll^s^
1 1 T^^^
1 1 1 ^*^*
                                                             Constants in Text
                                                             Corresponding to
                                                             Surface Species
                                 (To
" U
 h
                                                  Diffuse Layer
                                                 of Counter Ions
                                  Distance from Surface (x)
Figure 1.  Schematic Representation of  Surface Species and Surface Charge-
           Potential Relationship for the  Triple-Layer Model.   Brackets 1n
           the  'zero1 plane Indicate deprotonated surface sites.   (After
           Felmy,  G1rv1n and Jenne 1984)

-------
                 [SOH8]S (H )
          SKai	                                       (3)
                   [SOH2+]S
and
                 [S
-------
at the designated adsorption plane within the double layer; R Is the molar
gas constant (J K"1 mol"1); and T 1s the absolute temperature (°K).
Subsequently, the absence of superscript c Indicates that the work
necessary to bring the Ions from the bulk solvent Into the double layer
has been Included 1n the equilibrium constant.  Surface acidity constants
for reactions (1) and (2) are obtained by Inclusion of the exponential
Boltzmann factor Into equations (3) and (4), respectively,

          *sKai = *SK£1 exp(-F?/RT)                                 (6)

          *\2 = *§Ka2 exp(-Ff/RT).                                (7)

The net surface charge, aQ, 1s related to the electrostatic potential, by
the relationship a0 = C f0.
     Reviews of SCMs (Davis, James and Leckle 1978; Westall 1980; Westall
and Hohl 1980; Pelmy, G1rv1n and Jenne 1984; Sposito 1984) Indicate that
there are four or five types of these models.  According to Morel, Westall
and Yeasted  (1981) the principal differences among them are

       o  the set of surface species considered and the corresponding
          surface reactions

       o  the mathematical expression of free-energy minimum (I.e., the
          "mass law" as a function of surface site "concentration")

       o  the formulation of the coulomblc term (I.e., the electrostatic
          energy term).

These differences are considered In some detail by Morel, Westall and
Yeasted (1981) and Sposito (1984).  Earlier model formulations have
evolved to the point that either the constant capacitance or triple-layer
model Is generally used 1n current research.  The triple-layer model
consists of  two constant capacitance layers bounded by a diffuse layer
(see Figure  1).  Only H+ and OH'are considered to be specifically adsorbed
(I.e., those ions that have formed surface complexes and are located 1n
                                      6

-------
the surface "o" plane)  and contribute to the surface charge,  OQ, and
experience an electrostatic potential, j>0.   All  other specifically
adsorbed ions, including the swamping electrolyte ions,  are considered to
be located in the "b" or inner Helmholtz plane.   These ions contribute to
the charge, a^, and are subject to an electrostatic potential, fa.
Regions of constant capacitance, Ct and C2,  separate the "o"  and "b"
planes and the "b" and "d" planes, respectively.  In contrast, in the
constant capacitance model all specifically adsorbed ions (e.g., H+, OH",
Cd  ) are considered to be located in the surface "o" plane.   These ions
contribute to the surface charge, aQI and are subject to the  potential,
j>0.  In the constant capacitance model, the Inner Helmholtz plane (i.e.,
"b" plane) and C2 region (between the "b" and "d" planes) are omitted.
Ions that are not specifically adsorbed are assumed to be located in the
diffuse layer.
     The choice of which model to use is based upon pragmatic
considerations because Westall and Hohl (1980) concluded that none of the
of the SCMs were more correct than the others.  This conclusion was based
on the finding that all of the SCMs could be made to fit a given set of
experimental data equally well.  Importantly, they emphasize that the
surface complexatlon constants are highly model  specific and adsorption
constants generated with one model cannot be reliably used 1n other
models.
     If adsorption modeling 1s to be done in solutions of nearly constant
ionic strength and composition, the constant capacitance model has the
advantage of requiring one Instead of two capacitance values.  However,
because the surface complexatlon and acidity constants may vary with the
ionic strength and electrolyte composition, the constant capacitance model
is not appropriate for developing sediment criteria where the Ionic
strength varies from dilute fresh waters through estuaries to coastal
marine waters.  Therefore, 1t appears that the triple-layer model is the
model of choice for developing sediment criteria for metals because the
surface complexatlon constants for this model are expected to be nearly
independent of the electrolyte composition.  In addition, several research
groups are now using the triple-layer model, which means the database for

