Attach. 8, Schwartz et al, Leaching Potential and Redox Transformations of As and Se

Applied Geochemistry 67 (2016) 177-185
Contents lists available at ScienceDirect
Applied Geochemistry
ELSEVIER journal homepage: www.elsevier.com/locate/apgeochem
Leaching potential and redox transformations of arsenic and selenium
in sediment microcosms with fly ash
Grace E. Schwartz a,1, Nelson Rivera a, Sung -Woo Lee a, James M. Harrington b,
James C. Hower `, Keith E. Levine b, Avner Vengosh d, Heileen Hsu -Kim a, *
a Duke University, Department of Civil & Environmental Engineering, 121 Hudson Hall, Durham, NC 27708, USA
b RTI International, Analytical Sciences, 3040 East Cornwallis Drive, Research Triangle Park, NC 27709, USA
University of Kentucky, Center for Applied Energy Research, 2540 Research Park Drive, Lexington, KY 40511, USA
a Duke University, Division of Earth and Ocean Sciences, Nicholas School of the Environment, Durham, NC 27708, USA
A R T I C L E I N F O
Article history:
Received 6 November 2015
Received in revised form
8 February 2016
Accepted 21 February 2016
Available online 23 February 2016
Keywords:
Coal combustion residuals
Water quality
Solid waste disposal
1. Introduction
A B S T R A C T
® CrossMark
The unintended release of coal ash to the environment is a concern due to the enrichment of contam-
inants such as arsenic (As) and selenium (Se) in this solid waste material. Current risk assessments of coal
ash disposal focus on pH as the primary driver of leaching from coal ash. However, redox speciation of As
and Se is a major factor for their mobilization potential and has received much less attention for risk
assessments, particularly in disposal scenarios where coal ash will likely be exposed to microbially-
driven redox gradients. The aim of this study was to demonstrate the differences of aerobic and
anaerobic conditions for the leaching of As and Se from coal ash. Batch sediment -ash slurry microcosms
were performed to mimic an ash spill scenario and were monitored for changes in As and Se speciation
and mobilization potential. The results showed that the dissolved As concentrations were up to 50 times
greater in the anaerobic microcosms relative to the aerobic microcosms during the two week incubation.
This trend was consistent with As redox speciation determined by X-ray absorption spectroscopy, which
indicated that 55% of the As in the solid phase at the end of the experiment was present as As(III) (a more
leachable form of arsenic relative to As(V)). In the aerobic microcosms, only 13% of the As was As(III) and
the rest was As(V). More than half of the Se was present as Se(IV) in the original fly ash and in the aerobic
microcosms, while in the anaerobic microcosms Se was gradually transformed to less soluble Se(0)
species. Likewise, dissolved Se concentrations were up to 25 times greater in the aerobic microcosms
relative to anaerobic conditions. While the overall observations of As and Se mobilization potential from
coal ash were consistent with expectations for aqueous and solid phase speciation of these elements, the
findings directly show the relevance of these processes for coal ash disposal. These results highlight the
need to select appropriate environmental parameters to include in risk assessments as well as provide
potential geochemical monitoring tools through the use of dissolved Se/As ratios to determine the redox
conditions of ash storage and spill sites.
Coal ash is the solid waste by-product of coal combustion and
includes bottom ash, fly ash, and sludge from flue gas desulfur-
ization units. Every year, over 60 Mt of coal ash are disposed in 300
landfills and 600 holding ponds across the United States (American
* Corresponding author.
E-mail address: [e-mail] (H. Hsu -Kim).
Current Affiliation: Smithsonian Environmental Research Center, 647 Contees
Wharf Road, Edgewater, Maryland, 21037 USA.
http://dx.doi.org/10.1016/j.apgeochem.2016.02.013
0883-2927/0 2016 Elsevier Ltd. All rights reserved.
© 2016 Elsevier Ltd. All rights reserved.
Coal Ash Association; US EPA, 2013d). Ash wastes are enriched in
many potentially toxic elements, and the presence of arsenic (As)
and selenium (Se) are a particular concern because of their rela-
tively high mobilization potential at neutral to alkaline pH values
that are typical of ash disposal impoundments (Izquierdo and
Querol, 2012; Meij, 1994). Moreover, these elements have the ten-
dency to bioaccumulate in the aquatic food web and impart eco -
toxicological effects (Izquierdo and Querol, 2012; Lemly, 2004;
Rowe, 2014; Sharma and Sohn, 2009; Thorneloe et al., 2010).
Coal ash impoundments are not always closely monitored,
particularly for effluent discharge to surface waters, seepage to

178
G.E. Schwartz et al. / Applied Geochemistry 67 (2016) 177-185
groundwater, and structural integrity. Consequently, impoundment
effluent discharge is a major source of As and Se contamination to
certain aquatic environments, with approximately 36,000 kg of As
and 102,000 kg of Se discharged annually to surface waters in the
United States (US EPA, 2013a). Ash impoundments have been cited
in 132 documented cases of groundwater and surface water
contamination (US EPA, 2007, 2013a). Moreover, impoundment
failures and direct release of ash have been reported by at least 41
different power plants in the last 15 years (US EPA, 2012a). These
impoundment failures include the 2008 ash spill at the Tennessee
Valley Authority (TVA) Kingston Fossil Plant. Arsenic and selenium
originating from the spilled ash were two contaminants of concern
at the site (Ruhl et at., 2009, 2010).
The mobility of As and Se in the environment is intimately
linked to redox speciation of these elements and the propensity of
individual species to associate with soil and sediment particles
through adsorption/desorption reactions and precipitation/disso-
lution reactions (Fernandez -Martinez and Charlet, 2009;
Masscheleyn et al., 1991). The As(V) oxyanion arsenate tends to
sorb more strongly to mineral phases such as iron oxides when
compared to the reduced As(III) arsenite form (Goldberg and
Johnston, 2001; Masscheleyn et al., 1991; Raven et al., 1998). If
sulfide is present in sufficient quantities, insoluble arsenic -sulfide
species (As2S3(s)) and soluble thioarsenicals may form (O'Day
et al., 2004; Wilkin et al., 2003). In contrast to As, the Se(VI) oxy-
anion selenate has little tendency to adsorb to solids or to precip-
itate out of solution compared to selenite Se(IV), which has greater
sorption affinity to metal oxyhydroxides, clays, and organic matter
(Fernandez -Martinez and Charlet, 2009). In anaerobic conditions,
selenium can persist as elemental selenium Se(0) or metal selenide
mineral phases that are sparingly soluble in water. Organo -Se
compounds such as selenocysteine and selenomethionine are a
reduced forms of Se that are biologically active (Lemly, 1993).
Arsenic associated with coal ash exists mainly as As(V) species
while Se is typically found as Se(IV) and elemental Se(0) species
(Chappell et al., 2014; Deonarine et al., 2015; Huggins et al., 2007;
Liu et al., 2013). The mobilization of As and Se from coal ash is
typically assessed using deionized water under aerobic conditions
and perhaps under a wide range of pH values (Bednar et al., 2010;
Izquierdo and Querol, 2012; Liu et al., 2013; Thorneloe et al., 2010).
