United States
Environmental Protection
Agency
Office of
Research and
Development
Office of Solid Waste
and Emergency
Response
EPA/540/5-94/505
October 1994
                                   EPA  Ground  Water  Issue
                          Natural Attenuation  of Hexavalent
                          Chromium in Groundwater and Soils

                          Carl D. Palmer*  and Robert W. Puls**
Introduction

Chromium is an important industrial metal used in diverse
products and processes (Nriagu, 1988a,  b). At many locations,
Cr has been released to the environment via leakage, poor
storage, or improper disposal practices (Palmer and Wittbrodt,
1991; Calder, 1988). Within the environment, Cr  is found
primarily in two oxidation states: Cr(VI) and Cr(lll).Cr(VI) is
relatively mobile in the environment and is acutely toxic,
mutagenic (Bianci et al., 1983; Beyersmann et al., 1984;
Bonatti et al., 1976; Paschin et al., 1983), teratogenic (Abbasi
and Soni, 1984), and carcinogenic  (Mancuso and Hueper,
1951; Mancuso, 1951; Waterhouse, 1975; Yassi  and Nieboer,
1988; One, 1988). In contrast, Cr(lll) has relatively lowtoxicity
(van Weerelt et al., 1964) and is immobile under  moderately
alkaline to slightly acidic conditions.

Concerns about the impact of chromium  on human health and
the environment require an evaluation of the potential risk of
chromium entering the groundwater flow system and being
transported beyond compliance boundaries. At sites where
such potential exists, active remedial measures such as
excavation or pump-and-treat have  been undertaken.
Experience at sites where pump-and-treat remediation of
chromium-contaminated groundwater is currently underway
suggests that, although it is feasible to remove high levels of
chromium from the subsurface, as concentrations decrease it
becomes more difficult to remove the remaining chromium
(Wittbrodt and Palmer, 1992). While several new remedial
technologies are being investigated, there is still concern
about the cost of such remediation technology; and, at many
sites, there is a debate about the need for expensive
remediation.
Researchers have identified natural reductants that can
transform the more toxic hexavalent form of chromium to the
less toxic trivalent form. Under alkaline to slightly acidic
conditions, this Cr(lll) precipitates as a fairly insoluble
hydroxide, thereby immobilizing it within the soil. Such
"natural attenuation" of hexavalent chromium is of great
interest because it suggests that strict water-quality standards
do not have to be attained everywhere within and beneath the
site. If natural attenuation does occur, pump-and-treat
remediation could desist after the most contaminated
groundwater has been removed, even if the maximum
contaminant level (MCL) has not be achieved. Under certain
circumstances, expensive  remedial  measures may not even
be necessary.

In this paper, what is known about the transformation of
chromium in the subsurface is explored. This is an attempt to
identify conditions where it is most likely to occur, and
describe soil tests that can assist in determining the likelihood
of natural attenuation of Cr(VI) in soils.

The Geochemistry of Chromium

Chromium exists in oxidation states ranging from +6 to -2,
however, only the +6 and  +3 oxidation states are commonly
encountered in the environment. Cr(VI) exists in solution as
monomeric ions h^CrO^, HCiO^ (bichromate), and
(chromate), or as the dimeric ion Cr207" (bichromate)
.Environmental Science and Engineering, Oregon Graduate Institute
  of Sciences, Technology
** Robert S. Kerr Environmental Research Laboratory, U.S. EPA
                          Superfund Technology Support Center for
                          Ground Water

                          Robert S. Kerr Environmental
                          Research Laboratory
                          Ada, Oklahoma
                       Ota* of Sold Waste asd

-------
 (e.g. Palmer and Wittbrodt, 1991; Richard and Bourg, 1991).
 The monomeric species impart a yellow color to the water
 when the [Cr(VI)] is greater than 1 mg/L Water that contains
 high levels of Qcfft has an orange color.

 The monomeric chromate species are related through a series
 of acid dissociation reactions

            H2CrCg  a HCr04+H+    ;  K,
HCrO4
                                    ; K2
the pK values are -0.86 and 6.51 , respectively (Allison et al.,
 1990). The bichromate is the result of the polymerization of
the monomeric bichromate ions to form the dimer, QrzO?~,
                                                    1.0
                              Q) 0 8

                              i
                              Q.
                              (0 0.6
                              •5
                              C
                              O 0.4

                              1
                              £ 0.2
                                                    0.0
                                                         Cr(VI)tot= 5 mmol/L
                                                                                    CrOj
                                                                                10
                                                                                        12
         HCrO4+HCrO4
+ & O
                             ;Kd
                                                             Figure 1. Distribution of Cr(VI) species as a function of pH.
where pKd  is -1.54 (Allison et al., 1990). The relative
concentration of each of these species depends on both the
pH of the contaminated water (Fig. 1) and the total
concentration of Cr(VI) (Fig. 2). Significant concentrations of
h^CrdJ only occur under the extreme condition of pH c 1.
Above pH 6.5, CrOf generally dominates. Below pH 6.5,
HCrO4 dominates when the Cr(VI) concentrations are low (<30
mM); but Cr2O^ becomes significant when  concentrations are
greater than 1 mM, or it may even dominate when the total
Cr(VI) concentrations are greater than 30 mM.

In the Cr(lll)-H2O system, Cr(lll) exists predominantly as Cr3*
below pH 3.5. With increasing pH, hydrolysis of Cr3* yields
CrOH^Cr(OH)^ Cr(OH)3,and  Cr(OH)4(Rai, et al.,  1987).
At high concentrations, these ions impart a green color to the
solution. Under slightly acidic to alkaline conditions, Cr(lll)
can precipitate as an amorphous chromium hydroxide.
Amorphous Cr(OH)3 can crystallize as Cr(OH)3-3H2O or
Cr2O3 (eskolaite) under different conditions (Swayambunathan
et al., 1989). In  the presence of Fe(lll), trivalent chromium can
precipitate  as a solid solution. If the pH within the contaminant
plume is between 5 and 12, the aqueous concentration of
Cr(lll) should be less than 1 ^mole/L (<0.05 mg/L) (Fig. 3).

There are several mineral phases that contain Cr(VI)that may
be present  at chromium contaminated sites. Palmer and
Wittbrodt (1990) identified PbCrO4(crocoite),PbCrO4 . H,O
(iranite), and K2CrO4 (tarapacaite) in chromium sludge from a
hardchrome plating facility. CaCrO4 was found at a seepage
face in a drainage ditch where there was high evaporation.
Most of the contaminated groundwater was at equilibrium with
BaCr04(hashemite).BaCrO4 forms a complete solid solution
with BaS04(Rai et al., 1988) and can be a major impediment
to the remediation of chromium contaminated sites by pump-
and-treat (Palmer and Fish, 1992; Wittbrodl and  Palmer,
1992).
                                                   1.0


                                                   0.8
                                                Z
                                                O 0.4
                                                ^ 0.2
                                                cc "•*
                                                U_
                                                   0.0
                                                    0.01     0.10     1.0      10.      100.    1  )0
                                                              TOTAL Cr(VI)  (mmol/L)

                                               Figure 2. Fraction of bichromate (HCrO^ and bichromate (Cr2O^) at
                                                       pH 4 as a function of the total Cr(VI) concentration.
                                    -2
                                r^


                                O
                                 O) -6


                                    -8
                                                                       Cr(OH)3(am)
                                                                        	MOi,	
                                                                    Fe0.75Cr  o. 2 5(OH)^^ •
                                                             4      6     8    10    12    14
                                                                            PH
                                               Figure 3. Cr(lll) concentration in equilibrium with Cr(OH)3(am) and
                                                       Fe076 Cr026(OH)3 .based on data from Rai et al., (1 987) and
                                                       Sass and Rai (1987), respectively.