-------
this model is likely to significantly expand.   The following calculations
of surface complexation constants pertain to the triple-layer model.
                                      8

-------
                            TRIPLE-LAYER MODEL

     The conventional surface complexation reaction for the exchange of a
metal 1on with hydrogen at neutrally charged surface sites 1s written as

          SOHa(s) + M++(e) = SOM+(s) + H+(e)                       (8)

with the equilibrium constant designated as *SKU.  Metal plus hydroxyl
adsorption at a protonated site may result 1n the formation of a
hydroxylated surface species, as Illustrated by the reaction

          SOH0(s) + M++(e) + OH"(e) = SOMOH(s) + H+(e)             (9)

with the equilibrium constant designated by *SKUOH.
     It may be noted that, 1n contrast to reaction (8), some Investigators
write the metal 1on adsorption reaction 1n terms of the deprotonated site
as

          S(f(s) + M++(e) = SOM+(s)                               (10)

with the equilibrium constant designated'as SKU.
     The equilibrium constant for reaction  (8) 1s

                 [SOM+]S  (H+)e
          *SK«	                                   (11)
                 [SOH8]S  (M++)e

where the asterisk specifies that the reaction Is written  1n terms of a
neutral adsorption site  (SOH8) as 1n equation  (8), and  the absence of the
superscript c  Indicates  the Boltzmann function has been Included.  Note
that reaction  (8) 1s the  sum of reaction  (2)  (plus the  Boltzmann factor)
and  (10), I.e.,

          *SK« =  *SKai   V                                     (12)

-------
Similarly, the equilibrium constant for equation (9) Is

                      [SOMOH0] (H+)e
          *S|X	                            /-|0\
            NMOH ~ ------——-                                   \13)
                   [SON] (M++)e (OH-)e

or

            KHOH =  SKal * KyaH KHl                               (14)

where K^ is the first hydrolysis constant for M   in the bulk solution
and *SKUOH is the equilibrium constant for reaction (10) where the
hydrolyzed metal ion  MOH  has been substituted for M  .  The inclusion of
the metal-hydroxide surface complex sometimes Increases the accuracy with
which the adsorption  Isotherms can be simulated.
     Although there is little certainty regarding the formation of
bidentate metal ion complexes at solid surfaces, bidentate reactions are
sometimes Included in the system of adsorption reactions (Yeasted 1978;
Mihelcic and Luthy 1986).  The bidentate reaction 1s written

          2(SOH6)(s)  + M++(e) = (SO)2MB(s) + 2H+(e)               (15)

and the equilibrium constant is

                  [(SO)2M']S (H+)l
          **SK°	                                (16)
                   2[SOH8]  (M+*)e

where the double asterisks  indicate two SOH neutral sites are involved in
the reaction, and the reactivity of the bidentate sites 1s assumed to be
proportional to the total available sites rather than to the number of
sites to the second power  (Morel, Westall and Yeasted 1981; Mihelcic and
Luthy 1986).
     Completing the problem definition for adsorption of a metal requires
two additional relationships.  First, a mass balance equation for sites
                                      10

-------
1s needed

          [SOT] = [SOHa] + [SOH2+] + [SO"] + [SOM+] + [SOMOH0]     (17)

where SOT 1s the experimentally determined total concentration of  surface
sites.  If a bidentate reaction 1s Included 1n the system of adsorption
reactions, then a (SO)2Ma term is added to equation (17).  Second, a
charge balance equation 1s required,

          a0 = [SON2*] + [SOM+] - [SO"]                             (18)