Much less attention has been given to redox transformations that
can occur during ash disposal, even though these processes are
critical for As and Se mobilization, as stated above. Two studies
have attempted to address redox conditions by taking ash-
deionized water mixtures and purging a subset with nitrogen gas
(Bednar et al., 2010; Liu et al., 2013). The results of the work showed
no or minimal differences between the oxic and N2 -purged mix-
tures with respect to As and Se redox speciation and leaching po-
tential from the ash. These results were inconsistent with our
previous field studies at the TVA Kingston ash spill site and at
several North Carolina ash holding ponds, where the mobilization
of As and Se from coal ash appeared to change as a function of local
redox conditions (Ruhl et al., 2012, 2010, 2009). Comprehensive
measurements of As and Se speciation were not available from
these field sites to verify the mechanisms of leaching from the ash.
This study aimed to delineate the effects of redox gradients for
As and Se mobilization from coal ash using laboratory sediment
microcosms that more closely mimic the complexity of biogeo-
chemical redox processes in the environment. Another objective
was to improve our understanding of processes that were previ-
ously observed at coal ash spill sites and perhaps identify
geochemical tools for monitoring coal ash contaminants in redox
gradients. Batch sediment slurry microcosms were constructed
with aerobic and anaerobic conditions and were amended with fly
ash to simulate a coal ash spill into a benthic sediment -water
system. The microcosms were monitored for total dissolved con-
centrations of As and Se, speciation of these elements in the
aqueous and solid phases of the microcosms, and other water
chemistry variables relevant for As and Se leaching.
2. Materials and methods
2.1. Materials
All chemicals for reagents were purchased from Sigma Aldrich
(St. Louis, MO), unless otherwise stated. Trace metal grade acids
(Fisher Scientific, Pittsburgh, PA) were used for acid digestions and
pH adjustments of samples. All reagents and calibration standards
were prepared with >18 MQ -cm Milli -Q grade filtered water (EMD
Millipore).
The microcosms comprised of mixtures of sediment and water
from the Emory River (Tennessee, USA). Surface water and bottom
sediment samples for the microcosms were collected in April 2014
from mile marker 10 of the Emory River near Kingston
(35.9475941°, —84.53178889°), which is located several miles up-
stream of the TVA Kingston ash spill (Bartov et al., 2012; Deonarine
et al., 2013; Ruhl et al., 2009, 2010). The sediment was a mixture of
brown, organic fines and sand. Water samples were taken at 0.15 m
depth and were stored in acid -cleaned plastic jugs. Bulk sediment
was collected from the top layer of sediment (approximately 15 cm)
using a Ponar dredge (Wildco) and placed in screw top buckets. The
sediment and water samples were stored on ice for shipment to
Duke University and stored at 4 °C in the laboratory. These sedi-
ment and water samples were used within one month after
collection for the microcosm experiments. Prior to the construction
of the microcosms, the water was analyzed for trace element
concentration, pH, and conductivity.
The coal ash used for the microcosm experiments was collected
at the TVA John Sevier fossil plant in April 2011. The sample was a
composite of fly ash collected from electrostatic precipitator hop-
pers at each of the plant's four units. The composite fly ash sample
was characterized for major mineral oxide content (by X-ray fluo-
rescence) and for total As and Se concentrations (methods
described in Section 2.3).
2.2. Microcosm preparation and sampling
Sediment slurry microcosm experiments were conducted on
two separate occasions: the first under aerobic conditions, followed
by the second under anaerobic conditions. Each treatment type
(with and without ash; aerobic and anaerobic) was performed in
duplicate microcosms.
The microcosms were designed to mimic a stagnant, ash -
impacted environment. The sediment to surface water ratio in the
microcosms was chosen to provide an environment where suffi-
cient overlying water would be available for sampling and analyses
but microbial activity would not be limited. Each microcosm was
prepared in a 1-L acid -washed, glass jar and consisted of 240 g of
sediment (wet weight basis) and 600 mL of surface water. The
sediment was thoroughly homogenized by stirring before micro-
cosm construction. The river water was amended with a carbon
substrate for microbial activity, 0.5-mM pyruvate and 0.5-mM ac-
etate, immediately prior to microcosm construction.
The aerobic microcosms were continuously stirred and purged
with hydrated air using Teflon tubing and aquarium air stones
during the experiments. After an incubation period of three days,
56 g of coal ash was added to the microcosms designated for ash
amendment. This amount of coal ash corresponded to 40% (w/w) of
dry sediment, an amount that was observed at the TVA Kingston
ash spill site after dredging was completed (Deonarine et al., 2013).

G.E. Schwartz et al. / Applied Geochemistry 67 (2016) 177-185
A single replicate sediment -water microcosm containing
6 mg L-1 resazurin was also constructed to serve as an indicator of
redox conditions. This indicator microcosm was not used for the
ash experiments but was used only to infer aerobic conditions for
the other microcosms.
The anaerobic microcosm experiment was performed after the
completion of the aerobic experiment. These microcosms con-
tained the same amount of water and sediment (600 mL and 240 g,
respectively); however, they were constructed in 1-L glass pyrex
bottles with gas-tight caps and assembled inside an anaerobic
chamber (Coy Labs, Grass Lake, MI) containing an ambient atmo-
sphere of 90% Nzlgl, 5% COZIgy and 5% Hz(g). Surface water amended
with the carbon substrate (0.5-mM pyruvate and 0.5-mM acetate)
was purged with high purity NZ for at least 15 min immediately
prior to addition to the microcosms. A single replicate microcosm
with the resazurin redox indicator (6 mg L-1) and no ash was also
prepared for the anaerobic experiment. After assembly, the sealed
microcosms were stored in the laboratory under static conditions at
room temperature (22 °C). The microcosms were mixed end -over -
end once per day in addition to immediately prior to each sampling
time point. Anaerobic conditions (EH < —50 mV) were achieved in
approximately three days, as indicated by the resazurin indicator
microcosm turning from pink to a clear color. At this time, 56 g of
coal ash was added to the microcosms designated for the ash
amendments (performed in the anaerobic chamber).
At time points before (-72 h, -2 h) and after the ash amendment
(4, 24, 72,168, and 336 h), samples of the microcosm slurry were
collected (12-15 mL in most instances). A portion of the sample
was immediately filtered through a 0.2-µm nylon syringe filter
(VWR). This filtered fraction is herein referred as the "dissolved"
fraction. The solid phase of the slurry sample was collected by
centrifugation (3000 RPM for 15 min). For the aerobic microcosms,
the air bubblers and stir plates were turned off during the sampling.
For the anaerobic microcosms, the collection, filtration, and pres-
ervation of the sample were performed in the anaerobic chamber.
Samples for solid phase separation were capped in the anaerobic
chamber, centrifuged outside the chamber, and then returned
immediately to the anaerobic chamber for the remainder of the
sample preparation.
2.3. Sample preservation and chemical analyses
Filtered aqueous samples were split and analyzed for dissolved
As and Se concentrations, the speciation of dissolved As, pH, and
other relevant water quality variables. Samples for dissolved trace
element analysis (e.g., As, Se, and Fe) were immediately diluted in a
2% (v/v) HNO3/0.5% (v/v) HCl solution and analyzed by inductively
coupled plasma -mass spectrometry (ICP -MS, Agilent 7700). Sulfate
concentration was quantified by ion chromatography (Dionex). For
acid volatile sulfide (AVS), filtered water samples were preserved
with the addition of 10 mM ZnSO4 and 5 mM KOH (corresponding
to final concentrations in the sample) and stored at 4 °C until
analysis. AVS measurements were made using the method
described by Allen et al. (1993) and summarized in the Supporting
Information (SI).