-------
Reduction of Hexavalent Chromium

Cr(VI) is a strong oxidant and is reduced in the presence of
electron donors. Electron donors commonly found in soils
include aqueous Fe(l I), ferrous iron minerals, reduced sulfur,
and soil organic matter.

The reduction of Cr(VI) by ferrous iron can be written as
HCrO4 +  31%2*+ 71-T + Cr3*
                                           4H20
This reaction is very fast on the time scales of interest for most
environmental  problems with the reaction going to completion
in less than 5 minutes even in the presence of dissolved
oxygen (Eary and Rai, 1988). Only when the  pH is greater
than 1 0 or when P04 concentrations exceed 0. 1 molar does
the rate of oxidation of Fe2* by dissolved oxygen exceed the
rate of oxidation  by Cr(VI)(Eary and Rai, 1988). When the pH
of the groundwater is greater than 4, Cr(lll) precipitates with
the Fe(lll) in  a solid solution with the general composition
CrxFe,.x(OH)3 (Sass and Rai, 1987; Amonette  and Rai, 1990).
If the reduction of Cr(VI) by Fe(n) is the only source of Fe(lll)
and Cr(lll), a solid solution with the composition
Cr025Fe075(OH)3 forms via the reaction
             3Rj2+ + 3H20 + 5OH~-> 4Cr0.26Fe0.75(OH)3
(Eary and Rai, 1988; Sass and Rai, 1987). The volubility of
CrxFe1 x(OH)3 decreases as the mole fraction of Fe(lll) in the
solid increases. Therefore, if the pH is between 5 and 12, the
concentration of Cr(lll) is expected to be less than 1 0"6 molar.

Numerous minerals in geologic materials contain ferrous iron
that is potentially available for the reduction of hexavalent
chromium. These iron-containing minerals may be silicates,
oxides, or sulfides. Common ferrous iron-containing silicates
include olivine; pyroxenes such as augite and hedenbergite;
the amphiboles hornblende, cummingtonite, and grunerite;
micas such as biotite, phlogopite, and glauconite; chlorites,
and the smectite nontronite. Iron oxides such as magnetite
(Fe2+Fe!fO4) contain iron as a major constituent, however,
hematite (Fe3+O3) can contain small amounts of (FeO). in
sulfide minerals such as pyrite (FeS2), both the ferrous iron
and the sulfide are active in reducing hexavalent chromium.

Lancy (1966) suggested that pyrite could be used for treating
spent cooling waters that contain Cr(VI) as a corrosion
inhibitor.  He stated that the reduction of Cr(VI) occurs at the
pyrite surface rather than in solution. Lancy (1966) found that
reduction by pyrite occurred even in slightly alkaline  solutions;
however, the pyrite had to  be continuously abraded to remove
surface coatings.  Blowes and Ptacek (1992) conducted batch
tests in continuously agitated reaction vessels containing a
solution of 18 mg/LCr(VI) and pyrite both in the presence and
in the absence of calcite. In the experiments that used both
pyrite and calcite, 50% of the Cr was removed in less than 6.5
hours. Concentrations were < 0.05 mg/L after 20 hours.
Experiments conducted without the calcite attained 50%
removal in 1 hour and concentrations were < 0.05 mg/L in less
than 4 hours.

Cr(VI) reduction in the presence of iron oxides has been
observed in several experiments. White and Hochella (1989)
found that magnetite and ilmenite reduced Cr(VI)toCr(lll).
The reduction of Cr(VI) in the presence of hematite (Fe2O3)
was demonstrated by Eary and Rai (1989). They attribute the
reduction to the presence of a small amount of an FeO
component in the hematite. They suggest that reduction
occurs in solution after the FeO component has been
solubilized.

Reduction of Cr(VI) by ferrous iron-containing silicates has
been reported. Eary and Rai (1989) suggest that the
reduction of Cr(VI) in the presence of biotite occurs in solution
rather than at the mineral surface. They observed an increase
in the rate of reduction when  their suspensions were spiked
with Fe3*. They explain their results with the mechanism
proposed by White and Yee (1985) in which Fe3+ is reduced at
the  mineral surface by the reaction
                                                            [Fe(n)
                                                                      J biotite
                       Fe3
, [Fe(lll)]biotite
                                                       where the ions in the brackets denote ions within the crystal
                                                       structure of biotite. To maintain charge balance, K*is
                                                       released to solution as the iron in the crystal structure is
                                                       oxidized. The Cr(VI) in solution is then reduced by the Fe2+.
                                                       The Fe3* resulting from this reduction reaction is then
                                                       adsorbed to the surface of the biotite where it is again reduced
                                                       to Fe2*, thus setting  up a cycle that ultimately results in the
                                                       reduction of more Cr(VI)than is stoichiometrically possible for
                                                       the amount of iron that is in solution.

                                                       There are some key experimental difficulties in studying
                                                       groundwater/mineral interactions such as those just described
                                                       that have some bearing on the transfer of knowledge to the
                                                       field. Although the processes can in some cases be
                                                       interpreted from the data on mineral reactions, the rates
                                                       themselves may be quite useless. A key difficulty in studying
                                                       mineral reactions in the laboratory is that the rate of the
                                                       reaction depends on how the solid  phase was prepared. For
                                                       example, if the samples are ground and simply washed before
                                                       use, microparticles can adhere to the larger grain surfaces.
                                                       These microparticles have greater specific surface area and
                                                       can react at a much faster rate than the larger size particles.
                                                       Such experimental artifacts were observed in weathering
                                                       studies of pyroxenes (Schott et al., 1981).

                                                       Another important reductant in soils is organic matter.  In fact,
                                                       bichromate reduction has been used as a wet  combustion
                                                       method for the determination of soil organic carbon (Walkley
                                                       and Black, 1934). Bichromate can react with soil organic
                                                       carbon according  to
                                                                2Cr2O7" +  3C°+ 16t-T->4Cr3+ + 3C02+

-------
The Cr3* may hydrolyze and precipitate as Cr-hydroxide or it
may bind to the remaining soil organic carbon. Much of the
soil organic carbon is present as soil humic and fulvic acids.
Redox reaction with these materials has  been demonstrated
for several  redox reactive species.  Reduction of Cr(VI) by soil
humic and fulvic acids has been demonstrated by Bartlett and
Kimble(1 976), Bloomfield and Pruden (1980), Goodgameet
al. (1984), Boyko and Goodgame (1986), and Stollenwerk and
Grove (1985). The rate of reduction of Cr(VI) decreases with
increasing pH  (Fig. 4), increases with  the increasing initial
Cr(VI) concentration, and increases as the concentration of
soil humic substance increases. At neutral pH, many weeks
may be required for the Cr(VI) to be completely reduced
Cr(VI).

