Impact of brownmillerite hydration on Cr(VI) sequestration in chromite ore processing residue

Abstract

Experimental and modeling studies were conducted to delineate the reaction progress of chromite ore processing residue (COPR)
upon hydration and the roles of brownmillerite and calcium aluminum chromium oxide hydrates (CAC) in the scavenging of hexavalent
chromium. A kinetic study was conducted by preparing slurry samples with both synthetic brownmillerite and actual COPR samples
at ambient temperatures. The hydration reaction of brownmillerite using synthetic brownmillerite was very fast (within 1 hour)
and was completed within 2 days. However, the hydration of brownmillerite embedded in COPR to its hydration byproducts was
not clearly observed after 7 days of aging. Newly formed Ca4Al2O6(CrO4)·14H2O (CAC-14) was observed after 1 hour of aging in both samples. However, the rate of formation of CAC-14 with synthetic brownmillerite
was much faster than the COPR embedded brownmillerite. The reaction progress of synthetic brownmillerite and COPR upon chromate
influx was simulated by a reaction path modeling program. The phase transformation of both samples can be predicted by the
constructed model. Moreover, the formation of CACs upon chromate addition was predicted by the model, suggesting an effective
sink for Cr(VI).

Key wordschromite ore processing residue–brownmillerite–calcium aluminum chromium oxide hydrates–hexavalent chromium

Full text

Geosciences Journal
Vol. 15, No. 3, p. 287
− 296, September 2011
DOI 10.1007/s12303-011-0023-y
ⓒ The Association of Korean Geoscience Societies and Springer 2011
Impact of brownmillerite hydration on Cr(VI) sequestration in chromite
ore processing residue
ABSTRACT: Experimental and modeling studies were conducted
to delineate the reaction progress of chromite ore processing res-
idue (COPR) upon hydration and the roles of brownmillerite and
calcium aluminum chromium oxide hydrates (CAC) in the scav-
enging of hexavalent chromium. A kinetic study was conducted by
preparing slurry samples with both synthetic brownmillerite and
actual COPR samples at ambient temperatures. The hydration
reaction of brownmillerite using synthetic brownmillerite was very
fast (within 1 hour) and was completed within 2 days. However,
the hydration of brownmillerite embedded in COPR to its hydra-
tion byproducts was not clearly observed after 7 days of aging. Newly
formed Ca
4
Al
2
O
6
(CrO
4
)·14H
2
O (CAC-14) was observed after 1 hour
of aging in both samples. However, the rate of formation of CAC-
14 with synthetic brownmillerite was much faster than the COPR
embedded brownmillerite. The reaction progress of synthetic
brownmillerite and COPR upon chromate influx was simulated
by a reaction path modeling program. The phase transformation
of both samples can be predicted by the constructed model. More-
over, the formation of CACs upon chromate addition was pre-
dicted by the model, suggesting an effective sink for Cr(VI).
Key words: chromite ore processing residue, brownmillerite, calcium
aluminum chromium oxide hydrates, hexavalent chromium
1. INTRODUCTION
Millions of tons of chromite ore processing residue
(COPR) which is an alkaline material (pH > 12) were depos-
ited over a span of 50 years at numerous urban areas in the
USA, the UK, and elsewhere in the world (Burke et al.,
1991; Weng et al., 1994; James et al., 1996; Farmer et al.,
1999; Graham et al., 2006). There are more than 200 active
COPR sites in Hudson County, New Jersey, USA (Wazne et
al., 2007). COPR was beneficially used as structural fill
because of its favorable structural quality as a granulated
material. The COPR was generated by the high-temperature
roasting process of chromite ore with lime in order to extract
chromium as soluble sodium chromate. Specifically, the
roasting process was conducted at approximately 1200 °C
and the lime was added to act as a mechanical separator
allowing oxygen to react with the chromite and sodium car-
bonate. Lime also served as a sequestering agent, combin-
ing with various ore impurities to form insoluble compounds
(Allied Signal, 1982). The sodium chromate formed during
the roasting process was extracted with hot water as a weak
yellow liquor solution. The sodium chromate was then con-
verted into sodium dichromate by reaction with sulfuric
acid. After draining, the residue was discarded. The disposed
COPR contains unreacted chromite ore and un-extracted
chromate. Even though the high lime process has ceased in
the USA and the UK, it is still being used in China, Russia,
Kazakhstan, India and Pakistan (Darrie, 2001).