where a0 is the net surface charge.
                                      11

-------
                SURFACE  CQMPLEXATIQN AND ACIDITY CONSTANTS

     Applying an SCM to sediment-water systems requires acidity constants
for the absorbent(s) and surface complexation constants for the metal(s)
of Interest and the important competitive cations.  In calculating an
adsorption constant, the activity of the uncomplexed dissolved metal  is
used rather than the total dissolved concentration in order to remove  the
effect of solution composition in complexation constant calculations,  and
thereby increase the independence of these constants from the composition
of the aqueous phase.  Surface complexation constants are optimized to the
available experimental data with a computerized optimization model
(Westall and Hohi 1978), or by iterative use of the SCM embedded in an
equilibrium chemistry model (Benjamin and Bloom 1981).
     Table 1 lists all of the surface complexation constants found in  the
literature that were derived with the triple-layer model for Cd and Zn
sorption onto Fe oxides.  The database 1s rather small, although for
amorphlc Fe two sets of constants were found for Cd and three for Zn,
whereas a single set of Cd and Zn constants was found for goethite (o-
FeOOH).  Constants for the formation of a particular complex onto amorphlc
Fe oxide differ by more than a factor of ten.  Although these constants
will allow preliminary estimates of the activity of metals In pore waters,
steps should be taken to Improve their reliability.  The quality of the
database for Fe oxides could be Improved by recalculating the triple-layer
constants with a common set of reactions and auxiliary constants, in
conjunction with a consistent means of optimizing the surface complexing
constants.  This database could be supplemented by recalculating surface
complexation constants from adsorption data that were developed for the
constant capacitance model, and from studies for which surface
complexation constants were not calculated.  Surface complexation
constants can be calculated from information available in the literature
if a standard procedure was used to prepare the oxide and sufficient
system characterization data are available.  Thus, Cd and Zn surface
complexation constants could be calculated from the data of Balistrieri
and Murray (1982) for sorption onto goethite from 0.1N NaN03 solutions,
                                      12

-------
                             Table 1.   Surface Complexation Constants for Cd and Zn
Absorbent Surface Sites Total Time Temp. Electrolyte
Area Concentration Identity Cone.
of Absorbate
{m2 g"1) (mol kg"1) (M) (Hours) (°C) (M)
CADMIUM
Fe 0 «H 0(A) (b) 9.85 1 x 10~5 2 25 NaNO^ 0.1
Fe 0 'N 0 182(d) 9.85 2 x 10~* 2 25 NaNO 0.1
232 3
-* 0-FeOOH 51.8 0.22 3 x 10~6 2.5 25 NaCl 0.53
CO
ZINC
Fe 0 «H 0(A) (b) 9.85 1 x 10~5 2 25 NaN03 0.1
Fe 0 «H 0(A) 182 9.35 5 x 10 2 25 NaN03 0.1
Fe 0 'H 0(A) (b) 10? - 105 16-18 22-25 NaNO^ 0.01
O-FeOOH 51.8 0.22 3 x 10~6 2.5 25 NaCl 0.53
-Log Surface Reference
Complexation Constant
KM "SlOH KNa KAn
4.8 11.2 Benjamin and
Bloom (1981)
ic\ ic\
6.01 ' 9.3* 9.0 6.9 Davis and
Leckie (1978b)
(c) (c)
1.3 9.35 8.4 7.0 Balistrieri and
Murray (1982)
(c) (c)
4.8 10.5 Benjamin and
Bloom (1981)
9.4 9.0 6.4 Davis and
Leckie (1978b)
2.3 10.5 9.1 6.4 Dempsey and
Singer (1980)
9.15 8.4 7.0 Balistrieri and
Murray (1982)
(a) BET method using N
(b) Same preparation method as Davis and Leckie (1978)
(c) Calculated with the Stanford Version of the triple-layer model (Davis, James and Leckie 1978)
(d) An area of 300 m  g   obtained by negative adsorption