In addition to these abiotic reduction pathways, Cr(VI) can be
reduced by microbes in the subsurface (Martin et al., 1994).
Both aerobic and anaerobic reduction by microbes have be
observed, however, the latter is more common. The
mechanisms for Cr(VI) by these microbes is not well known. It
may be part of a detoxification mechanism that occurs
intracellularly. Alternatively, the  chromate may be utilized as a
terminal electron acceptor-as part of the cell' s-metabolism. A
third possibility is that reduction  is an extracellular reaction
with excreted waste products such as H2S. In addition to two
strains of Gram-positive bacteria, Martin et al., (1994) found a
fungus in contaminated soil that was capable of reducing
Cr(VI) under anaerobic conditions.
025
i
J 00201
0
E O.ffitS
zz
0 0616
Cr(VI) Reduction bya
Soil Humic Acid
100 mg SHAA
• ^™
• A A
• . pH 5
A r
• A A
    0.005
    0.000 •
        o   200 400 600 aoo 10001200 1400 1SOO 1800 2000
                         Time (hrs)
Figure 4. Reduction of Cr(VI) in a suspension of 100 mg/L soil humic
        acid (SHA) at pH 3 and 5 (Wittbrodt and Palmer, in
        preparation).
Oxidation of Cr(lll)

Any evaluation of the natural attenuation of Cr(VI) must
consider the potential oxidation of the Cr(lll) to the toxic Cr(VI)
form. In contrast to the numerous pathways for the reduction
of Cr(VI), there are very few mechanisms for the oxidation of
Cr(lll). Only two constituents in the environment are known to
oxidize Cr(lll) to Cr(VI): dissolved oxygen and manganese
dioxides (MnO2)(Eary and Rai, 1987). Studies of the reaction
between dissolved oxygen and Cr(lll) revealed very little
(Schroeder and Lee, 1975) or no  (Eary and Rai, 1987)
oxidation of Cr(lll) even for experiments conducted at pH as
great as 12.5 for 24 days. Therefore, the transformation of
Cr(lll) by dissolved oxygen is not likely to bean important
mechanism for the oxidation of Cr(lll).

Oxidation of Cr(lll) has been  observed in several soils and
sediments. The oxidation of the Cr(lll) can be relatively slow
requiring several months (Fig. 5). For example, Palmer and
Wittbrodt (1 990) monitored Cr(VI) concentrations in  batch
tests using three different geologic media from a site in
Corvallis, OR.  They observed increases in Cr(VI)
concentrations over a 300-day period with Cr(VI)
concentrations becoming as great as 7 mg/L in one
experiment. Bartlett and Kimble (1976) did not observe
oxidation of Cr(lll) in their dried soils but Bartlett and James
(1979, 1983a)  did observe oxidation in soils that were
maintained  in a moist state.
        ID'1
                                                                  O
                                                                      iii naii
                                                                         0   50  100  150  200 250  300  350
                                                                                    TIME (Days)
                                                              Figure 5. Cr(lll) oxidized to Cr(VI) in a suspension of Willamette silt
                                                                      loam.
Bartlett and James (1979) observed a correlation  between the
amount of Cr(lll) oxidized by soils and the amount of
hydroquinone-reduced manganese in soils and  suggested the
oxidation of Cr(lll) is the result of interaction with manganese
dioxides. This hypothesis has been experimentally verified
using B-MnO2 or pyrolusite (Eary and Rai, 1 987) and 8-MnO2
(Fendorf and  Zasoski, 1992; Riser and Bailey, 1992). There is
an increase in the rate and amount of Cr(lll) oxidation as pH
decreases, and  the surface area to solution volume increases.
Experimental results indicate  that the  oxidation follows the
reaction
                  1.5 S-MnO2 -> HCiO4 + 1.5Mn2+
Significant oxidation of Cr(lll) was observed in less than 1
hour (Fendorf and Zasoski, 1 992) and continued for more than
600 hours (Eary and Rai, 1987). Eary and Rai (1987)
developed an empirical rate law for the oxidation of Cr(lll) by

-------
B-MnO2; however, the zero point charge for this phase is quite
different than birnessite which is more commonly found in
soils. Therefore, this rate law may not be applicable to
manganese dioxides in soils.

Perspective on the Natural Attenuation of Cr(VI)

If hexavalent  chromium can be reduced and immobilized in
the subsurface as a result of interaction with naturally existing
reductants, then expensive remedial measures may not be
required at certain sites. In principle, the natural attenuation of
Cr(VI) in the subsurface is feasible. There are several natural
reductants that can  transform Cr(VI) to Cr(lll). If the pH of the
contaminant plume is between about 5 and 12, Cr(lll)
precipitates as Cr(OH)3 or as part of a solid solution with
Fe(lll), thereby keeping Cr(lll) concentrations below 1 jimole/L
(0.05 mg/L). Whether or not  natural attenuation at a particular
site is a viable option depends on the characteristics of both
the aquifer and the  contaminant plume under investigation.

The potential reductants of Cr(VI) include aqueous species,
adsorbed  ions, mineral constituents, and organic matter.
When a contaminant plume containing hexavalent chromium
enters  the subsurface, it displaces the groundwater containing
the dissolved reductants. There is little mixing of the waters
containing the reducing agents and the  Cr(VI)-contaminant
plume. What mixing does occur will be driven by molecular
diffusion at the front of the plume or from the edges of the
plume  and diffusion from lower permeability lenses containing
relatively immobile water. Thus, aqueous reductants such as
Fe2* are not going to be important in reducing hexavalent
chromium. Mixing of reductants and Cr(VI) in the plume are
going to occur primarily though the interactions of the plume
with the immobile soil matrix. Such interactions include
desorption of reductants such as Fe2* from mineral surfaces,
direct and indirect surface redox reactions between Cr(VI) and
the mineral surfaces, and reduction by soil organic matter.
Thus, it is the soil matrix that is most important with regards to
redox transformations of chromium in the subsurface. This
argument  is further  supported by studies that clearly
demonstrate that groundwater contributes less that 1  % of the
oxidation capacities (equivalents of Cr oxidized per gram of
soil) and  reduction capacities (equivalents of Cr reduced per
gram of soil)  of aquifer systems while the soil matrix
contributes the remaining fraction (Barcelona and Helm,
1991).  Thus, any discussion of redox transformations of
chromium  in the subsurface must focus on the soil matrix.

Three key factors must be addressed in considering the
potential use  of natural attenuation of Cr(VI) in the subsurface.
Firstly, the reduction capacity of the aquifer, Rc, must  be great
enough to reduce all of the Cr(VI) that passes through it. If Xc
is the distance from the source to the point of compliance (Fig.
6), the  total mass of Cr(VI) from the source, MO, must be less
than the total mass  of Cr(VI), Mr, that can be reduced by the
aquifer material between the source and XC:
where A is the cross-sectional area of the plume normal to the
direction of groundwater flow and p,, is the dry bulk density of
the aquifer. As XJncreases, the mass of Cr(VI) that can be
reduced increases. A key difficulty in applying this criterion is
in  providing a reasonable estimate of M0. In the absence of
other reactions such  as adsorption or precipitation,  the
minimum  rate of movement of the Cr(VI) front through the
aquifer, v^, computed by assuming the reductant reacts
instantaneously with  the Cr(VI), is
                                                       (2)
                              eu c.
where pb and 8V are the dry bulk volumetric water content of
the porous medium, Vwis the velocity of the groundwater, and
C0is the concentration of the chromium in the contaminant
plume.
                                     CROSS-
                                   SECTIONAL   POINT OF
                                     AREA, A    COMPLIANCE
 Figure 6. Cr-plume moving from the source area to the point of
         compliance. The initial Cr(VI) concentration in the source
         area is MO, vw is the groundwater velocity, vmln is the velocity
         of the Cr(VI) front assuming instantaneous reduction of the
         Cr(VI), and Xcis the distance from the source area to the
         point of compliance.

The second key factor in the application of natural attenuation
of Cr(VI) is the rate of reduction relative to the rate of
advective transport in the  subsurface. The time for the
reduction reaction to decrease the concentration from its initial
concentration, CO,  to some target concentration, Cs, such as a
drinking water standard, should be less than the residence
time of the contaminated water in the portion of the aquifer
between the source of the Cr(VI) and the  point of compliance.
For example, if the rate of reduction of Cr(VI) follows a first-
order rate equation
                          dt
                                                       (3)
                    M 0< Mr =
                                                       (D
the time for the concentration of Cr(VI) to decrease from C. to
Cs must be less than the residence time of the contaminated
parcel of water within the aquifer

-------
                                                                         R0x = kox(K/K*)(A/Vr[H+]"+'
                                                       (9)
If natural attenuation is to be a viable option, this criterion
must be met. Difficulties in utilizing this criterion arise in
applying the appropriate rate equation and obtaining the
pertinent rate coefficients.