Chromium exists mainly in two oxidation states as
hexavalent [Cr(VI)] and trivalent chromium [Cr(III)] in common
environments. Cr(VI) is highly mobile, severely toxic at mod-
erate doses, and classified as a respiratory carcinogen in
humans. In contrast, Cr(III) is used as a dietary element at
low doses, and in most environmental systems is immobile
(Higgins et al., 1998). The COPR, which contains both Cr(III)
and Cr(VI), was discovered not as benign as initially
thought; yellow chromate solutions are often observed to
leach from locations where COPR is deposited and elevated
Cr(VI) concentrations are measured in ground water and
water bodies in the proximity of these sites (Geelhoed et al.,
2002). Additionally, structures built on sites where COPR
was used as fill experienced catastrophic failures due to
heave and uncontrolled expansion of the COPR material.
Consequently, COPR has become a major geoenvironmen-
tal and geotechnical hazard in many urban areas.
The main solid phases that comprise COPR at the New
Jersey site are brownmillerite (Ca
4
Fe
2
Al
2
O
10
), periclase (MgO)
and quicklime (CaO) with minor impurities (Chrysochoou et
al., 2005b). Moreover, approximately 31% of brownmillerite
was observed among those phases (Jagupilla et al., 2009).
Brownmillerite is known to be unstable under environmental
aqueous conditions and should hydrate to hydrogarnet
[Ca
3
Al
2
(OH)
12
], portlandite [Ca(OH)
2
] and hematite [Fe
2
O
3
]
according to the following reaction:
Ca
4
Fe
2
Al
2
O
10
+7 H
2
O
Brownmillerite
→ Ca(OH)
2
+Fe
2
O
3
+ Ca
3
Al
2
(OH)
12
. (1)
Portlandite Hematite Hydrogarnets
Deok Hyun Moon*
Mahmoud Wazne
Department of Environmental Engineering, Chosun University, Gwangju 501-759, Republic of Korea
W.M. Keck Geoenvironmental Laboratory, Center for Environmental Systems, Stevens Institute of Tech-
nology, Hoboken, NJ 07030, USA
*Corresponding author: dmoon10@hotmail.com

288 Deok Hyun Moon and Mahmoud Wazne
However, brownmillerite is still present at numerous sites
some 80 years after its deposition. It has been reported that
the major COPR phases (brownmillerite, periclase, quicklime)
are unstable in aqueous environments and eventually must
react entirely to form hydration products (Moon et al., 2005b).
Among these three phases, only lime hydrates quickly to
form portlandite whereas brownmillerite and periclase react
much slower to form their respective hydration products.
Especially, brownmillerite hydration in actual COPR material
is extremely slow at ambient temperature (Moon et al.,
2008). This could be due to the iron coating on the actual
COPR which delays the dissolution process (Taylor, 1997).
Calcium aluminum chromium oxide hydrates [Ca
4
Al
2
O
6
(CrO
4
)·nH
2
O, CACs] were also identified in COPR mate-
rials as the main reaction products of brownmillerite. The
CACs could be found at different hydration states (9, 12, or
14 H
2
O) (Chrysochoou et al., 2005a). The transformation of
brownmillerite to CAC-14 [Ca
4
Al
2
O
6
(CrO
4
)·14H
2
O] upon
chromate influx can be written as:
Ca
4
Fe
2
Al
2
O
10
+CrO
4
2–
+16H
2
O+2H
+
→
Ca
4
Al
2
O
6
(CrO
4
)·14H
2
O+2Fe(OH)
3
. (2)
In COPR materials, Cr(VI) is found mainly in CACs and
is partially present in hydrotalcites [Mg
6
Al
2
(CO
3
)(OH)
16
·
4H
2
O] and hydrogarnets through anionic substitution. These
minerals could be embedded in brownmillerite (Kremser et al.,
2005), and therefore, the rate of Cr(VI) release to the environ-
ment would be dependent on the rate of brownmillerite hydration.
Therefore, the hydration pathway of brownmillerite in
COPR during long-term weathering is not well understood. It
is worth studying the hydration pathway of synthetic brown-
millerite which does not have an iron coating or other impu-
rities to determine the time needed for hydration to occur. This
information can be used as reference for long-term hydration
of brownmillerite in COPR at ambient temperature. Also,
hydration of brownmillerite is needed in order to understand
Cr(VI) sequestration in COPR and the reaction pathway of
brownmillerite to form CAC compounds upon hydration.