-------
and compared with the values they report for adsorption from 0.53N NaCl
solutions.
     There 1s little basis for selecting particular constants from among
those given in Table 1.  The constants of Benjamin and Bloom (1981)  are
provisionally selected on the basis that they appear reasonable in
comparison with the other values, which are also tabulated.   In addition,
a degree of internal consistancy 1s obtained by selecting all four
constants from the same authors.
     It should be noted that according to Anderson and Benjamin (1985),
the fractional adsorption of Cd, Co, Cu, and Zn per mole of  Fe In fine-
grained goetMte (a-FeOOH) or aged Fe oxyhydroxide precipitate was greater
at any given pH than the fractional adsorption on fresh amorphlc Fe oxide.
Thus, using the equilibrium constants of Benjamin and Bloom  (1981) for
amorphlc Fe oxide to represent the partitioning of metals between
sediments and associated pore water may yield a higher estimate of the
desorbed metal activity 1n pore water.  This method is expected to result
1n a conservative estimate of the potential toxldty of metals desorbed to
the pore water.
     Acidity  (I.e., protonatlon/deprotonatlon) constants for the triple-
layer model are determined experimentally by potentlometrlc  tltratlon of
the solid at a series of Ionic strengths.  The tltrations are carried out
1n the absence of specifically adsorbing Ions other than H+  and OH~ and  in
the absence of other Ions except the weakly adsorbing swamping electrolyte
Ions.  Selected acidity constants appropriate to the triple-layer model
are given 1n Table 2.  There 1s little difference between the two sets of
constants for amorphlc Fe oxide.  The acidity constants of Davis and
Leckle (1978b) are selected because these particular constants have been
preferentially used by subsequent investigators.
                                      14

-------
                Table 2.  Surface Acidity Constants for Iron Oxides
Oxide P*SK,i P*S|Ca9 Surface
al " Area
Fe203-H20(A) 5.1 10.7 600
Fe203-H20(A) 5.3 10.6 (a)
a-FeOOH 4.2 10.8 48
a-FeOOH 4.9 32
a-FeOOH 5.57 9.52
Source
Davis and
Leckie (1978b)
Dempsey and
Singer (1980)
Yates (1975)
Hingston et al.
(1968)
Balistrieri and
Murray (1981)
                                          2   .1
(a)  Assumes specific  surface  area of  600 m  g
                                        15

-------
                            OTHER DATA NEEDED

     The site-specific data needed to apply an SCM to a sediment-water
system include the following:  1) concentrations of adsorption sites for
each of the principal adsorbents, 2) concentrations of sorbed metals,  and
3) potentlometric titration and surface area data to calculate the
appropriate surface charge and electrostatic potentials.  The total  number
of adsorption sites in single solid-phase systems can be obtained by
short-term tritium or deuterium exchange with surface hydroxyls or by
acid/base titrations; however, there 1s currently no method for directly
determining the number of sites for individual adsorbent phases 1n a
heterogenous sediment.  One possible approach is to use literature data to
determine the moles of sites per mole of amorphic Fe.  Additional research
will be required to determine the error associated with such an approach
(i.e., determine the variation in numbers of sites under differing
conditions of oxide preparation and extent of aging).  Such an approach to
estimating site concentration is compatible with the hot acidic
hydroxylanrtne hydrochlorlde extraction method proposed for estimating the
total quantity of sorbed metals and amorphic Fe and cryptocrystalline Mn
oxide adsorbentsz'3'.
     Double-layer capacitance cannot be obtained by direct experimental
measurement.  A value of 20 pf cm"2 1s generally used for the outer layer
capacitance (Davis, James and Leckle 1978; Bal1str1eri and Murray 1981).
A value of 140 /*F cm"2 was used by Davis, James and Leckle (1978) and
Bal1str1er1 and Murray (1981) for Inner-layer capacitance.
     For the electrostatic potential calculation for Fe oxides, some
Investigators use the Davis and Leckle (1978b) surface area value of
600 m2 g"1 (e.g., Dempsey and Singer 1980; Dzombak and Morel 1985).
 (a)  Jenne, E. A.   1987.  Sediment Quality Criteria for Metals; II. Review
 of Methods for Quantitative Determination of Important Adsorbents and
 Sorbed Metals in Sediments.  Submitted by Battelle, Washington
 Environmental Program Office, Washington, D.C. to the U.S. Environmental
 Protection Agency,  Criteria and Standards Division.
                                      16