Aliquots of the filtered sample were also analyzed for aqueous
phase As speciation (i.e., arsenate and arsenite) via ultra perfor-
mance liquid chromatography -ICP -MS (Waters ACQUITY UPLC)
coupled with a Thermo ELEMENT 2 sector field ICP -MS (Kim et al.,
2013; Milstein et al., 2003). These samples were preserved with
0.125-M EDTA, in accordance to Bednar et al. (2002). UPLC-ICP-MS
instrument parameters are shown in Table S2. Additional details of
aqueous sample preservation and analysis methods can be found in
the SI section.
Total Se and As concentrations in the original sediment and fly
179
ash samples were quantified by heated HNO3 acid digestion (85 °C
for 6 h) and analysis by ICP -MS. The same acid extraction procedure
was performed for a soil standard reference material (San Joaquin
Soil NIST SRM 2709) and a coal fly ash reference material (NIST SRM
1633c). The recovery of certified As concentration values were 84%
and 90% for the soil and fly ash references, respectively. Se re-
coveries were 84% and 105%, respectively. For the fly ash sample,
the major mineral oxide content was also characterized via X-ray
Fluorescence following the ASTM standard method for ash analysis.
Se and As speciation in the solid phase from the ash -amended
microcosms was determined using K -edge X-ray absorption near
edge structure (XANES) spectroscopy. The pellets obtained after
centrifugation were subsequently packed into sample holders as
wet pastes, covered with Kapton tape, and stored at 4 'C until
analysis. XANES analyses were also performed for the original
sediment and fly ash endmembers for the microcosms but were not
performed on the ash -free microcosms due to the relatively low
concentration of Se (<0.5 µg g-1) and arsenic (<2.5 µg g-1) and
limitations of synchrotron beamtime required for these
measurements.
XANES spectra were collected in fluorescence mode at Beamline
11-2 at the Stanford Synchrotron Radiation Lightsource (SSRL) in
Menlo Park, CA. Samples were held in a liquid NZ cryostat during
analysis, and XANES spectra were collected with the use of a Si(220)
(phi = 90°) monochromator and a 100 -element solid state Ge de-
tector array. Successive scans were collected to ensure that no
changes in the sample occurred during data collection. Speciation
of As and Se in the samples was quantified by linear combination
fitting (LCF) of reference spectra to the sample spectra. For sele-
nium, the references included sodium selenate, sodium selenite,
selenite sorbed to aluminosilicate glass (Rivera et al., 2015), SO (S),
FeSe(s), and seleno-L-cystine. Arsenic references included As(V)-
and As(III)-oxides, arsenate adsorbed to aluminosilicate glass
(Rivera et al., 2015), arsenite adsorbed to ferrihydrite (Root et al.,
2007), orpiment (As2S3) and realgar (ASS). Additional details on
XANES sample preparation and analysis can be found in the SI
section.
3. Results and discussion
3.1. Characteristics of the ash, sediment, and water used for the
microcosms
Total As and Se concentrations in the ash were 44 µg g-1 dry
weight (dw) and 19 µg g-1 dw, respectively, while in the original
river sediment, total As and Se were 2.38 µg g-1 wet weight (ww)
and 0.27 gg g-1 ww, respectively. Characteristics of the surface
water included pH 7.5, 125.6 µS cm -1 conductivity, less than
4 µg L-1 As, and <0.4 µg C1 Se. With these concentrations quan-
tified in the original materials, each sediment microcosm without
ash was estimated to contain 570 µg of As and 65 µg of Se. Micro-
cosms with sediment and ash each contained approximately
3040 µg total As and 1130 µg total Se, with 81% and 94% of the As
and Se originating from the fly ash. The fly ash sample comprised
primarily of silica-, aluminum-, and iron -oxides (56%, 28%, and
6.7%, respectively) (Table S1), characteristics that are typical for a
Class F fly ash.
3.2. Leaching potential of selenium in anaerobic and aerated
microcosms
Upon addition of the coal ash to both anaerobic and aerobic
microcosms, dissolved Se immediately increased to concentrations
that were 150 -times greater than pre -amendment measurements
(Fig. 1). The extent of dissolved Se release varied according to redox

180
(a) 120
J
rn
m
a�
0
0
m
100
80
60
40
20
G.E. Schwartz et al. / Applied Geochemistry 67 (2016) 177-185
my
Ash
-100 0 100 200 300 400
Hours after Coal Ash Amendment
(b) 120 Anaerobic
�1 100
6)
80 -D `Sediment -Only
0 Sediment + Ash
> 60
0
ij 40
20
0
-100 0 100 200 300 400
Hours after Coal Ash Amendment
Fig. 1. Total dissolved selenium concentrations (<0.2 gm filtered fraction) in sediment -
ash microcosms: (a) Aerobic treatments; (b) Anaerobic treatments. Each data point
represents the average of duplicate microcosms. Error bars represent the range of
duplicates.
state, but in both cases, dissolved Se concentrations were greater in
microcosms with ash amendments than in the respective ash -free
control microcosms (<1.5 µg L-1 throughout the experiment).
Dissolved Se was generally greater in the aerobic microcosms
than in the anaerobic microcosms (with the exception of the first
time point at 4 h) (Fig. 1). In the aerobic ash -amended microcosms,
total Se concentration reached a maximum of 53 µg L-1 at 24 h after
ash addition and decreased to 29 gg L-1 over the course of the two-
week experiment. While a decrease in soluble Se could indicate
reductive transformation of Se to less soluble, lower oxidation
states such as elemental selenium, other water chemistry variables
such as constant levels of dissolved sulfate (Fig. SI) and low levels
of dissolved iron (<0.02 mg L-1) (Fig. S2) throughout the experi-
ment indicated that aerobic conditions were maintained. The
decrease in dissolved Se could instead be a result of re -adsorption
of Se(IV) species onto coal ash and sediment particles, which has
been shown to occur in aerobic systems (Fan et al., 2002; Simmons
and Wallschlager, 2005). Measurements of dissolved Se speciation
were attempted, but the concentrations were below the limit of
quantification for our UPLC-ICP-MS system (<50 µg L-1).
In anaerobic ash -amended microcosms (Fig. 1b), there was an
immediate spike in total dissolved Se after the coal ash amend-
ment, but then Se concentration decreased from 94 µg L-1 at 4 h to
1.8 gg L-1 at 336 h. This decline in dissolved Se concentration was
more drastic than that observed in the aerobic microcosms
amended with ash.
The pH of the water could influence leaching of Se-oxyanions,
which tend to desorb in greater amounts from coal ash as pH in-
creases (Liu et al., 2013). In the ash -amended microcosms, the pH
was 7.3-7.5 in aerobic conditions and similar to pH values in the
anaerobic experiment (pH 7.1-7.4) (Fig. S3). These results indicated
that the differences between the aerobic and anaerobic microcosms
for dissolved Se could not be explained by pH.
The measured amount of dissolved Se was always less than 5% of
the total Se in the microcosm, and the bulk of the selenium
remained in the solid phase. Therefore, we examined the speciation
of solid phase Se as a means to determine the longer term leaching
potential of Se. In the original fly ash, approximately 65% (±0.4%) of
the Se was Se(IV) species, as indicated by LCF models of the Se K-
edge XANES spectra (Fig. 2, Table S3). A smaller proportion, 26%
(±0.4%) and 9% (±0.5%), was Se(0) and Se(VI), respectively. In all ash
and microcosm samples, the best fits were obtained with the use of
selenite sorbed to aluminosilicate rather than sodium selenite for
the Se(IV) reference. This suggests that the sorbed selenite material
was a better approximation of Se(IV) species in the coal ash -
sediment matrix than the sodium selenite standard.