The purpose of this study was to investigate brownmil-
lerite hydration reaction pathways using synthetic brown-
millerite and field COPR samples to assess the hydration
process under controlled conditions. The hydration of
brownmillerite was also investigated in the presence of
chromate in order to examine Cr(VI) sequestration in COPR
materials. Reaction path kinetics geochemical modeling
using the geochemical modeling code, EQ3/6 (Wolery and
Jarek, 2003), was used to simulate the synthetic brownmil-
lerite and the COPR-water interactions.
2. EXPERIMENTAL METHODOLOGY
2.1. Materials and Reagents
Synthetic brownmillerite samples and COPR samples
containing high brownmillerite concentrations were used to
investigate the hydration of brownmillerite and the formation
of CAC. The synthetic brownmillerite was obtained from
CTL (Skokie, IL) with a mean particle size of 20 µm. The
COPR samples were obtained from Study Area 7 (SA 7),
located in Hudson Country, New Jersey, USA. All the COPR
samples were pulverized and sieved by hand through a 100-
mesh sieve (0.15 mm diameter opening). Deionized (DI)
water was all used in this study. Potassium chromate
(K
2
CrO
4
) salts obtained from Fisher Scientific (Suwanee,
GA) dissolved in DI water were used as a chromate source
for the investigation of CAC formation.
2.2. Slurry Sample Preparation
In order to investigate katoite [Ca
3
Al
2
(SiO
4
)
3–x
(OH)
4x
(x =
1.5 to 3.0), a hydrogarnet] and CAC formation upon
brownmillerite hydration, slurry samples were prepared
using 10 grams of either synthetic brownmillerite or actual
COPR material with DI water and potassium chromate
solution using a 2:1 liquid to solid ratio. The liquid to solid
ratio of 2 to 1 was used to provide sufficient water and to
promote full hydration (Moon et al., 2005a). The concen-
tration of chromate was 1 wt% of the dry sample, to rep-
resent the average Cr(VI) concentration in COPR. The
method of slurry sample preparation used for CAC formation
was similar to the method used for ettringite formation
where sodium sulfate decahydrate (Na
2
SO
4
10H
2
O) was used
as a sulfate source (Moon et al., 2005a). The time intervals
for aging at ambient temperature were 1 hour, 8 hours, 1
day, 2 days, and 7 days. All of the slurry samples were con-
tinuously mixed for 1 hour using a Toxicity Characteristic
Leaching Procedure (TCLP) tumbler at 30 rpm. After 1 hour
of continuous mixing, samples were periodically shaken
during aging.
2.3. X-ray Powder Diffraction (XRPD) Analyses of the
Slurry Samples
About one gram of homogenized slurry sub-samples
were drawn after the designated curing time using a 5 mL
pipette. The samples were then filtered using a 0.4 µm pore
size membrane filter. The residue was analyzed using XRPD.
Step-scanned X-ray patterns were collected with a Rigaku
DXR-3000 computer-automated diffractometer. XRPD anal-
yses were conducted at 40 kV and 40 mA using a diffracted
beam graphite-monochromator with Cu radiation. XRPD
data was collected in a range of 5° to 65° 2θ with a step size
of 0.02° and a count time of 3 seconds per step. Qualitative
analysis of XRPD patterns were conducted using the JADE
software version 7.1 (MDI, 2005) using the International
Center for Diffraction Data database, (ICDD, 2002) refer-
ence patterns.

Impact of brownmillerite hydration on Cr(VI) sequestration in chromite ore processing residue 289
2.4. Mineralogical Characterization of the COPR Sample
The initial COPR sample was homogenized and air dried
for 24 hours. Two grams of the homogenized air dried sam-
ple was pulverized by a McCrone micromill for 10 min
together with 7 mL cyclohexane. The resulting slurry was
air dried then mixed with 20% by weight of corundum (α-
Al
2
O
3
) as an internal standard. The sample was then sub-
jected to XRPD analysis. The qualitative and quantitative
analyses of the XRPD patterns were performed using the
Jade software (MDI, 2005), version 7.1 and the Whole Pat-
tern Fitting function of Jade, which is based on the Rietveld
method (Rietveld, 1969). The reference databases for pow-
der diffraction and crystal structure data were the Interna-
tional Center for Diffraction Data database, (ICDD, 2002)
and the Inorganic Crystal Structure Database (ICSD, 2005),
respectively.