A third factor concerning the natural attenuation of Cr(VI) is
the possible oxidation of Cr(lll) to the more toxic hexavalent
form. While contamination is actively entering the subsurface,
conditions  may favor the reduction of Cr(VI)to Cr(lll). After
the source  of the  active contamination is removed, however,
chemical parameters within the aquifer, particularly pH, may
be altered. Under the new conditions, oxidation of Cr(lll) may
be favored. Thus, soil containing Cr(lll) formed during the
active contamination phase may  become a source of Cr(VI).

Both oxidation and reduction of chromium are occurring
simultaneously within the subsurface as part of a geochemical
cycle. As the Cr(lll) is oxidized to Cr(VI) by  manganese
dioxides in the soil,  Cr(VI) can be reduced to Cr(lll) by some
reductant such as soil organic carbon or pyrite. The rate of
change in [Cr(VI)] (d[Cr(VI)]/dt) is the sum of the rate of
reduction of Cr(VI),Rre0 and decreases when
Rred + Rox < O. Ultimately, the [Cr(VI)] will reach a steady state,
i.e., d[Cr(VI)]/dt = O. At this time, the rate of loss of Cr(VI) via
reduction is balanced  by the rate of production by the
oxidation of Cr(lll):

                       -Rred-  ROX                      (6)

Wittbrodt and Palmer (1  994) suggest that the reduction of
Cr(VI) by soil  fulvic acid can be represented by


              Rred = -kred Xe [HCrO;][SHS][H+]P          (7)

where [SHS] is the concentration of soil humic substance and
Xe denotes the equivalent fraction of the humic substance that
has been oxidized. Fendorf and Zasoski (1992) suggest that
CrOH2* is the reactive species in the oxidation of Cr(lll) by
MnO2. For illustrative  purposes, assume that the oxidation
reaction follows a rate equation of the form
            Rox = kox [CrOH2+] [A/ Vf [H+]n              (8)

where (A/V) denotes the surface area of the MnO2 per unit
volume  of solution. If we further assume that the solution is
equilibrated with Cr(OH)3(am), then
where Kw is the dissociation constant for H20 and K is the
equilibrium constant for the reaction


          Cr(OH)3(am)^Cr(OH)2+ + 2(OH")
Equating R  with -R  . and rearranging the terms yields
         [HCrO;] =
                                           __
                                            2
                               [SHS]
                                                           in+2-p
                                                                         (lo)
Although some of the specific points of rate equations
presented here are  debatable,  equation 10  does illustrate
aspects of natural attenuation in soils that contain both a
reductant and MnO2. The key point is that as long as the
supply of reductant  and MnO2 have not been significantly
depleted, [HCtO^] does not converge to zero with increasing
residence time within the aquifer as one would expect for a
first order reaction that only considers reduction of Cr(VI).
Rather, [HCKXJ converges to some steady-state
concentration that is> O that may or may  not be above the
MCL. This steady-state concentration increases with
increasing koj/kred> and (A/V) m/[SHS] and it varies with pH.
Thus, in principle, if the rate equations are correct and all of
the parameters are  known, one could calculate the steady-
state Cr(VI) concentration and determine  if natural attenuation
could achieve compliance goals. Studies of the kinetics of
these coupled processes needs be done to verify the general
forms of the rate equations and to determine the appropriate
rate coefficients.

Determining the  Potential for Natural Attenuation

If "natural attenuation" is to be considered an alternative to
expensive remediation efforts,  additional characterization is
required  to demonstrate that the expectations are likely to be
met. There is no single test that can tell us if natural
attenuation of Cr(VI) will occur at a particular site. Several
tests are briefly described which have been utilized to address
key factors affecting Cr(VI) transport in the subsurface and
describe how the results can be utilized in determining the
potential for the natural attenuation of Cr(VI) in the subsurface.

Ideally, it must be demonstrated that 1 ) there are natural
reductants present within the aquifer, 2) the amount of Cr(VI)
and other reactive constituents do not exceed the capacity of
the aquifer to reduce them, 3) the rate of Cr(VI) reduction is
greater than the rate of transport of the aqueous Cr(VI) from
the site, 4) the Cr(lll) remains immobile, and 5) there is no net
oxidation of Cr(lll)to Cr(VI). Some of these criteria are
relatively simple while others require additional tests and
interpretation. Additional tests that will be required include
tests of the oxidizing and reducing capacities of the aquifer.

-------
Mass ofCr(VI) at the Source

It must be demonstrated that the amount of Cr(VI) in the
aquifer does not exceed the capacity of the soil for reducing
this chromium. Therefore an important first step in evaluating
the potential for natural attenuation is to determine the mass
of Cr(VI) in the soil. Chromium exists in the subsurface either
in solution or in association with the  solid phase. Cr(VI) in
solution can be determined by the diphenylcarbazide (DPC)
method (APHA, 1989).  Aqueous samples are most often
obtained from monitoring wells. Alternatively, water separated
from the soil matrix either by centrifugation or by squeezing.
The pH of these waters should be measured to determine if it
is within the proper range (5.5 to 12)  to insure the Cr(lll)
concentrations are less than 1  \M (0.05 mg/L).

Cr(VI) associated with the soil  matrix maybe adsorbed to
mineral surfaces (particularly iron oxides) or precipitated  as
chromate minerals. There is no precise method for
determining each of these fractions of Cr(VI); nonetheless,
determinations have been  made using  sequential  extractions.
An initial water extraction serves to remove remaining pore
water and dissolve highly soluble chromium  minerals present
in the soil or that may have precipitated as the result of
evaporation during sample handling  and storage.  This water
extraction also removes some adsorbed ions.

Following the water extraction, a phosphate extraction is used
as a measure of the "exchangeable"  chromate in the soil
(Bartlett and James, 1988). The test is conducted by adding
phosphate to the soil and equilibrating for 24 hours. The
water is then separated from the slurry and Cr(VI)  is measured
by the DPC method (Bartlett and Kimble, 1976; Bartlett and
James, 1988c). The increase in the chromate concentration is
the amount of "exchangeable" chromate. Amacher and Baker
(1982) found optimal extraction using 0.01  M monobasic
potassium phosphate  (KH2PO4). James and Bartlett (1983b)
used a solution of 0.005 M KH2PO4 and 0.05 M K2HPO4 to
yield a pH of 7.2. James and Bartlett (1983b) stated that
doing the extraction at pH 7.2  is preferred because there is
less likelihood of chromate reduction  than at lower pH.
However, decreasing the pH of the soil slurry can result in
dissolution of BaCrO4from the soil. Moreover, if the pH of the
soil water was initially low,  then increasing the pH to 7.2 can
cause  precipitation of BaCrO4, thereby complicating the
interpretation of the results. When the soil water is not
equilibrated with BaCrO4> the phosphate extraction method of
James and Bartlett (1983b) primarily  measures the amount of
adsorbed Cr(VI) in  the soil. The phosphate  removes chromate
by both directly competing for the adsorption sites in the soil
and indirectly (in some cases)  by  increasing the pH.