-------
                             DISCUSSION

     Surface complexation models have most frequently been applied to
Individual crystalline or amorphic oxides in efforts to further understand
the reactions and improve the description of the adsorption process (e.g.,
Davis, James and Leckie 1978; Davis and Leckie 1978a, 1978b, 1979, 1980;
Benjamin and Leckie 1981; Balistrieri and Murray 1981).  These models have
also been applied to clays (James and Parks 1982) and to mixed systems
containing two oxides or an oxide plus a clay (Honeyman 1984).
     Surface complexation models have also proved useful in describing the
adsorption of both anions and cations in systems containing heterogenous
solids and mixed electrolytes.  Adsorption studies of heterogenous systems
include the following:  Cd onto soil (Honeyman 1984);  phosphate onto non-
calcareous soils (Goldberg and Sposito 1984);  Cr by in situ Fe oxide
precipitation, as used in conventional coagulation treatment of waste
waters (Dzombak and Morel 1985); and  Pb and Zn onto the filter cake
(primarily hematite and magnetite) that results from scrubbing blast-
furnace off gases (Mlheldc and Luthy 1986).  The adsorption of Cd, Cu,
Pb, and Zn onto goethite was successfully modeled in synthetic seawater
(Balistrierl and Murray 1982).  In most instances, the model simulations
do not exactly reproduce experimental curves.  However, the fit of the SCM
to the experimental data in these diverse applications Indicates the broad
applicability of the methodology to soil and sediment systems.
     The  key assumptions Inherent in an SCM are as follows:  1) adsorption
can be represented as a surface complexation reaction with surface
hydroxyl  groups, 2) the chemical potential of the adsorbing species can be
represented by an activity of the uncomplexed metal, and 3) all adsorption
sites are equivalent.  The third assumption appears to be true for
amorphic  Fe oxide at low saturations and 1s a useful approximation at
higher loadings  (e.g., Benjamin and Leckie 1981).  These models contain
two additional features of particular interest.  First, it is assumed that
only the  proton  and hydroxyl ions form Inner-sphere complexes and that the
metals form only outer-sphere complexes  (Cotton and Wilkinson 1980).
Second, the reference state  for the  ion activities is  Infinite dilution.
Where the SOM(s) term is a small percentage of the total neutral  sites,
                                      17

-------
metal adsorption Is essentially Independent of competition with major
cations at the concentrations at which they occur in freshwater.  Under
these conditions, cation exchange reactions at classical electrostatic
sites that result from broken surface bonds or positive-charge deficiency
associated with structural cations do not need to be Included in the
modeling of trace metals.  However, competition from major cations (e.g.,
Mg) for neutral SOH8 sites needs to be considered when modeling adsorption
1n marine waters (Bal1str1er1 and Murray 1982).
     The tabulated surface complexation constants for the same oxide may
vary depending upon the following:  1) the correctness of the aqueous
spedatlon calculations used to obtain the activity of the uncomplexed
metal; 2) the method used for estimating the concentration of surface
sites, since these estimates vary with the method of their determination;
3) the omission of reactions for the swamping electrolyte Ions; 4) the
differences among different preparations of a given oxide; 5) the reaction
times 1n the experimental systems used to determine the surface
complexation equilibrium constants; 6) the other methodological
variations, such as solids concentration or the degree of saturation of
surface complexation sites; and 7) the associated constants  (e.g.,
capacitance, surface area) originally used to optimize the surface
complexation equilibrium constant.  It Is likely that the variability 1n
these constants could be significantly reduced if they were all
recalculated using a coherent set of reactions and associated constants
and optimization constants.  Additional Cd and Zn surface complexation
constants can be obtained by recalculation of those developed for the
constant capacitance model.
     General application of an SCM for developing sediment criteria for
metals and other Important environmental problems requires that several
methodological problems be solved.  It also requires the development of an
Improved database both by further evaluation and recalculation of those
constants available in the literature, and by determining missing values.
Methodological problems Include verification of the proposed method for
estimating sorbed metals and the development of an approach  for estimating
the  concentration of  sites on individual solids in heterogenous systems.
                                      18

-------
                               REFERENCES


Anderson, P. R., and M. M. Benjamin.   1985.  "Effects of Silicon on the
     Crystallization and Adsorption Properties of Ferric Oxides."
     Environ. Sci.  Technol.  19:1048-1053.