2.5. Scanning Electron Microscopy (SEM) Analyses
Prior to SEM analysis, all samples were dried for 3 days
in a nitrogen-purged chamber and prepared using double-
side carbon tape. SEM analyses were conducted using a
LEO-810 Zeiss microscope equipped with an ISIS-LINK
system.
2.6. Geochemical Modeling
Geochemical modeling was used to simulate the reaction
progress of brownmillerite and COPR. The modeling results
can be used to assess the experimental results and to predict
the hydration progress of brownmillerite and COPR. The
reaction progress of synthetic brownmillerite and COPR in
the presence of a chromate solution was simulated by using
the EQ3/6 software package, version 8.0 (Wolery and Jarek,
2003). In this modeling package the reactants are added
incrementally and thermodynamic equilibrium is enforced
at every incremental step. This modeling procedure allows
the reactants to have different reaction rates and facilitates
the assessment of the reaction products at each incremental
step and not necessarily only at equilibrium.
3. RESULTS
The transformation of synthetic brownmillerite to katoite
was clearly observed after 1 hour of continuous mixing, and
was completed after 2 days of curing (Fig. 1). The SEM
images of synthetic brownmillerite (initial condition) and
Fig. 1. XRPD patterns of hydrating synthetic brownmillerite for curing time up to 7 days.

290 Deok Hyun Moon and Mahmoud Wazne
katoite (2 days of curing) are presented in Figures 2 and 3,
respectively. In addition to katoite, paraalumhydrocalcite
(CaAl
2
(CO
3
)
2
(OH)
4
·6H
2
O) and iron carbonate hydroxide
[Fe
6
(OH)
12
(CO
3
)] were also identified.
Compared to that of synthetic brownmillerite, the major
phases identified in the COPR samples before and after
hydration were brownmillerite, periclase, brucite, katoite,
hydroandradite, quartz, Ca
4
Al
2
(OH)
12
CrO
4
·12H
2
O (CAC-12),
CAC-14 and stichtite [Mg
6
Cr
2
CO
3
(OH)
16
·4H
2
O] (Fig. 4).
Upon hydration, the transformation of brownmillerite was
not clearly observed within 7 days (Fig. 4).
Fig. 2. SEM image of synthetic brownmillerite as received (irreg-
ular shape).
Fig. 3. SEM image of katoite (hexagonal in shape) derived from
synthetic brownmillerite hydration (2 days of curing).
Fig. 4. XRPD pattern of the actual COPR materials upon hydration.

Impact of brownmillerite hydration on Cr(VI) sequestration in chromite ore processing residue 291
Upon addition of chromate to synthetic brownmillerite,
the major phases identified after 1 hour of continuous mix-
ing were brownmillerite, CAC-14, bernalite [Fe(OH)
3
]
and calcium aluminum iron carbonate hydroxide hydrate
[Ca
8
Al
2
Fe
2
O
12
CO
3
(OH)
2
22H
2
O, CAFC] (Fig. 5). The SEM
image of the transformed CAC-14 upon chromate influx is
presented in Figure 6. Katoite formation was observed with
minor peaks after 1 day of aging.
Upon the addition of chromate to the COPR samples,
no obvious changes in peak intensities for most of the
phases were identified, except for portlandite and CAC
compounds. Portlandite dissolved and newly formed CAC-14
was observed, after 1 hour of aging (Fig. 7).
4. DISCUSSION
Paraalumhydrocalcite was identified upon the hydration
of synthetic brownmillerite. It was stable up to 8 hours of
aging, but then it dissolved after 1 day of aging. One pos-
sible explanation is that paraalumhydrocalcite is a meta-sta-
ble phase, and that the Ca
2+
and Al
3+
from this phase would
be consumed to form katoite. However, it should be noticed
that this carbonate phase was more thermodynamically
favored than calcite because no calcite was observed after
7 days. Even though it has been reported that iron does not
Fig. 5. XRPD pattern of synthetic brownmillerite upon chromate addition.
Fig. 6. SEM image of CAC-14 (hexagonal plates) formed by
exposing synthetic brownmillerite to K
2
CrO
4
solution (7 days).