BaCr04 is a likely chromate mineral phase that can be a
source of Cr(VI) in  contaminated aquifers. There is no direct
test for BaCrO4 in soils, however, when the groundwater is
equilibrated with this phase and the source of the Ba2* is
entirely from the clays in the natural soil, the maximum
amount of BaCrO4 in the aquifer is equal to the ammonium
acetate exchangeable  Ba2* (Thomas, 1982) in background
soils. For example, Palmer and Wittbrodt (1990) found that
the amount of exchangeable Ba2* was useful in estimating the
number of pore volumes required to flush Cr(VI) from soil
columns.

At many sites, the total Cr(VI) associated with the soil matrix is
the sum  of the BaCrO4 and the PO4-extractableCr(VI). This
sum, S, is often reported in units of mass per gram  of soil.
The total concentration of Cr(VI) in the soil, Cr(VI)tot is the sum
of the aqueous Cr(VI) and the matrix associated Cr(VI), S,
which can be reported in common units of mass per unit
volume of water by
             Cr(VI)tot = [Cr(VI)]
1000.QH.
  0,
(11)
where the dry bulk density of the soil, pb, is in g-cnr3. The
total mass of Cr(VI) in the site soils can then be estimated by
integrating the concentrations over the volume  of
contaminated soil.

Mass ofCr(lll)in the Subsurface

If all of the chromium that entered the soil was Cr(VI), then
demonstrating the presence of Cr(lll) in the soil would prove
that reduction is occurring. The mass of Cr(lll) in the soil can
provide a measure  of the amount of reduction that has
occurred. Although proof of chromate reduction is necessary,
it is not sufficient for demonstrating that natural attenuation will
adequately  protect the environment.

The total amount of Cr(lll) present in  the soil is the sum of the
mass in solution as well  and  mass associated with the solid
phase. Total chromium in solution can be determined by
atomic absorption spectrophotometry (AAS) or inductively
coupled plasma spectroscopy (ICP). When total chromium is
statistically greater than  Cr(VI),Cr(lll) can be simply
determined  by difference.

The amount of Cr(lll) associated with the soil matrix has
ostensibly been determined using several techniques. An
ammonium oxalate (0.1 M) extraction serves to remove
amorphous hydroxides of Cr,  Fe, and Al  (Ku et al., 1978;
Borggaard, 1988). Bartlett (1991) suggests that a K, H-citrate
extraction provides a measure of the Cr(lll)that is potentially
removable by low molecular weight organic molecules. A
dithionate-citrate-bicarbonate (DCB) extraction is conducted
by adding 0.3 M sodium  citrate and 0.1 M sodium bicarbonate
to the soil sample and heating to 80°C for 20 minutes. One
gram of sodium dithionate is then added and the soil slurry is
stirred for another 15 minutes. The DCB extraction removes
the crystalline forms of the Cr-, Fe-, and Al-oxyhydroxides (Ku
et al., 1978; Borggaard,  1988). The  dithionate reduces
crystalline iron (goethite) in the soil and Cr, Fe, and Al are
complexed  by the citrate. In addition  to the Cr(lll)
oxyhydroxides, the DCB  method also extracts sparingly
soluble Cr(VI) mineral phases such as BaCrO4, thereby
complicating interpretation of the results. Bartlett (1991) uses

-------
40 mL per gram of soil of 0.7 M NaOCI solution (undiluted
laundry bleach) at pH 9.5 to extract chromium. The slurry is
placed in a boiling water bath for 20 minutes before the liquid
is separated and  Cr is determined by AAS or ICP. This
method is useful in determining total chromium in the soil
because it readily oxidizes and removes Cr(lll)that is not
removed by other methods.

Identification of Potential Reductants

The presence of Cr(lll) in the soil maybe indicative of active
reduction in the soil, or it may be the result of the
neutralization of acidic waters containing Cr(lll) with
subsequent precipitation of chromium hydroxides.  Therefore,
identification of specific reductants within the aquifer is
warranted.  The identification of some potential reductants at a
site can be fairly simple in  some cases. For example, pyrite
(FeS2), a common constituent in many geological materials, is
readily identifiable by its visual characteristics. Other mineral
phases capable of reducing Cr(VI) can be identified using
classical  petrographic techniques or powder x-ray diffraction.
Scanning electron microscopy (SEM) can be utilized to identify
crystallite morphology. SEMS equipped with energy
dispersive x-ray spectroscopy can also provide information
about the elemental composition of these crystallite. Electron
diffraction patterns obtained from transmission electron
microscopes  provide crystallographic information. Such
electron microscopy methods can, however, be relatively
expensive. A fairly simple and inexpensive test for organic
carbon can provide a measure of the amount of carbon
available for reduction of Cr(VI). Knowledge of the specific
reductant within the aquifer is useful in determining the time
scale for the reduction of Cr(VI) based on studies that are
reported in the literature. Soils containing iron  sulfides or
organic matter are more likely to reduce Cr(VI) on the time
scales  of interest than soils containing ferrous iron  silicates.

Reduction Capacity of the Aquifer

Adequate protection of the environment by natural  attenuation
of Cr(VI)  requires that the soil possess  a large  enough
reducing  capacity to reduce all the hexavalent  chromium in the
source area. Several measures  for predicting reduction of
Cr(VI) in  soil are presented by Bartlett and James (1 988) and
Bartlett (1991). A measure of the maximum amount of Cr(VI)
that can be reduced per unit mass of aquifer, the "total Cr(VI)
reducing capacity", can  be obtained  using the classical
Walkley-Black method for determining soil organic carbon
(Bartlett and James, 1988). In this method, 2 to 3 grams of
soil are reacted with a mixture of 1 N K2Cr2O7 in IN H2SO4for
30 minutes (Walkley and Black,  1934; Nelson and Sommers,
1982; Bartlett and James,  1988). The Cr(VI) concentration is
measured using the diphenylcarbazide (DPC) method (APHA,
1989) and the decrease in the mass of Cr(VI) in the reaction
vessel  per gram of soil used in the test is the reduction
capacity. Although this method of determining  soil organic
carbon has its limitations (e.g. Nelson and Sommers, 1982), it
is a direct measure of how much Cr(VI) can be reduced by a
soil at extreme acid concentrations. Variations on this method
use heat or a combination of heat and pressure (Nelson and
Sommers, 1982). Barcelona and Helm (1991) used a
modified closed-tube chemical oxygen  demand procedure
(U.S.  EPA, 1979) to determine reduction capacities.

The extreme conditions of pH and temperature used in the
total Cr(VI) reducing capacity test may yield a greater reducing
capacity than would be available under  most environmental
conditions. The "available  reducing capacity"  test of
Bartlett and James (1988)  determined the reduction capacity
by reacting about 4 to 5 grams of moist soil in a solution of
10 mM H, PO, and K2Cr.,O7 for 18 hours. The H3PO4 is added
to buffer the pH and to compete with the Cr(VI) for the
adsorption sites. When KH2PO4 is used Bartlett and James
referred to it as the "reducing intensity". These tests are
designed to determine the reducing capacity at pH values
more  likely to be encountered in the field. However, in long
term reduction tests at near neutral  pH,  Palmer and Wittbrodt
(unpublished data) observed reduction  occurring  after 250
days (Fig. 7). Such long-term reduction tests are not practical
at most waste sites.
       0.08
            0     50  100  150  200    250   300
                          TIME (Days)

 Figure 7.  Cr(VI) reduced to Cr(lll) in Willamette silt loam.


 Oxidation Capacity

A potential limitation to the use of natural attenuation of Cr(VI)
in soil is the oxidation of the Cr(lll) to Cr(VI) by Mn02. If the
oxidizing capacity of the  soil is greater than the  reduction
capacity, then as the chromium  is  cycled in the soil it could
exhaust the soil reductant and be oxidized and ultimately
mobilized in the  soil. It is important, therefore, to determine
the capacity of the aquifer to oxidize Cr(lll).