In aerobic ash -amended microcosms, Se(IV) was the dominant
form of Se in the solid phase at all time points (Fig. 2, Table S3). This
result is consistent with the Se speciation of the original ash sam-
ple. The proportion of Se(0) also appeared to grow over the course
of the experiment (from 19 ± 0.6% at 24 h to 38 ± 0.5% at 336 h).
While abiotic reduction of Se is not expected in the aerated con-
ditions of the microcosms, selenite-reducing microorganisms are
capable of producing Se(0) in oxic conditions (Antonioli et al., 2007;
Hunter and Kuykendall, 2007; Hunter and Manter, 2009; Zheng
et al., 2014).
In the anaerobic microcosms amended with ash, Se speciation of
the solid phase was drastically different from the original fly ash
sample. LCF models of the XANES spectra showed that the majority
of selenium in the solids was Se(0) and FeSe and that the proportion
of Se(0)+FeSe increased from 68% at 4 h to 79% at 336 h (Fig. 2,
Table S3). The formation of FeSe was further supported by the
dissolved Fe data (Fig. S2). In the anaerobic experiment, dissolved
Fe concentrations immediately decreased to values below
1.5 mg L-1 after the ash amendment, even though dissolved Fe in
the ash -free anaerobic control was greater than 10 mg L-1 and
increasing with time. Sulfate reduction was also occurring in the
anaerobic microcosms (Fig. S1); thus, dissolved Fe was likely
precipitating out of solution as FeS and FeSe particles.
Collectively, the aqueous and solid phase Se speciation data
indicated that a portion of the Se in fly ash readily leached from the
ash under aerobic conditions. However, in anaerobic settings, the
results indicated that Se originating from the ash was transformed
to species of lower oxidation states, subsequently diminishing the
leaching potential of Se.
3.3. Arsenic dissolution and speciation in aerobic and anaerobic
microcosms
The leaching of As in the microcosms was also dictated by redox
potential, with much greater dissolved As concentrations observed
in anaerobic conditions than in aerobic conditions (Fig. 3). In the
aerobic microcosms (Fig. 3a), the addition of ash resulted in an
increase of total dissolved As concentration to a maximum value of
12 µg L-1 at 336 h (Fig. 3a). In the anaerobic experiment (Fig. 3b),
the total dissolved As concentration after the addition of ash was
greater: 157 µg L-1 was detected at 4 h and a maximum dissolved
As concentration of 498 µg L-1 was observed at 72 h (Fig. 3a). This
amount of dissolved As represented 9.8% of the total mass of As in
the microcosm container (570 gg As from the original sediment;
2460 µg As from the ash). After 72 h, total dissolved As concen-
trations declined slightly over the remainder of the anaerobic

G.E. Schwartz et al. / Applied Geochemistry 67 (2016) 177-185
(a) Sample - - - LCF (b)
Se(IV)
Se(0)
FeSe
12650 12660 12670 12680
Energy (eV)
0% 20% 40% 60% 80% 100%
Percent Composition
® FeSe
D Elemental Se
INSe(IV)
MSe(VI)
181
Fig. 2. Solid phase speciation of selenium: (a) Normalized Se K -edge XANES spectra for solids from the ash -amended microcosms and models of the data using linear combination
fitting (LCF) of reference spectra; (b) The relative proportions of iron selenide (FeSe), elemental $e(0), Se(IV) (as selenite sorbed to aluminosilicate), and Se(VI) (sodium selenate). The
total Se in the solid phase of the microcosm was estimated to be 3.8 µg g-1, of which 94% originated from the coal ash.
experiment.
The speciation of dissolved and solid phase As was largely
consistent with expectations for the tested redox condition. In the
aerobic ash -amended microcosms, more than 92% of the dissolved
As was in the oxidized form, As(V) (Fig. 4a). Similarly, As speciation
in the original fly ash and the sediment -ash mixture of the aerobic
microcosm was predominantly As(V) (85-89%, Fig. 5 and Table S3).
The best model fits for the As K -edge XANES spectra were obtained
with arsenate -sorbed to aluminosilicate glass as the As(V) refer-
ence material and with arsenite sorbed to ferrihydrite as the As(III)
reference material. We observed poorer fits with the use of As -
oxide compounds as the As(V) and As(III) references. This result
suggested that the sorbed standards were better mimics of As
species in the coal ash and ash -sediment matrices.
In the anaerobic experiment, the dissolved As after the addition
of ash was present mostly as arsenite (>86%) (Fig. 4b). As(V) was
consistently less than 15% of the dissolved As in the anaerobic
microcosms with ash. We note, however, that for the amount of
dissolved sulfide in the anaerobic microcosms with fly ash
(2-8 µM, Fig. S4), thioarsenite and thioarsenate species were
possible. Previous studies have shown that thio -arsenic species are
not preserved by the EDTA reagent and can be converted to arsenite
and arsenate prior to the analysis (Suess et al., 2011). While thio -
arsenicals were a possibility in the anaerobic microcosms, they
nevertheless were not expected to be dominant based on ther-
modynamic considerations (Wilkin et al., 2003).
In the solid phase of the ash -amended anaerobic microcosms,
As(III) was also the dominant form of As (55-73%) while As(V) was
less than 41% at all time points (Fig. 5, Table S3). This result indi-
cated relatively rapid transformation of As(V) from the fly ash to
As(III) species. The amount of As -sulfide solids in the anaerobic
experiment also increased from 2% at 24 h to 19% at 336 h. The
formation of relatively insoluble As -sulfide minerals such as orpi-
ment can result in a decrease of dissolved As (Burton et al., 2014;
O'Day et al., 2004). Thus, the production of As -sulfides species in
the anaerobic microcosm could explain the decrease of dissolved As
after the 72-h time point (Fig. 3b). Collectively the dissolved and
solid phase As species distribution in the anaerobic experiments
demonstrated reductive transformation of As from the ash material
and was consistent with expectations for As speciation in a strongly
reducing environment where sulfate reduction was occurring
(Smedley and Kinniburgh, 2002).
Unexpectedly, in the anaerobic ash -free microcosms, As(V) was
found to be the dominant form (70% and greater) of dissolved As
(Fig. 4c). With relatively low amounts of dissolved sulfide in the
ash -free microcosm (less than 0.1 µM at the end, Fig. S4), the dis-
solved As(V) was likely to primarily consist of arsenate rather than
thioarsenate. The total dissolved As and dissolved Fe concentra-
tions were also increasing with time in the anaerobic ash -free
microcosms. Thus, leaching of As was likely occurring through
reductive dissolution of iron oxides and release of As(V) sorbed to
these minerals. We note, however, that the reduction of Fe(III)-
oxides and As(V) occur in the same Eh range (0-100 mV) for
neutral pH conditions (Masscheleyn et al., 1991), and the reduction
potential of the anaerobic microcosms was likely to be less
than —50 mV, as indicated by the resazurin. Thus, it is unclear why
As(V) remained dominant in the anaerobic ash -free microcosms.