292 Deok Hyun Moon and Mahmoud Wazne
form phases separately during brownmillerite hydration
(Chatterji and Jeffery, 1962), iron carbonate hydroxide was
stable throughout the experiment (7 days), suggesting that
iron liberated by the synthetic brownmillerite hydration pre-
cipitated to form iron carbonate hydroxide though released
iron may have also resulted in the formation of amorphous
iron compounds.
It appears that CAC-12 was transformed into CAC-14
upon the hydration of the COPR samples. It has been reported
that CAC hydration states can be fluctuated depending on
available moisture (Chrysochoou et al., 2005). No apparent
changes in the peak intensity for brownmillerite and katoite
were observed upon the hydration of the COPR material.
This indicated that the dissolution of brownmillerite in the
COPR samples is very slow. This could be due to the iron
coating on the COPR samples which may hinder or slow
down the dissolution process (Taylor, 1997). Research also
showed that Mg
2+
substitution for Ca
2+
in brownmillerite
results in a slower hydration process (Jupe et al., 2001).
Alternatively, this could be due to the relatively larger par-
ticle size of the COPR samples used in this study compared
to the synthetic brownmillerite samples (150 µm versus 20
µm). However, if the relatively larger particle size was the
cause, then some brownmillerite hydration should have
been observed after aging the samples for 7 days.
Upon chromate influx to synthetic brownmillerite, CAC-
14 was identified after 1 hour of mixing. The formation of
CAC-14 is mainly due to the reaction mentioned in Equation
2. The SEM image of transformed CAC-14 upon chromate
influx (Fig. 6) indicated that the Fe
3+
liberated by the syn-
thetic brownmillerite accumulated as bernalite and CAFC,
while Ca
2+
was converted to CAC-14 and CAFC. Moreover,
the quantities of CAC-14 and CAFC increased upon aging.
However, it was not clear whether the peak intensities for
bernalite increased or whether this was due to small peak
intensities. Katoite formation observed with minor peaks
after 1 day of aging suggested that CAC, bernalite and CAFC
formation was thermodynamically favored over katoite upon
chromate addition. Moreover, it was noticed that brown-
millerite was not completely transformed into other phases
at 7 days of aging, suggesting that the reaction is slow but
ongoing.
Upon chromte influx to the actual COPR material, Port-
landite dissolved and CAC-14 was newly formed (Fig. 7)
after 1 day of aging. This phase formation is mainly due to
hydration. Moreover, CAC-14 intensities increased with
Fig. 7. XRPD pattern of the actual COPR materials upon chromate addition.

Impact of brownmillerite hydration on Cr(VI) sequestration in chromite ore processing residue 293
curing, indicating that chromate addition mainly contributed
to the formation of this phase.
Overall, following chromate addition to the synthetic
brownmillerite and the COPR samples, a newly formed CAC-
14 compound was observed in both samples. However, the
rate of CAC-14 formation was much faster in synthetic
brownmillerite than in the COPR samples. No apparent peak
reductions of brownmillerite were observed in the COPR
samples, indicating that synthetic brownmillerite was more
thermodynamically or kinetically unstable than brownmil-
lerite in the COPR samples.
5. GEOCHEMICAL MODELING
The mineralogical characterization of the COPR sample
indicated a high brownmillerite content at approximately
39% of the crystalline phases (Table 1). Even though brown-
millerite is not stable thermodynamically under aqueous
conditions it is still present at the site some 80 years after
its deposition. During the model simulation, the rate of
brownmillerite hydration was assumed two orders of mag-
nitude slower than the reaction rate of the other minerals.
This was incorporated into the model using the relative
kinetic rates feature of EQ3/6. Similarly, periclase is also
not stable at the site conditions but its slow hydration
explains its presence at the site. Its reaction rate was treated
similar to that of brownmillerite. The amorphous content
was determined at 30% using a corundum internal standard.
This amorphous fraction was assumed to consist of the iron
oxide hydroxide phase and goethite during the simulation
runs.
There is no separate thermodynamic data for CAC-12
and CAC-14 in the thermodynamic database of EQ3/6 or in
the surveyed literature. Therefore, Cr(VI)-hydrocalumite
thermodynamic data was used to represent both of these
minerals during the model simulation. Similarly, there is no
thermodynamic data for stichtite but the Cr(III)-hydrotalcite
phase and Al-hydrotalcite thermodynamic data were used
instead. It is also worth mentioning that cationic and anionic
substitution may exist and some solid phases may be solid
solutions rather than pure solids; however, the extent of
these substitutions was not investigated in this study.