Bartlett and James (1988) suggest a relatively simple test for
the amount of Cr(lll) that can  be oxidized by a soil. The
method involves  adding 2.5 grams of soil to a solution
containing 25 ml of 1 mM CrCI3. After shaking for 15 minutes,
a solution of KI-^PC^ . K2HPO4 is added to the  reaction
vessel, the slurry centrifuged or filtered, and the Cr(VI)

-------
 measured using the DPC method. Moist soils should be used
 in these tests. Drying the soils alters the surfaces of the
 manganese  dioxides making them less reactive (Bartlett and
 James,  1979).

 Barcelona and Helm (1991) used a solution of  chromous
 (Cr(ll)) ion to measure the oxidation capacity of soils. They
 added about 1 g of soil to cuvettes containing the Cr(ll)
 solution. The work was performed in a glove box to prevent
 oxygen from reacting with the Cr(ll). The cuvettes were
 sealed, shaken,  and allowed to react for 2 hours. The
 samples were centrifuged and the Cr(ll) measured
 spectrophotometrically. The loss of Cr(ll) is then used as a
 measure of the oxidation capacity of the soils.

 Each of these methods has some problems. Palmer and
 Wittbrodt (unpublished data) conducted  oxidation tests  similar
 to the Bartlett and James (1 988) method except that the
 concentration of Cr(VI)was monitored over nearly a year.
 Cr(VI) concentrations in these tests continued to  increase up
 to 100 days  and  may have been continuing after  300 days.
 For one soil, the short-term oxidation test of Bartlett and
 James underestimated the amount of oxidation obtained in the
 long-term tests by more than an order of magnitude. It is not
 clear in the Barcelona and Helm method whether the Cr(ll) is
 being oxidized to Cr(lll) or Cr(VI) or some combination of the
 two. If both products are forming, the results are more difficult
 to interpret.

 If the only mechanism for the oxidation of Cr(lll) in soils is
 oxidation by manganese oxides, then using extraction
 methods specifically designed for this purpose  may be a good
 way of determining the oxidation  capacity of the soils.  One
 very simple  extraction technique (Chao, 1972; Gambrell and
 Patrick, 1982) utilizes 0.1  M hydroxylamine hydrochloride
 (NH2OH . HCI) in 0.1 M HNO3. About 0.5 g of  soil  are added
 to 25 mL NH2OH . HCI of the solution and shaken for 30
 minutes and the Mn concentration is measured. The number
 of moles of Cr(lll) that can  be oxidized is then computed by
 dividing the  number of moles of Mn2+ per gram  of soil obtained
 in the extraction  by 1.5. The hydroxylamine hydrochloride test
 is fast, easy, and specifically targets the phase that promotes
 the oxidation of Cr(lll).

 Rates of Oxidation and Reduction

 Key factors in the suitability of natural attenuation as an option
 for chromium contaminated soils are the rates of  oxidation and
 reduction of chromium. This information is the  most difficult to
 obtain. Scientists are only now learning about the form of the
 applicable rate eauations and the aDDrooriate rate coefficients
that may apply. Such kinetic studies are an area where
 research has lagged behind the practical need  for the
 information.  While rates can be obtained from the technical
 literature, one must be careful to use rates of reduction for
 materials that are most likely controlling the Cr(VI) reduction  in
the site  soil.  In addition, because the rates depend on the
 concentration (surface area per liter of solution) of the
reductant and pH, it is important to obtain rate coefficients that
were acquired under conditions similar to those at the site.
Many rate studies have considered only a limited set of
conditions such as a single pH value or one reductant
concentration. Consequently, the reported rate coefficients
are apparent values that are strictly valid only under the
conditions of the experiment. Thus, the experimental factors
must be taken into account before the rate coefficients can be
applied  to field problems.

When MnO2 is present, Cr(lll) maybe oxidized back to Cr(VI)
and the net rate of reduction will be less than that obtained
from experiments that only utilize reductants. Further, many
rate experiments are conducted in stirred reactors that can
abrade  reactive surfaces. In soils, the rate of reaction may
become surface limited as adsorbed ions and precipitates
cover the  reactive surfaces.

One method of obtaining the net rate of reduction is through
tests on uncontaminated soils obtained from the site. These
soils should be similar to those through which the contaminant
plume will be migrating. Cr(VI) can be added to the soil slurry
and the Cr(VI) concentrations monitored over time. The
reaction vessels must exclude  light  to prevent photoreduction
reactions and the slurry must have the same pH as the
contaminant plume. A key limitation to such experiments is
that they require several months to a year to complete.

Estimating Reduction  from Monitoring Well Data

In principle, Cr(VI) reduction can be estimated from the
decrease in the mass of Cr(VI) in the aquifer (e.g., Henderson,
1994). The key difficulty in such an  approach is to  estimate
the mass of Cr(VI) using the aqueous concentrations. The
total mass of Cr(VI) in the aquifer is  the sum of the  mass that
is in solution, the mass that is adsorbed to the aquifer matrix,
and the mass that is precipitated within the aquifer. The  mass
of Cr(VI) in solution is obtained by integrating the Cr(VI)
concentrations over the volume of the contaminated aquifer
                     Maq=9vCV                     (12)
where V is the volume of aquifer containing a plume with a
Cr(VI) concentration of C.

The mass of Cr(VI) adsorbed to the soil matrix, Mads, can be
computed from the adsorption isotherm. For example, if
Cr(VI) follows a Langmuir isotherm, then over a volume of
aquifer, V, with constant aqueous Cr(VI) concentration, C, Ma
can be computed as
         M
'ads    Pb
Pmax
+><
"""KadsC V(i-ev
                                                     (13)
where Kads is the Langmuir adsorption constant, S^ is the
maximum amount of contaminant that can be adsorbed to the
soil.

-------
There is no unique amount of Cr(VI) precipitate for a given
hexavalent chromium concentration. Therefore it is
impossible to estimate mass of this fraction of Cr(VI) is the
subsurface using only the measured  concentrations in
monitoring wells. Thus, natural attenuation of Cr(VI)from
mass balances using monitoring well data can only be  used
when it can be reasonably demonstrated the Cr(VI)
precipitates cannot form within the aquifer.

Even when it is demonstrated that the formation of precipitates
within the aquifer is unlikely, there are inherent problems with
any monitoring system that can create uncertainties in  the
estimated mass of Cr(VI)  during a sampling round. In the
three-dimensional flow field, the highest concentrations from
one sampling period may  migrate between the discrete
monitoring points  of the next sampling round. The undetected
mass is not included in the total mass estimates in the  second
sampling round and may be mistakenly interpreted as  mass
loss due to Cr(VI)  reduction.

Summary

Under certain conditions, toxic Cr(VI) can be reduced to the
less toxic Cr(lll) in soils and precipitated as an insoluble
hydroxide phase.  The possibility of relying on such "natural
attenuation" of Cr(VI) is attractive because of the great
expense of remediating chromium contaminated sites. Before
such an option is adopted, however, it should be
demonstrated that natural attenuation is likely to occur under
the specific conditions at the site being investigated.

If natural attenuation is to be considered a viable option for
chromium contaminated sites, then ideally, it must be
demonstrated that 1) there are  natural reductants present
within the aquifer,  2) the amount of Cr(VI) and other reactive
constituents do not exceed the capacity of the aquifer to
reduce them, 3) the time scale required to  achieve the
reduction of Cr(VI) to the target concentration is less than the
time scale for the transport of the aqueous Cr(VI) from  source
area to the point of compliance, 4) the Cr(lll) will remain
immobile, and 5) there is no net oxidation of Cr(lll) to Cr(VI).
The most difficult information to obtain is the time scales for
the reduction and  oxidation of chromium in the soil.