One potential explanation is that Fe(III) outcompeted As(V) as an
electron acceptor for microbial respiration. The kinetics for As(V)
reduction to As(III) are also known to be relatively slow, which may
Aerobic: 336h
Aerobic: 168h
11- Aerobic:24h
7
li
lilt
t
Anaerobic: 336h
"
Anaerobic: 168h
LU
Z
�'
I Anaerobic:24h
Q
X
.I
I
Sediment
EE
�
Ash
o
Z
n
i
"
HA
Se(VI)
Se(IV)
Se(0)
FeSe
12650 12660 12670 12680
Energy (eV)
0% 20% 40% 60% 80% 100%
Percent Composition
® FeSe
D Elemental Se
INSe(IV)
MSe(VI)
181
Fig. 2. Solid phase speciation of selenium: (a) Normalized Se K -edge XANES spectra for solids from the ash -amended microcosms and models of the data using linear combination
fitting (LCF) of reference spectra; (b) The relative proportions of iron selenide (FeSe), elemental $e(0), Se(IV) (as selenite sorbed to aluminosilicate), and Se(VI) (sodium selenate). The
total Se in the solid phase of the microcosm was estimated to be 3.8 µg g-1, of which 94% originated from the coal ash.
experiment.
The speciation of dissolved and solid phase As was largely
consistent with expectations for the tested redox condition. In the
aerobic ash -amended microcosms, more than 92% of the dissolved
As was in the oxidized form, As(V) (Fig. 4a). Similarly, As speciation
in the original fly ash and the sediment -ash mixture of the aerobic
microcosm was predominantly As(V) (85-89%, Fig. 5 and Table S3).
The best model fits for the As K -edge XANES spectra were obtained
with arsenate -sorbed to aluminosilicate glass as the As(V) refer-
ence material and with arsenite sorbed to ferrihydrite as the As(III)
reference material. We observed poorer fits with the use of As -
oxide compounds as the As(V) and As(III) references. This result
suggested that the sorbed standards were better mimics of As
species in the coal ash and ash -sediment matrices.
In the anaerobic experiment, the dissolved As after the addition
of ash was present mostly as arsenite (>86%) (Fig. 4b). As(V) was
consistently less than 15% of the dissolved As in the anaerobic
microcosms with ash. We note, however, that for the amount of
dissolved sulfide in the anaerobic microcosms with fly ash
(2-8 µM, Fig. S4), thioarsenite and thioarsenate species were
possible. Previous studies have shown that thio -arsenic species are
not preserved by the EDTA reagent and can be converted to arsenite
and arsenate prior to the analysis (Suess et al., 2011). While thio -
arsenicals were a possibility in the anaerobic microcosms, they
nevertheless were not expected to be dominant based on ther-
modynamic considerations (Wilkin et al., 2003).
In the solid phase of the ash -amended anaerobic microcosms,
As(III) was also the dominant form of As (55-73%) while As(V) was
less than 41% at all time points (Fig. 5, Table S3). This result indi-
cated relatively rapid transformation of As(V) from the fly ash to
As(III) species. The amount of As -sulfide solids in the anaerobic
experiment also increased from 2% at 24 h to 19% at 336 h. The
formation of relatively insoluble As -sulfide minerals such as orpi-
ment can result in a decrease of dissolved As (Burton et al., 2014;
O'Day et al., 2004). Thus, the production of As -sulfides species in
the anaerobic microcosm could explain the decrease of dissolved As
after the 72-h time point (Fig. 3b). Collectively the dissolved and
solid phase As species distribution in the anaerobic experiments
demonstrated reductive transformation of As from the ash material
and was consistent with expectations for As speciation in a strongly
reducing environment where sulfate reduction was occurring
(Smedley and Kinniburgh, 2002).
Unexpectedly, in the anaerobic ash -free microcosms, As(V) was
found to be the dominant form (70% and greater) of dissolved As
(Fig. 4c). With relatively low amounts of dissolved sulfide in the
ash -free microcosm (less than 0.1 µM at the end, Fig. S4), the dis-
solved As(V) was likely to primarily consist of arsenate rather than
thioarsenate. The total dissolved As and dissolved Fe concentra-
tions were also increasing with time in the anaerobic ash -free
microcosms. Thus, leaching of As was likely occurring through
reductive dissolution of iron oxides and release of As(V) sorbed to
these minerals. We note, however, that the reduction of Fe(III)-
oxides and As(V) occur in the same Eh range (0-100 mV) for
neutral pH conditions (Masscheleyn et al., 1991), and the reduction
potential of the anaerobic microcosms was likely to be less
than —50 mV, as indicated by the resazurin. Thus, it is unclear why
As(V) remained dominant in the anaerobic ash -free microcosms.
One potential explanation is that Fe(III) outcompeted As(V) as an
electron acceptor for microbial respiration. The kinetics for As(V)
reduction to As(III) are also known to be relatively slow, which may

182
(a) 16 Aerobic
J
6) 12 T
Q
G.E. Schwartz et al. / Applied Geochemistry 67 (2016) 177-185
8 I
> -a -Sediment Only
0
f Sediment + Ash
0 4
- - - -
0 a -
-100 0 100 200 300 400
Hours after Coal Ash Amendment
(b) 600 Anaerobic m❑ -Sediment Only
—W- Sediment+Ash
J
400
Q
Q)
N
O
W 200
0
-------❑
0 IT I
-100 0 100 200 300 400
Hours after Coal Ash Amendment
Fig. 3. Total dissolved arsenic concentrations (<0.2-µm filtered fraction) in sediment -
ash microcosms: (a) Aerobic treatments; (b) Anaerobic treatments. Data points and
error bars represent the average and range of duplicate microcosms.
have contributed to the presence of both species in the ash -free
microcosm (Masscheleyn et al., 1991; Smedley and Kinniburgh,
2002). Likewise, the ash -free microcosms were poised at moder-
ately reducing conditions while the ash -amended microcosms
were poised at lower redox potential (as indicated by reduction of
sulfate and production of sulfide in the presence of ash, Figs. SI and
S4). Lower Eh values in the ash amendments could lead to greater
conversion of As(V) to As(III) compared to the ash -free control.
In summary, the microcosm experiments demonstrated that the
leaching potential of As from coal ash was greater in anaerobic
conditions than aerobic conditions, due to redox transformations of
As. However, the large amount of sulfur from the ash could
contribute to secondary precipitation reactions of As -sulfides if the
ash was released or stored in sufficiently reducing conditions.
3.4. Implications for ash spill settings
The results of this study showed that in a system buffered at
neutral pH, redox potential had a major influence on the release of
Se and As from coal ash, with increased As release under anaerobic
conditions and increased Se release under aerobic conditions.
Furthermore, this study provided clues to the impact of coal ash on
the geochemistry of the benthic environment and subsequent im-
plications for As and Se speciation and solubility. For example, the
microcosm experiments showed that coal ash dramatically
increased dissolved sulfate concentrations. In anaerobic environ-
ments with active microbial populations, reduction of sulfate can
result in formation of sulfide and the sequestration of As(III) in
sulfide mineral phases.
The results also shed light on possible field-based tools to
(a) Aerobic: Ash -Amended Microcosms
12
10
Q
6
0
n 4
2
0
(b) 600
500
400
N
Q
300
0
w 200
100
0
(o) 80
-1 60
-o 40
0
Ch
0 20
0
.... n, 'I 'I A01
4 24 72 168 336
Hours after Coal Ash Amendment
Anaerobic: Ash -Amended Microcosms
124%
102%
11 104%Ih
105%
105%
4 24 72 168 336
Hours after Coal Ash Amendment
Anaarnhir• Cariimant r)nly Mirror -c
0 As(V)
El As(III)
■ As(V)
El As(III)
L1 As(V)
ElAs(III)
4 24 72 168 336
Hours after Coal Ash Amendment
Fig. 4. Dissolved arsenic as arsenate As(V) and arsenite As(Ill) in: (a) Aerobic ash -
amended treatment; (b) Anaerobic ash -amended treatment; (c) Anaerobic sediment
only (no ash) treatment. Bars represent the average of duplicate microcosms. The
percentages above each bar is the recovery of total dissolved As (quantified inde-
pendently by ICP -MS). Dissolved As concentrations in the ash -free aerobic microcosms
were below detection limits for speciation analysis (<12 µg L-1).
evaluate the processes controlling As and Se mobilization from coal
ash. The Se/As ratios in the aerobic and anaerobic experiments
showed opposite trends, depending on the redox state of the
experiment (Fig. 6). If the original Se/As ratio is known for a spilled
coal ash, one could delineate the conditions that control Se and As
mobilization based on the changes in their ratios relative to the
ratios in the original coal ash. This tool might provide an indirect
measurement of the redox state of the system and predictions for
future fluctuation in Se and As contents based on redox conditions.