The model predicted the transformation of brownmillerite
to katoite, portlandite, hematite and Cr(VI)-Hydrocalumite
(a CAC). As shown in Figure 8, these phases formed as
soon as brownmillerite started to dissolve. The model
differed from the experimental data by not predicting the
formation of bernalite and calcium aluminum iron carbon-
ate hydroxide hydrate. It is no surprise that these two phases
were not predicted by the model since their thermodynamic
data is not present in the thermodynamic data base of EQ3/
Table 1. The percentages of the crystalline phases for the initial
COPR sample as quantified by Rietveld analysis
Species Molecular formula Percent
Brownmillerite
Pordlandite
Calcite
Periclase
Brucite
Katoite
Hydrotalcite
Hydrocalumite
Quartz
Ca
2
(Al, Fe)O
5
Ca(OH)
2
CaCO
3
MgO
Mg(OH)
2
Ca
3
Al
2
(SiO
4
)
1.5
(OH)
6
Mg
6
Al
2
(CO
3
)(OH)
16
4H
2
O
3CaO Al
2
O
3
CaCrO
4
14H
2
O
SiO
2
39
7.8
8.1
3.6
9.3
16.6
7.8
3.1
4.6
Fig. 8. EQ3/6 simulation of the reaction progress of 500 g of synthetic brownmillerite with chromate solution with chromate content
at 1 percent of dry weight brownmillerite and liquid to solid ratio of 2.

294 Deok Hyun Moon and Mahmoud Wazne
6, which is the most reliable and extensive mineral database
available. Furthermore, no thermodynamic data was avail-
able in the surveyed literature for these two phases. One
model simulation was run without chromate (simulation not
presented) and the model predicted katoite, portlandite and
hematite. During this run, paraalumhydrocalcite and iron
carbonate hydroxide were not predicted by the model due to
the unavailability of their thermodynamic database in the
EQ3/6 thermodynamic data base in a similar fashion to the
first simulation.
The presence of CAFC and paraalumhydrocalcite in the
experimental data indicated a carbonate source; however,
carbonate was not added to the system. This may be caused
by atmospheric carbon dioxide contamination enhanced by
the high alkaline pH of the slurry. Conversely, when 0.1%
of carbonate was added in the model simulation, calcite was
predicted as the carbonate phase (simulation not shown)
rather than CAFC and paraalumhydrocalcite. As for the
iron phases, the experimental data showed the sequestration
of iron in bernalite and iron carbonate hydroxide whereas
the model predicted the formation of hematite. Bernalite and
iron carbonate hydroxide thermodynamic data were also not
available in the EQ3/6 thermodynamic database.
As soon as chromate was introduced into the system, all
of the chromate (0.043 moles, 1% w/w) was sequestered
immediately by the newly formed Cr(VI)-hydrocalumite.
Cr(VI)-hydrocalumite remained stable throughout the reac-
tion progress of the hydration of brownmillerite. This is sig-
nificant as the model predicted that Cr(VI)-hydrocalumite,
which is a hydration byproduct of the dissolution of brown-
millerite, is expected to scavenge all of the chromate in the
system constrained only by stoichiometric limitations.
For the reaction progress of COPR in the presence of a
chromate solution, the model predicted the complete disso-
lution of quartz, periclase, brownmillerite, brucite and goet-
hite (the surrogate amorphous phase). The new phases
predicted by the model are andradite (Ca
3
Fe
2
Si
3
O
12
) and
hematite. The model predicted an increase in the content of
portlandite, katoite and hydrotalcite. However, the portlan-
dite content decreased initially at reaction progress less than
10% before it started to increase again. Portlandite may have
served as an immediate calcium source for the formation of
andradite and Cr(VI)-hydrocalumite before the start of the
dissolution of brownmillerite. As soon as a significant per-
centage of brownmillerite dissolved the percentages of port-
landite and katoite started to increase.
The model also predicted the increase in the concentra-
tion of Cr(VI)-hydrocalumite upon the introduction of the
chromate solution which is similar to the experimental results
in Figure 6. The total number of moles of chromate (0.059)
was sequestered in Cr(VI)-hydrocalumite and remained sta-
ble throughout the simulation, similar to the hydration of
synthetic brownmillerite in the presence of a chromate solu-
tion in the previous section.