Bal1strier1, L. S., and J. W. Murray.   1981.  "The Surface Chemistry of
     Goethlte (a-FeOOH) 1n Major Ion  Seawater."  Amer. J. Sci. 281:788-
     806.

Ballstrleri, L. S., and J. W. Murray.   1982.  "The Adsorption of Cu, Pb,
     Zn, and Cd on  Goethlte from Major Ion Seawater."  Geochem. et
     Cosmochim. Acta 46:1253-1265.

Benjamin, M. M., and N. S. Bloom.  1981.  "Effects of Strong Binding of
     Anionlc Adsorbates on Adsorption  of Trace Metals on Amorphous Iron
     Oxyhydroxide."  In Adsorption  From Aqueous Solutions, ed. P. H.
     Tewari, pp. 41-60.  Plenum Press, New York.

Benjamin, M. M., and J. 0. Leckle.   1981.   "Multiple-Site Adsorption of
     Cd, Cu, Zn and Pb on Amorphous Iron Oxyhydroxide".  J. Colloid
     Interface Sci. 79:209-221.

Cotton, F. A., and  G. Wilkinson.   1980.  Advanced Inorganic Chemistry; 4th
     Edition.  John Wiley & Sons, NY.

Davis, J. A., R. 0. James and J.  0. Leckle.  1978.  "Surface lonization
     and CompTaxation at the Oxide/Water Interface. I. Computation of
     Electrical Double Layer Properties 1n Simple Electrolytes."  iL.
     Colloid Interface Sc1.  63:480-499.

Davis, J. A., and J. 0. Leckle.  1978a.  "Effect of Adsorbed Complexing
     Llgands on Trace Metal  Uptake by Hydrous Oxides."  Environ. Sci.
     Technol.  12:1309-1315.

Davis, J. A., and J. 0. Leckle.  1978b.  "Surface lonlzation and Comple-
     xation at the Oxide/Water Interface.  II. Surface Properties of Amo-
     rphous  Iron Oxyhydroxide and Adsorption of Metal Ions."  J. Colloid
     Interface Sci. 67:90-107.

Davis, J. A., and J. 0. Leckle.  1979.  "Spedation of Adsorbed Ions at
     the Oxide/Water Interface."  In Chemical Modeling in Aqueous Systems,
     ed. E. A. Jenne, pp. 299-317,  Symp. Series 93.  American Chemical
     Society.

Davis, J. A., and J. 0. Leckle.  1980.  "Surface lonlzation and
     Complexation at the Oxide/Water Interface. 3.Adsorption of Anions."
     J. Colloid Interface Sci. 74:32-43.

Dempsey, B. A., and P. C. Singer.  1980.  "The Effects of Calcium on the
     Adsorption of Zinc by MnOx(s)  and Fe(OH)3(am)."  In Contaminants and


                                      19

-------
     Sediments, ed. R. A. Baker, pp. 333-352.  Ann Arbor Science, Ann
     Arbor, Michigan.

Dzombak, D. A., and F. M. M. Morel.  1985.  Adsorption of Inorganic
     Contaminants In Ponded Effluents from Coal Fired Power Plants.
     PB 86103736, National Technical Information Service, Springfield,
     Virginia.

Felmy, A. R., D. C. Glrvln and E. A. Jenne.  1984.  MINTEQt A Computer
     Program for Calculating Aqueous Geochemlcal Equilibria.  PB 84157148,
     National Technical Information Service, Springfield, Virginia.

Goldberg, S., and G. Sposlto.  1984.  "A Chemical Model of Phosphate
     Adsorption by Soils: II. Noncalcareous Soils."  Soil Scl. Soc.
     48:772-778.

Hlngston, F. J., A. M. Posner, and J. P. Quirk.  1968.  "Adsorption from
     Aqueous Solution".  Am. Chem. Soc., Adv. Chem. Ser. 79:82.

Honeyman, B. D.  1984.  Cation and Am'on Adsorption at the Oxide/Solution
     Interface 1n Systems Containing Binary Mixtures of Adsorbents.  Ph.D.
     Thesis, Stanford University, Palo Alto, California.