Demonstrating Cr(VI) reduction  in aquifer by mass balances
that rely primarily  on the aqueous concentrations from
monitoring well networks are valid only if it is demonstrated
thatCr(VI) precipitates are not forming in the aquifer. The
monitoring network must be sufficiently dense that estimates
of Cr(VI) are accurate.

Several soil tests are described that are  useful in determining
the mass of Cr(VI) and Cr(lll) in the source areas and the
reduction and oxidation capacities of the aquifer materials.
Some simple conceptual models are  presented whereby this
information, combined with knowledge of the residence time  of
the chromium between the source and the point of compliance
can be used to determine  the feasibility of natural attenuation
of Cr(VI). The major limitation to this approach is the lack of
information about the rate of oxidation and reduction of
chromium under conditions likely to encountered by plumes
emanating from chromium sources. Without better information
about these rate processes under a wider range of conditions
with respect to pH, the use of the natural attenuation option for
contaminated  soils will continue to  be a highly debated issue.

References

Abbasi, S.A. and R.  Soni, 1984. Teratogenic Effects of
Chromium(VI) in the Environment as Evidenced by the Impact
of Larvae of Amphibian Randtigrina: Implications in the
Environmental Management of Chromium.  Int. J.
Environmental Studies,  23: 131-137.

Allison, J. D.,  D.S. Brown, K.J. Novo-Gradac, 1990.
MINTEQA2/PRODEFA2, A Geochemical Assessment Model
for Environmental Systems:  Version 3.0. U.S. Environmental
Protection Agency, Athens, GA.

Amacher, M.C. and D.E. Baker,  1982. Redox Reactions
Involving Chromium, Plutonium  and Manganese in Soils.
DOE/DP/0451 5-1. Institute for Research on Land and Water
Resources, Pennsylvania  State  University and  U.S.
Department of Energy, Las Vegas, NV.

Amonette, J.E. and D. Rai, 1990. Identification of
noncrystalline (Fe, Cr)(OH)3 by Infrared Spectroscopy. Clays
and Clay Minerals, 38(2):  129-136.

APHA, 1989.  Standard Methods for the Examination of Water
and Wastewater, 17th Edition, LS. Slesceri.A.E. Greenberg,
and R.R. Trussell (Editors). American Public Health
Association, Washington, D.C.

Barcelona, M.J. and T.R. Helm,  1991. Oxidation-Reduction
Capacities of Aquifer Solids. Environmental Science and
Technology, 25:1565-1572.

Bartlett, R. J., 1991.  Chromium  Cycling in Soils: Links, Gaps,
and Methods. Environmental Health Perspectives, 92: 17-24.

Bartlett, R.J. and J.M. Kimble, 1976.  Behavior of Chromium in
Soils:  11: Hexavalent Forms. J.  Environmental Quality,
5(4):383-386.

Bartlett, R.J. and B.R. James, 1979.  Behavior of Chromium in
Soils:  III.  Oxidation. J. Environmental Quality, 8(1):31-35.

Bartlett, R.J. and B.R. James, 1988.  Mobility and
Bioavailability of Chromium  in Soils. IN: Chromium in the
Natural and Human Environments, Vol. 20 (J.O. Nriagu and E.
Nieboer, editors). John Wley&  Sons, New  York: 267-306.

Beyersmann,  D.A. Koester,  B. Buttner, and P.  Flessel, 1984.
Model  Reactions of Chromium Compounds  with Mammalian
and Bacterial  Cells. Toxicol. Environ. Chem., 8: 279-286.
                                                         10

-------
Bianchi.V., A. Zantedeschi, A. Montaldi, and F. Majone, 1984.
Trivalent Chromium is Neither Cytotoxic nor Mutagenic in
Permealized Hamster Fibroblasts. Toxicological Letters, 23:
51-59.

Bloomfield, C. and G. Pruden, 1980. The Behavior of Cr(VI) in
Soil under Aerobic and Anaerobic Conditions. Environmental
Pollution (Series A), 23: 103-114.

Blowes, D.W. and C.J. Ptacek,  1992. Geochemical
Remediation of Groundwater by Permeable Reactive Walls:
Removal of Chromate by Reaction with Iron-Bearing  Solids.
Proceeding of the Subsurface Restoration Conference, June
21-24,  1992, Dallas, TX: 214-216.

Bonatti, S., M. Meini, and A. Abbondandolo, 1976. Genetic
Effects of Potassium Chromate in Schizosaccharomyces
pombe. Mutat. Res., 38: 147-149.

Borggaard, 0. K., 1988. Phase Identification by Selective
Dissolution Techniques. IN: Iron in Soils and Clay Minerals,
J.W. Stuck! et al. (editors). Reidel Publishing Co., pp. 83-98.

Boyko, S.L and D.M.L Goodgame, 1986. The Interaction of
Soil Fulvic Acid and Chromium (Vi) Produces Relatively Long-
Lived Water Soluble Chromium(V) Species. Inorg. Chim Acta,
123: 189-191.

Calder, L.  M., 1988. Chromium Contamination of
Groundwater. IN: Chromium in the Natural and Human
Environments, Vol. 20 (J.O. Nriagu and E. Nieboer, editors).
John Wley & Sons, New York:  215-230.

Chao.T.T, 1972.  Selective Dissolution of Manganese Oxides
from Soils and  Sediments  with Acidified Hydroxylamine
Hydrochloride.  Soil Science Society of America Proceedings,
36:764-768.

Eary, L.E. and D. Rai., 1989. Kinetics of Chromate Reduction
by Ferrous Ions Derived from Hematite and Biotite at25°C.
American Journal of Science, 289: 180-213.

Eary, L.E. and D. Rai, 1987. Kinetics of Chromium(lll)
Oxidation to  Chromium(VI) by Reaction with Manganese
Dioxides. Environmental Science and Technology,
21 (12):1187-1193.

Eary, L.E. and D. Rai, 1988. Chromate Removal from
Aqueous Wastes  by Reduction with Ferrous Iron.
Environmental Science and Technology, 22(8): 972-977.

Fendorf, S.E. and R.J. Zasoski, 1992. Chromium (III)
Oxidation by (3-MnO2. 1. Characterization. Environmental
Science and Technology, 26: 79-85.

Gambrell.R.P. and W.H. Patrick,  1982.  Manganese. IN:
Methods of Soil Analysis, Part 2,  Chemical and Microbiological
Properties, Second Edition. A.L.  Page (editor). Agronomy,
No. 9, Part 2, American Society of Agronomy, Soil Science
Society  of America, Madison,WI:313-322.

Goodgame, D. M. L., P.B. Hayman, and D.E. Hathway, 1984.
Formation of Water Soluble Chromium(V) by the Interaction of
Humic Acid and the Carcinogenic Chromium(VI). Inorg. Chim.
Acta, 91:1 13-115.

Henderson, T., 1994. Geochemical Reduction of Hexavalent
Chromium in the Trinity Sand. Ground Water 32(3): 477-486.

James, B.R. and Bartlett, R.J., 1983a. Behavior of Chromium
in Soils. VI. Interactions Between Oxidation-Reduction and
Organic Complexation. J. Environmental Quality 12: 173-176
(1983).

James, B.R. and Bartlett, R. J., 1983b. Behavior of Chromium
in Soils: VII. Adsorption and Reduction  of hexavalent forms. J.
Environmental Quality 12(2): 177-181.