The model simulation of COPR hydration without the
addition of a chromate solution (simulation not shown) was
similar to the run with the chromate solution except for the
increase in the content of Cr(VI)-hydrocalumite and a smaller
drop in the portlandite concentration upon chromate solu-
tion addition, which indicates that the increase in Cr(VI)-
hydrocalumite content in the previous simulation upon the
addition of chromate partially derived its calcium from port-
Fig. 9. EQ3/6 simulation of the reaction progress of 500 g of COPR with chromate solution with chromate content at 1 percent of dry
weight COPR and liquid to solid ratio of 2.

Impact of brownmillerite hydration on Cr(VI) sequestration in chromite ore processing residue 295
landite. Also, during the same run, katoite content increased
slightly at a reaction progress of 10% instead of the small
drop shown in Figure 9, which indicates that katoite may
have served as the aluminum source for the formation of
Cr(VI)-hydrocaulmite.
6. CONCLUSIONS
The transformation of brownmillerite to katoite and CAC
compounds was specifically investigated using synthetic
brownmillerite and actual COPR materials. DI water and a
chromate solution were introduced to synthetic brownmil-
lerite and actual COPR materials to investigate the reaction
pathways of katoite and CAC compounds, respectively. The
geochemical model was used to predict the product phases
of COPR upon complete hydration. Upon hydration of syn-
thetic brownmillerite, katoite formation was thermodynam-
ically favored after 1 hour of continuous mixing. The trans-
formation of brownmillerite to katoite was completed after
2 days of aging. However, no major changes in katoite peak
intensities in the actual COPR materials were observed
within 7 days of aging. This indicated that the dissolution of
brownmillerite in the actual COPR materials was very slow.
Upon chromate influx to the synthetic brownmillerite and
the actual COPR materials, CAC-14 was identified in both
samples after 1 hour of continuous mixing. The peak inten-
sities of CAC-14 increased upon aging in both samples.
However, the rate of CAC-14 formation was much faster in
synthetic brownmillerite than in the actual COPR material,
suggesting that the dissolution of brownmillerite in the
actual COPR material was slower than in synthetic brown-
millerite. The geochemical model predicted the formation
of Cr(VI)-hydrocalumite upon the introduction of the chro-
mate solution, and more importantly, the model predicted
that the byproducts of the hydration of brownmillerite and
COPR would scavenge chromate and thus control its sol-
ubility.
REFERENCES
Allied Signal, 1982, Process Description-Baltimore Works.
Burke, T., Fagliano, J., Goldoft, M., Hazen, R.E., Iglewicz, R., and
McKee, T., 1991, Chromite ore processing residue in Hudson
County, New Jersey. Environ Health Perspect, 92, 131–137.
Chatterji, S. and Jeffery, J.W., 1962, Studies of early stages of paste
hydration of cement compounds. Journal of the American
Ceramic Society, 45, 536–543.
Chrysochoou, M., Dermatas, D., Moon, D.H., and Wazne, M.,
2005a, Role and properties of calcium aluminum oxide hydrates
in chromium waste stabilization. In: Proceedings of the 1
st
International Conference on Environmental Science and Technology,
New Orleans, Louisiana, January 23–26.
Chrysochoou, M., Moon, D.H., Dermatas, D., Wazne, M., Christod-
oulatos, C., Meng, X., Kaouris, M., Morris, J., French, C., and
Sass, B., 2005b, Mineralogical analysis of Chromite Ore
Processing Residue by X-ray Powder Diffraction. In: Proceed-
ings of the 8
th
International In Situ and On-Site Bioremediation
Symposium, Baltimore, Maryland, June 6–9.
Darrie, R.G., 2001, Commercial extraction technology and process
waste disposal in the manufacture of chromium chemicals from
ore. Environmental Geochemistry and Health, 23, 187–193.
Farmer, J.G., Graham, M.C., Thomas, R.P., Licona-Manzur, C.,
Paterson, E., and Campbell, C.D., 1999, Assessment and mod-
elling of the environmental chemistry and potential for remedi-
ative treatment of chromium-contaminated land. Environmental
Geochemistry and Health, 21, 331–337.