The design of this study best mimics stagnant ash -impacted

G.E. Schwartz et al. / Applied Geochemistry 67 (2016) 177-185
a
()b
Sample - - - LCF
1XW
11840 11860 11880 11900 11920
Energy (eV)
0% 20% 40% 60% 80% 100%
Percent Composition
® As -Sulfide
ElAs(III)
R As(V)
183
Fig. 5. Solid phase speciation of arsenic: (a) Normalized As K -edge XANES spectra for solids from the ash -amended microcosms and models of the data using linear combination
fitting (LCF); (b) The relative proportions of As(V) (arsenate sorbed to aluminosilicate glass), As(Ill) (arsenite sorbed to ferrihydrite), and As -sulfide (as orpiment). Total As in the solid
phase of the microcosms was -10 µg g-1 and approximately 80% of the As originated from the coal ash.
10 aerobic
•
0 1
original ash
Q 0.1
A
A
0.01
anaerobic A
I�IZI)yl
0 100 200 300 400
Hours after ash addition
Fig. 6. Dissolved Se/As concentration ratios in the aerobic and anaerobic microcosms
amended with ash.
environments, and our results support observations from our pre-
vious field studies of coal ash impacted environments with limited
water exchange (Ruhl et al., 2009, 2010). For example at the TVA -
Kingston ash spill site, pore water extracted from buried
sediment -ash mixtures was found to have much higher total dis-
solved As concentration (mean = 324 µg L-1) than standing surface
water at the site (mean = 53.3 µg L-1) (Ruhl et al., 2010). Addi-
tionally, a study of North Carolina surface waters receiving coal ash
effluent revealed that both As and Se accumulated in lake bottom
sediments and were released into the water column during sea-
sonal thermal stratification and fluctuations of redox potential in
the water column (Ruhl et al., 2012). Our microcosm study repre-
sents only a 2 week snapshot of fly ash weathering, so some caution
is warranted in extrapolating the results to the long term fate of
contaminants at ash spill sites. Nevertheless, data from these ex-
periments strongly suggest that redox transformations of coal ash
contaminants should be considered when assessing remediation
options for ash -impacted environments, especially when balancing
the risks of natural attenuation and alternative measures such as
dredging.
Due to redox -induced adsorption/desorption reactions and
precipitation/dissolution reactions, As concentrations can be very
high in sediment pore water even though contaminant concen-
trations in overlying oxic surface waters may be well -below EPA
guidelines. These high concentrations could present a risk for bio -
magnification in the benthic aquatic food web. Furthermore, the
release of As from sediments into overlying surface waters during
thermal stratification events has implications for communities that
use the surface water for recreational use and as a drinking water
reservoir. The consequences of seasonal As release would be miti-
gated to some degree by dilution, but, nevertheless, the long-term
cycling of As in the environment should be taken into account
when communities consider plans for remediating coal ash
impacted sites (Ruhl et al., 2012).
In the case of Se, Se oxyanion species in the aerobic water col-
umn can be taken up by aquatic biota and converted to
Aerobic: 336h
"
Aerobic:168h
Aerobic: 24h
VVIA,,
Anaerobic: 336h
W
Anaerobic: 168h
Z
Anaerobic: 24h
�
'll
N
E
`0
Sediment
Z
"
"'
Ash
As(V)
.1
As(111)
As -Sulfide
1XW
11840 11860 11880 11900 11920
Energy (eV)
0% 20% 40% 60% 80% 100%
Percent Composition
® As -Sulfide
ElAs(III)
R As(V)
183
Fig. 5. Solid phase speciation of arsenic: (a) Normalized As K -edge XANES spectra for solids from the ash -amended microcosms and models of the data using linear combination
fitting (LCF); (b) The relative proportions of As(V) (arsenate sorbed to aluminosilicate glass), As(Ill) (arsenite sorbed to ferrihydrite), and As -sulfide (as orpiment). Total As in the solid
phase of the microcosms was -10 µg g-1 and approximately 80% of the As originated from the coal ash.
10 aerobic
•
0 1
original ash
Q 0.1
A
A
0.01
anaerobic A
I�IZI)yl
0 100 200 300 400
Hours after ash addition
Fig. 6. Dissolved Se/As concentration ratios in the aerobic and anaerobic microcosms
amended with ash.
environments, and our results support observations from our pre-
vious field studies of coal ash impacted environments with limited
water exchange (Ruhl et al., 2009, 2010). For example at the TVA -
Kingston ash spill site, pore water extracted from buried
sediment -ash mixtures was found to have much higher total dis-
solved As concentration (mean = 324 µg L-1) than standing surface
water at the site (mean = 53.3 µg L-1) (Ruhl et al., 2010). Addi-
tionally, a study of North Carolina surface waters receiving coal ash
effluent revealed that both As and Se accumulated in lake bottom
sediments and were released into the water column during sea-
sonal thermal stratification and fluctuations of redox potential in
the water column (Ruhl et al., 2012). Our microcosm study repre-
sents only a 2 week snapshot of fly ash weathering, so some caution
is warranted in extrapolating the results to the long term fate of
contaminants at ash spill sites. Nevertheless, data from these ex-
periments strongly suggest that redox transformations of coal ash
contaminants should be considered when assessing remediation
options for ash -impacted environments, especially when balancing
the risks of natural attenuation and alternative measures such as
dredging.
Due to redox -induced adsorption/desorption reactions and
precipitation/dissolution reactions, As concentrations can be very
high in sediment pore water even though contaminant concen-
trations in overlying oxic surface waters may be well -below EPA
guidelines. These high concentrations could present a risk for bio -
magnification in the benthic aquatic food web. Furthermore, the
release of As from sediments into overlying surface waters during
thermal stratification events has implications for communities that
use the surface water for recreational use and as a drinking water
reservoir. The consequences of seasonal As release would be miti-
gated to some degree by dilution, but, nevertheless, the long-term
cycling of As in the environment should be taken into account
when communities consider plans for remediating coal ash
impacted sites (Ruhl et al., 2012).
In the case of Se, Se oxyanion species in the aerobic water col-
umn can be taken up by aquatic biota and converted to

184
G.E. Schwartz et al. / Applied Geochemistry 67 (2016) 177-185
organoselenium species, which are highly bioaccumulative (Fan
et al., 2002; Simmons and Wallschlager, 2005). Accumulation of
Se in sediments also presents a risk to benthic organisms that
ingest Se(0) and Se( -II) species and convert them to organo -
selenium species (Fan et al., 2002). Anaerobic sediments also act
as a source of selenium to the water column if the sediments are
disturbed in a way that results in the oxidation and remobilization
of reduced Se species (Belzile et al., 2000; Simmons and
Wallschlager, 2005). The constant risk of bioaccumulation, and
the latent risk of remobilization in Se -contaminated sediments
should be a major consideration for ash spill remediation.