Ku, H. F. H., B.G. Katz, D.J.Sulam.and R.K. Krulikas, 1978.
Scavenging of Chromium and Cadmium by Aquifer Material,
South Farmingdale Massapequa Area,  Long Island, New
York. Ground Water 16 (2): 112-118.

Lancy, L. E., 1966. Treatment of Spent Cooling Waters. U.S.
Patent  3,294,960.

Mancuso, T. F., 1951. Occupational Cancer and Other Health
Hazards in a Chrome Plant. A Medical Appraisal. II. Clinical
and Toxicological Aspects. Ind. Med. Surg. 20: 393-407.

Mancuso, T.F. and W.C. Heuper, 1951. Occupational Cancer
and Other Health Hazards in a Chrome Plant. A Medical
Appraisal.  L Lung  Cancers in Chromate Workers. Ind.  Med.
Surg., 20: 358-363.

Martin, C., D.R. Boone, and C.D. Palmer, 1994. Chromate-
Resistant Microbes from Contaminated Soil and Their
Potential for Bioaugmented Reduction of Cr(VI). Proceeding
of the Eighth National Outdoor Action Conference and
Exposition, Minneapolis, MN, May 23-25, 1994. National
Ground  Water Association: 191-204.

Nelson, D.W. and L.E. Sommers, 1982. Total Carbon, Organic
Carbon, and Organic Matter. IN: Methods of Soil Analysis,
Part2(R.C. Dinauer, editor), American Society of Agronomy,
Inc. and Soil Science Society of America, Inc., Madison, Wl:
539-580.

Nriagu, J. O., 1988a. Historical Perspectives. IN: Chromium
in the Natural and Human Environments, Vol. 20 (J.O. Nriagu
and E.  Nieboer, editors). John Wiley & Sons, New York: 1-20.

Nriagu, J. O., 1988b. Production and Uses of Chromium. IN:
Chromium in the Natural and Human Environments, Vol. 20
(J.O. Nriagu and E. Nieboer, editors). John Wiley& Sons,
New York: 81-104.
                                                        11

-------
One, B.-L, 1988. Genetic Approaches in the Study of
Chromium Toxicity and Resistance in Yeast and Bacteria. IN:
Chromium in the Natural and Human Environments, Vol. 20
(J.O. Nriagu and E. Nieboer, editors). John WileyS Sons,
New York: 351-368.

Palmer, C.D. and W. Fish, 1992. Chemically Enhanced
Removal of Metals from the Subsurface. Proceeding of the
Subsurface Restoration Conference, June  21-24, 1992,
Dallas, TX: 46-48.

Palmer, C.D. and P.R. Wittbrodt, 1990. Geochemical
Characterization of the United Chrome Products Site,  Final
Report. IN: Stage 2 Deep Aquifer Drilling Technical Report,
United Chrome Products Site, Corvallis, OR, September 28,
1990, CH2M Hill, Corvallis, OR.

Palmer, C.D. and P.R. Wittbrodt, 1991. Processes Affecting
the Remediation of Chromium-Contaminated Sites.
Environmental Health Perspectives, 92: 25-40.

Paschin, Y.V., V.I. Kozachenko,  and LE.  Sal'nikova, 1983.
Differential Mutagenic Response at the HGPRT  Locus  in V-79
and OHO Cells after Treatment with Chromate.  Mutat. Res.
122: 361-365.

Rai, D., B.M. Sass, and D.A. Moore, 1987. Chromium(lll)
Hydrolysis Constants and Volubility of Chromium(lll)
Hydroxide Inorg. 26(3): 345-249.

Rai, D., J.M. Zachara, LE. Eary, C.C. Ainsworth, J.E.
Amonette, C.E. Cowan, R.W. Szeimeczka, C.T.Resch.R.L.
Schmidt, S.C. Smith, and D.C. Girvin, 1988. Chromium
Reactions in Geologic Materials. Electric Power Institute Res.
Inst. Rept. EA-5741, Palo Alto, CA, 287 pp.

Richard, F.C. and A.C.M. Bourg, 1991. Aqueous
Geochemistry of Chromium: A Review. Water Research,
25(7): 807-816.

Riser, J.A. and G.W. Bailey, 1992. Spectroscopic Study of
Surface Redox Reactions with Manganese Oxides. Soil
Science Society of America Journal, 56: 82-88.

Sass, B.M. and D. Rai, 1987. Volubility of Amorphous
Chromium(lll)-lron( ill) Hydroxide Solid Solutions. Inorg.
Chem., 26(14): 2228-2232.

Schott, J. R.A. Berner and E. L. Sjoberg, 1981. Mechanism of
Pyroxene and Amphibole Weather!ng-1. Experimental
Studies with Iron-free Minerals. Geochimica et Cosmochimica
Act, 45:2123-2135.

Schroeder, D.C. and G.F. Lee, 1975. Potential
Transformations of Chromium in Natural Waters. Water, Air,
Soil Pollution, 4: 355-365.

Stollenwerk, K.G. and D.B. Grove, 1985.  Reduction of
Hexavalent Chromium in Water Samples Acidified  for
Preservation. J. Environmental Quality, 14(3): 396-399.
Swayambunathan,  V., Y.X.Liao, and D. Meisel, 1989. Stages
in the Evolution of Colloidal Chromium(lll) Oxide. Langmuir,
5(6): 1423-1427.

Thomas, G.W., 1982. Exchangeable Cations.  IN: Methods of
Soil Analysis, Part 2, Chemical and Microbiological Properties,
Second Edition. A.L. Page (editor). Agronomy, No. 9, Part 2,
American Society of Agronomy, Soil Science Society of
America, Madison, Wl: 159-165.

U.S. Environmental Protection Agency, 1979. Chemical
Oxygen Demand, EPA Method 410.4, Methods for Chemical
Analysis of Water and Wastes, U.S. Government Printing
Office: Washington, D.C.

van Weerelt, M., W.C. Pfeiffer, and M. Fiszman, 1984. Uptake
and Release of 51Cr(VI)and 51 Cr(lll) by Barnacles. (Balanus
sp). Mar. Environ. Res. 11 :201-211.

Walkley, A. and LA. Black, 1934. An Examination of the
Degtjareff Method for Determining Soil Organic Matter and a
Proposed  Modification of the Chromic Acid Titration Method.
Soil Science, 37:29-38.

Waterhouse, J.A.H., 1975. Cancer among Chromium Platers.
Br. J.  Cancer, 32: 262.

White, A.F. and M.F. Hochella, 1989. Electron  Transfer
Mechanisms Associated with the Surface Oxidation and
Dissolution of Magnetite and llmenite. Proc. 6th International
Symposium on Water-Rock Interaction:  765-768.

White, A.F. and A. Yee, 1985. Aqueous Oxidation-Reduction
Kinetics Associated with Coupled Electron-Cation Transfer
from Iron Containing Silicates at 25°C. Geochimica et
Cosmochimica Acta, 49:  1263-1275.

Wittbrodt, P.R. and C.D. Palmer, 1992.  Limitations to Pump-
and-Treat Remediation of a Chromium Contaminated Site.
Aquifer Restoration: Pump-and-Treat and the Alternatives.
National Ground Water Association  National Convention, Las
Vegas, NV, Sept. 30- Oct. 2, 1992.

Wittbrodt, P.R. and C.D. Palmer, 1994. Reduction of Cr(VI) in
the Presence of Excess Soil Fulvic Acid. Environmental
Science and Technology,  (accepted for publication).

Yassi, A. and E. Nieboer,  1988. Carcinogenicity of Chromium
compounds. IN: Chromium in the Natural and Human
Environments  (J.O. Nriagu and E. Nieboer, editors): 443-496.
                                                        12

-------