Geelhoed, J.S., Meeussen, J.C.L., Hillier, S., Lumsdon, D.G., Tho-
mas, R.P., Farmer, J.G., and Paterson, E., 2002, Indentification
and geochemical modeling of processes controlling leaching of
Cr(VI) and other major elements from chromite ore processing
residue. Geochimica et Cosmochimica Acta, 66, 3927–3942.
Graham, M.C., Farmer, J.G., Anderson, P., Paterson, E., Hillier, S.,
Lumsdon, D.G., and Bewley, R., 2006, Calcium polysulfide
remediation of hexavalent chromium contamination from chromite
ore processing residue. Science of the Total Environment, 364,
32–44.
Higgins, T.E., Halloran, A.R., Dobbins, M.E., and Pittignano, A.J.,
1998, In-situ reduction of hexavalent chromium in alkaline soils
enriched with chromite ore processing residue. Air & Waste
Management Association, 48, 1100–1106.
Inorganic Crystal Structure Database, 2005, Fachinformationszen-
trum Karlsruhe, Germany.
International Centre for Diffraction Data (ICDD), 2002, Powder Dif-
fraction File, PDF-2 Database Release.
Jagupilla, S.C., Moon, D.H., Wazne, M., Christodoulatos, C., and
Kim, M.G., 2009, Effects of particle size and acid addition on the
remediation of chromite ore processing residue using ferrous sul-
fate. Journal of Hazardous Materials, 168, 121–128.
James, B.R., 1996, The challenge of remediating chromium-contam-
inated soil. Environmental Science & Technology, 30, 248A–
251A.
Jupe, A.C., Cockcroft, J.K., Barnes, P., Colston, S.L., Sankar, G., and
Hall, C., 2001, The site occupancy of Mg in the brownmillerite
structure and its effect on hydration properties: an X-ray/neutron
diffraction and EXAFS study. Journal of Applied Crystallogra-
phy, 34, 55–61.
Kremser, D.D., Sass, B.M., Clark, G.I., Bhargava, M., and French,
C., 2005, Environmental forensics investigation of buried
chromite ore processing residue. In: Proceedings of the 8
th
Inter-
national In Situ and On-Site Bioremediation Symposium, Balti-
more, Maryland, June 6–9.
Material’s Data Inc., 2005, Jade, Version 7.1, California, U.S.A.
Moon, D.H., Chrysochoou, M., Dermatas, D., Christodoulatos, C.,
Kaouris, M., and Morris, J., 2005a, Investigation of Ettringite
Formation in Chrome Ore Processing Residue. In: Proceedings
of the 8
th
International In Situ and On-Site Bioremediation Sym-
posium, Baltimore, Maryland, June 6–9.
Moon, D.H., Dermatas, D., Sanchez, A.M., Chrysochoou, M., Wazne,
M., and Grubb, D.G., 2008, Assessment of brownmillerite and
periclase hydration in chromite ore processing residue at elevated
temperature. In: Proceedings of GeoCongress 2008, New Orleans,
Louisiana, March 9–12.
Moon, D.H., Dutko, P., Dermatas, D., Chrysochoou, M., Christodoulatos,
C., Kaouris, M., and Morris, J., 2005b, Ettringite-Induced
Laboratory Swelling in Chromite Ore Processing Residue. In:
Proceedings of the 8
th
International In Situ and On-Site Biore-

296 Deok Hyun Moon and Mahmoud Wazne
mediation Symposium, Baltimore, Maryland, June 6–9.
Rietveld, H.M., 1969, Aprofile refinement method for nuclear and
magnetic structures. Journal of Applied Crystallography, 2, 65–71.
Taylor, H.F.W., 1997, The Chemistry of Cement, 2
nd
edition, Aca-
demic Press, London, 418 p.
Wazne, M., Jagupilla, S.C., Moon, D.H., Jagupilla, S.C., Christodou-
latos, C., and Lim, M.G., 2007, Assessment of calcium polysul-
fide for the remediation of hexavalent chromium in chromite ore
processing residue (COPR). Journal of Hazardous Materials,
143, 620–628.
Weng, C.H., Huang, C.P., Allen, H.E., Cheng, A.H.D., and Sanders,
P.F., 1994, Chromium leaching behaviour in soil derived from
chromite ore processing waste. Science of the Total Environ-
ment, 154, 71–86.
Wolery, T. and Jarek, R., 2003, EQ3/6 software package, version 8.0.
Manuscript received February 1, 2010
Manuscript accepted June 24, 2011