3.5. Implications for coal ash management
This study's confirmation that redox potential is a key param-
eter in controlling As and Se mobilization during an ash spill brings
into question the applicability of the leaching tests currently used
to assess environmental risks. Coal ash management is guided by
the EPA's Toxicity Characteristic Leaching Protocol (US EPA, 1992), a
leaching test performed under aerobic conditions at a single pH
value (pH = 4.9 or 2.4). Other EPA methods such as the Leaching
Environmental Assessment Framework evaluates contaminant
leaching over a wide range of pH values, but the tests still fail to
account for complexity in the real environment (US EPA, 2012b,
2012c, 2013b, 2013c). Likewise, previous experiments with N2 -
purged water -ash mixtures did not result in changes to Se and As
speciation in coal ash (Bednar et al., 2010; Liu et al., 2013). The data
from our study suggests that the absence of oxygen, alone, is
insufficient for testing contaminant mobilization in anaerobic
conditions relevant for ash impoundments and ash spill sites.
Instead, microbially-driven redox transitions, which can be stimu-
lated by sulfate from the coal ash, are more environmentally rele-
vant and necessary for attaining sufficiently reducing conditions for
transformations of As, Se, and possibly other contaminants (e.g.,
mercury, chromium, etc.). Moreover, the impacts of the redox
transitions are likely to vary in degree according to the geochemical
properties of the coal ash, the sediment, as well as the composition
of the microbial community. All these considerations are needed in
the future improvements of standardized methods for coal ash risk
assessments.
Finally, these results are helpful for identifying suitable closure
methods for ash impoundments. The U.S. EPA now requires the
closure of ash ponds that show a risk of groundwater contamina-
tion or that are improperly sited (US EPA, 2014b). Likewise, recent
regulations in North Carolina require the closure of all the State's
ash impoundments by 2029; those designated as high-risk must be
closed by 2020 (US EPA, 2014a). One proposed closure method is
the "Cap in Place" approach, where the ash pond would be de -
watered and then covered with a porous or non -porous cap
(Duke Energy, 2015). One concern with this method is that the cap
could alter redox conditions in the impoundment, and this study
shows that such changes could enhance the release of soluble
arsenic into local groundwater. Thus, even if no previous ground-
water contamination issues have been reported, capping methods
that might induce anaerobic conditions should be avoided in the
closure of unlined impoundments. Overall, this research shows the
need to consider both coal ash characteristics and environmental
parameters when assessing the environmental risks of ash disposal.
Acknowledgments
We thank the Tennessee Valley Authority, Restoration Services,
and Environmental Services for their assistance with field sample
collection. We also thank Kaitlyn Porter for her assistance with ICP -
MS measurements. This work was supported by the National
Science Foundation (CBET-1235661). G. Schwartz was also partly
supported by a doctoral scholarship from the Environmental
Research and Education Foundation.
Appendix ASupplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.apgeochem.2016.02.013.
References
Allen, H.E., Fu, G., Deng, B., 1993. Analysis of acid -volatile sulfide (AVS) and
simultaneously extracted metals (SEM) for the estimation of potential toxicity
in aquatic sediments. Environ. Toxicol. Chem. 12,1441-1453.
American Coal Ash Association, Coal Combustion products production & use sta-
tistics; https://www.acaa-usa.org/Publications/Production-Use-Reports.
Antonioli, P., Lampis, S., Chesini, L, Vallini, G., Rinalducci, S., Zolla, L., Righetti, P.G.,
2007. Stenotrophomonas maltophilia SeITE02, a new bacterial strain suitable
for bioremediation of selenite-contaminated environmental matrices. Appl.
Environ. Microbial. 73, 6854-6863.
Bartov, G., Deonarine, A., Johnson, T.M., Ruhl, L., Vengosh, A., Hsu -Kim, H., 2012.
Environmental impacts of the Tennessee valley authority Kingston coal ash
spill. 1. Source apportionment using mercury stable isotopes. Environ. Sci.
Technol. 47, 2092-2099.
Bednar, A.J., Chappell, MA., Seiter, J.M., Stanley, J.K., Averett, D.E., Jones, W.T.,
Pettway, B.A., Kennedy, A.J., Hendrix, S.H., Steevens, J.A., 2010. Geochemical
investigations of metals release from submerged coal fly ash using extended
elutriate tests. Chemosphere 81, 1393-1400.
Bednar, A.J., Garbarino, J.R., Ranville, J.F., Wildeman, T.R., 2002. Preserving the dis-
tribution of inorganic arsenic species in groundwater and acid mine drainage
samples. Environ. Sci. Technol. 36, 2213-2218.
Belzile, N., Chen, Y. -W., Xu, R., 2000. Early diagenetic behaviour of selenium in
freshwater sediments. Appl. Geochem. 15, 1439-1454.
Burton, E.D., Johnston, S.G., Kocar, B.D., 2014. Arsenic mobility during flooding of
contaminated soil: the effect of microbial sulfate reduction. Environ. Sci.
Technol. 48,13660-13667.
Chappell, M.A., Seiter, J.M., Bednar, A.J., Price, C.L., Averett, D., Lafferty, B.,
Tappero, R., Stanley, J.S., Kennedy, A.J., Steevens, J.A., Zhou, P., Morikawa, E.,
Merchan, G., Roy, A., 2014. Stability of solid -phase selenium species in fly ash
after prolonged submersion in a natural river system. Chemosphere 95,
174-181.
Deonarine, A., Bartov, G., Johnson, T.M., Ruhl, L., Vengosh, A., Hsu -Kim, H., 2013.
Environmental impacts of the Tennessee valley authority Kingston coal ash
spill. 2. Effect of coal ash on methylmercury in historically contaminated river
sediments. Environ. Sci. Technol. 47, 2100-2108.
Deonarine, A., Kolker, A., Doughten, M.W., 2015. Trace elements in coal ash. US
Geological survey.
Duke Energy, 2015. Ash Management.
Fan, T.W.M., Teh, S.J., Hinton, D.E., Higashi, R.M., 2002. Selenium biotransformation
into proteinaceous forms by foodweb organisms of selenium -laden drainage
waters in California. Aquat. Toxicol. 57, 65-84.
Fernandez -Martinez, A., Charlet, L., 2009. Selenium environmental cycling and
bioavailability: a structural chemist point of view. Rev. Environ. Sci. Biotechnol.
8,81-110.
Goldberg, S., Johnston, C.T., 2001. Mechanisms of arsenic adsorption on amorphous
oxides evaluated using macroscopic measurements, vibrational spectroscopy,
and surface complexation modeling. J. Colloid Interface Sci. 234, 204-216.
Huggins, F.E., Senior, C.L., Chu, P., Ladwig, K., Huffman, G.P., 2007. Selenium and
arsenic speciation in fly ash from full-scale coal -burning utility plants. Environ.
Sci. Technol. 41, 3284-3289.
Hunter, W.J., Kuykendall, L.D., 2007. Reduction of selenite to elemental red selenium
by Rhizobium sp. strain B1. Curr. Microbial. 55, 344-349.
Hunter, W.J., Manter, D.K., 2009. Reduction of selenite to elemental red selenium by
Pseudomonas sp. strain CAS. Curr. Microbio]. 58, 493-498.