James, R. 0., and T. W. Healy.  1972.  "Adsorption of Hydrolyzable Metal
     Ions at the Oxide/Water Interface.  III. A Thermodynamlc Model of
     Adsorption."  J. Colloid Interface Sc1. 40:65-81.

James, R. 0., and G. A. Parks.  1982.  "Characterization of Aqueous
     Colloids by Their Electrical Double-Layer and Intrinsic Surface Che-
     mical Properties."  Surface and Colloid Science 12:119-217.

Jenne, E. A., D. M. D1 Toro, H. E. Allen, and C. S. Zarba.  1986. "An
     Activity-Based Model for Developing Sediment Criteria for Metals: I.
     A New Approach," pp. 560-562.  In Proceedings of the Chemicals 1n the
     Environment;  International Conference, eds. J. N. Lester, A. Perry,
     and R. M. Sterrltt. Selper Ltd., London.

Mlheldck, J. R., and R. G. Luthy.  1986.   "Adsorption of Lead and Zinc on
     Blast Furnace Iron Oxide Sol Ids."  J. Water Poll. Control Fed.
     58:242-249.

Morel, F. M. M., J. C. Westall and J. G. Yeasted.  1981.  "Adsorption
     Models: A Mathematical Analysis In the Framework of General
     Equilibrium Calculations."  In Adsorption of Inorganics at Solid-
     Liquid Interfaces, ed. M. A. Anderson  and A. J. Rubin, pp. 263-294.
     Ann Arbor Science, Ann Arbor, Michigan.

Schlndler, P. W., B.  Furst, R. D1ck and P.  N. Wolf.  1976.  "Llgand
     Properties of Surface Sllanol Groups:  I. Surface Complex Formation
     with Fe   , Cu2*, Cd   , and Pb   ."  J.  Colloid Interface Scl. 55:469-
     475.
                                      20

-------
Sposito, G.  1984.  The Surface Chemistry of Soils.  Oxford University
     Press, New York.

Stern, 0.  1924. "The Theory of the Electrolytic Double-Layer."
     Z. fur Electrochinrie. 30:508-516.

Stumrn, W., H. Hohl and F. Dalang.  1976.  "Interaction of Metal Ions with
     Hydrous Oxide Surfaces."  Croat. Chem. Acta. 48:491-504.

Stuiran, W., R. Kummert and L. Sigg.  1980.  "A Ligand Exchange Model for
     the Adsorption of Inorganic and Organic Ligands at Hydrous Oxide
     Interfaces."  Croatica Chemica Acta. 53:291-312.

Westall, J.  1980.  "Chemical Equilibrium Including Adsorption on Charged
     Surfaces."  In Particulates in Water, ed. M. C. Kavanaugh and J. 0.
     Leckie, pp. 33-44, Adv. Chem. 189.  American Chemical Society.

Westall, J., and H. Hohl.  1978.  "A General Method for the Computation of
     Equilibria and Determination of Equilibrium Constants for Adsorption
     at Hydrous Oxide Surfaces."  Paper presented at Symposium on Chemical
     Modeling-Sped ation, Sorption, Solubility and Kinetics, ACS Annual
     Meeting, Miami, Florida, Sept. 10-15.

Westall, J., and H. Hohl.  1980.  "A Comparison of Electrostatic Models
     for the Oxide/Solution Interface."  Adv. Colloid Interface Sci.
     12:265-294.

Yates, D.  E. 1975.  The Structure of the Oxide/Aqueous Electrolyte
     Interface.  Ph. D. Thesis. University Melbourne. Australia.

Yates, D.  E.f S. Levine and T. W. Healy.  1974.  "Site-binding Model of
     the Electrical Double Layer at the Oxide/Water Interface."  Chem.
     Soc.  Faraday Trans. 70:1807.

Yeasted, J. G.  1978.  The Modeling of Lake Response to Phosphorus Loa-
     dings; Empirical, Chemical and Hydrodynamic Aspects.  Ph.D. Thesis,
     Dept. of Civil Engineering, M.I.T., Cambridge, Massachusetts.
                                      21

-------