Izquierdo, M., Querol, X., 2012. Leaching behaviour of elements from coal com-
bustion fly ash: an overview. Int. J. Coal Geol. 94, 54-66.
Kim, N.H., Mason, C.C., Nelson, RG., Afton, S.E., Essader, A.S., Medlin, J.E.,
Levine, K.E., Hoppin, J.A., Lin, C., Knowler, W.C., 2013. Arsenic exposure and
incidence of type 2 diabetes in southwestern American Indians. Am. J. Epi-
demiol. 177, 962-969.
Lemly, A.D., 1993. Guidelines for evaluating selenium data from aquatic monitoring
and assessment studies. Environ. Monit. Assess. 28, 83-100.
Lemly, A.D., 2004. Aquatic selenium pollution is a global environmental safety issue.
Ecotoxicol. Environ. Saf. 59, 44-56.
Liu, Y. -T., Chen, T. -Y., Mackebee, W.G., Ruhl, L., Vengosh, A., Hsu -Kim, H., 2013. Se-
lenium speciation in coal ash spilled at the Tennessee valley authority Kingston
site. Environ. Sci. Technol. 47,14001-14009.
Masscheleyn, P.H., Delaune, RD., Patrick, W.H., 1991. Effect of redox potential and
pH on arsenic speciation and solubility in a contaminated soil. Environ. Sci.
Technol. 25,1414-1419.
Meij, R., 1994. Trace element behavior in coal-fired power plants. Fuel Process.

G.E. Schwartz et al. / Applied Geochemistry 67 (2016) 177-185
Technol. 39,199-217.
Milstein, L.S., Essader, A., Pellizzari, E.D., Fernando, R.A., Raymer, J.H., Levine, K.E.,
Akinbo, 0., 2003. Development and application of a robust speciation method
for determination of six arsenic compounds present in human urine. Environ.
health Perspect. 111, 293.
O'Day, P.A., Vlassopoulos, D., Root, R., Nelson, R., Turekian, K.K., 2004. The influence
of sulfur and iron on dissolved arsenic concentrations in the shallow subsurface
under changing redox conditions. Proc. Natl. Acad. Sci. U. S. A. 101,
13703-13708.
Raven, K.P., Jain, A., Loeppert, R.H., 1998. Arsenite and arsenate adsorption on fer-
rihydrite: Kinetics, equilibrium, and adsorption envelopes. Environ. Sci. Tech-
nol. 32, 344-349.
Rivera, N., Kaur, N., Hesterberg, D., Ward, C.R., Austin, R.E., Duckworth, O.W., 2015.
Chemical composition, speciation, and elemental associations in coal fly ash
samples related to the Kingston ash spill. Energy & Fuels 29, 954-967.
Root, R.A., Dixit, S., Campbell, K.M., Jew, A.D., Hering, J.G., O'Day, P.A., 2007. Arsenic
sequestration by sorption processes in high -iron sediments. Geochimica Cos-
mochimica Acta 71, 5782-5803.
Rowe, C.L., 2014. Bioaccumulation and effects of metals and trace elements from
aquatic disposal of coal combustion residues: Recent advances and recom-
mendations for further study. Sci. Total Environ. 485-486, 490-496.
Ruhl, L., Vengosh, A., Dwyer, G., Hsu -Kim, H., Schwartz, G., Romanski, A., Smith, S.D.,
2012. The impact of coal combustion residue effluent on water resources: a
North Carolina example. Environ. Sci. Technol. 46,12226-12233.
Ruhl, L., Vengosh, A., Dwyer, G.S., Hsu -Kim, H., Deonarine, A., 2010. Environmental
impacts of the coal ash spill in Kingston, Tennessee: an 18 -Month survey. En-
viron. Sci. Technol. 44, 9272-9278.
Ruhl, L., Vengosh, A., Dwyer, G.S., Hsu -Kim, H., Deonarine, A., Bergin, M.,
Kravchenko, J., 2009. Survey of the potential environmental and health impacts
in the immediate aftermath of the coal ash spill in Kingston, Tennessee. Envi-
ron. Sci. Technol. 43, 6326-6333.
Sharma, V.K., Sohn, M., 2009. Aquatic arsenic: Toxicity, speciation, transformations,
and remediation. Environ. Int. 35, 743-759.
Simmons, D.B.D., Wallschlager, D., 2005. A critical review of the biogeochemistry
and ecotoxicology of selenium in lotic and lentic environments. Environ. Tox-
icol. Chem. 24,1331-1343.
Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour and dis-
tribution of arsenic in natural waters. Appl. Geochem. 17, 517-568.
Suess, E., Wallschlager, D., Planer -Friedrich, B., 2011. Stabilization of thioarsenates in
185
iron -rich waters. Chemosphere 83,1524-1531.
Thorneloe, S.A., Kosson, D.S., Sanchez, F., Garrabrants, A.C., Helms, G., 2010. Evalu-
ating the fate of metals in air pollution control residues from coal fired power
plants. Environ. Sci. Technol. 44, 7351-7356.
US EPA, 1992. Method 1311: Toxicity Characteristic Leaching Procedure. United
States Environmental Protection Agency, Washington, D.C.
US EPA, 2007. Coal Combustion Waste Damage Assessments. United States Envi-
ronmental Protection Agency, Washington, DC.
US EPA, 2012a. Information Request Responses from Electric Utilities: Dataset
Results.
US EPA, 2012b. Method 1313: Liquid Solid Partitioning as a Function of Extract pH
using a Parallel Batch Extraction Procedure. United States Environmental Pro-
tection Agency, Washington, D.C.
US EPA, 2012c. Method 1316: Liquid -Solid Partitioninig as a Function of Liquid -To -
Solid Ratio in Solid Materials Using a Parallel Batch Procedure. United States
Environmental Protection Agency, Washington, D.C.
US EPA, 2013a. Environmental Assessment for the Proposed Effluent Limitations
Guidelines and Standards for the Steam Electric Power Generating Point Source
Category. United States Environmental Protection Agency, Washington, DC.
US EPA, 2013b. Method 1314: Liquid -Solid Partitioning as a Function of Liquid -Solid
Ratio for Constituents in Solid Materials Using an Up -Flow Percolation Column
Procedure. United States Environmental Protection Agency, Washington, D.C.
US EPA, 2013c. Method 1315: Mass Transfer Rates of Constituents in Monolithic or
Compacted Granular Materials Using a Semi -Dynamic Tank Leaching Procedure.
United States Environmental Protection Agency, Washington, D.C.
US EPA, 2013d. Technical Development Document for the Proposed Effluent Limi-
tations Guidelines and Standards for the Steam Electric Power Generating Point
Source Category. United States Environmental Protection Agency, Washington,
D.C.
US EPA, 2014a. The Coal Ash Management Act of 2014 (Session Law 2014-122,
Senate Bill 729). North Carolina State Legislature, Raleigh, NC.
US EPA, 2014b. Hazardous and Solid Waste Management System; Disposal of Coal
Combustion Residuals from Electric Utilities. U.S. Environmental Protection
Agency, Washington, D.C.
Wilkin, R.T., Wallschlager, D., Ford, R.G., 2003. Speciation of arsenic in sulfidic
waters. Geochemical Transactions 4,1-7.
Zheng, S., Su, J., Wang, L., Yao, R., Wang, D., Deng, Y., Wang, R., Wang, G., Rensing, C.,
2014. Selenite reduction by the obligate aerobic bacterium Comamonas tes-
tosterone S44 isolated from a metal -contaminated soil. BMC microbiol.14, 204.