Chemosphere 78 (2010) 12951300

Contents lists available at ScienceDirect

Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere

Ultrasonic extraction of arsenic and selenium from rocks associated

with mountaintop removal/valley lls coal mining: Estimation of bioaccessible
concentrations
I. Pumure a,*, J.J. Renton b, R.B. Smart c
a
College of Science and Technology, Department of Biochemistry, Chemistry and Physics, University of Central Missouri, PO Box 800, Warrensburg, Mo 64093, USA
b
Department of Geography and Geology, West Virginia University, PO Box 6300, Morgantown, WV 26506, USA
c
C. Eugene Bennett Department of Chemistry, West Virginia University, PO Box 6045, Morgantown, WV 26506, USA

a r t i c l e i n f o a b s t r a c t

Article history: Ultrasonic extraction (UE) was used to estimate the total bioaccessible fractions of arsenic and selenium
Received 21 July 2009 released from rocks associated with mountaintop removal/valley ll coal mining. The combined readily
Received in revised form 14 January 2010 bioaccessible amounts of arsenic and selenium in water soluble, exchangeable and NaOH fractions can be
Accepted 14 January 2010
extracted from the solid phase within a 20 or 25 min application of 200 W cm2 ultrasound energy in
Available online 2 February 2010
nanopure water for selenium and arsenic, respectively. Application of a two-way ANOVA predicted that
there are no signicant differences (p 6 0.001, n = 12) in the extracted arsenic and selenium concentra-
Keywords:
tions between the combined bioaccessible and ultrasonic extracts. The mechanisms for the UE of arsenic
Ultrasonic extraction
Selenium
and selenium are thought to involve the formation of secondary minerals on the particle surfaces which
Arsenic eventually dissolve with continued sonication. This is supported by the presence of transient SiO
Mountaintop removal/valley ll stretching and OH absorption and bending ATR-FTIR peaks at 795.33 cm1, 696.61 cm1 and
Sequential extraction 910.81 cm1. The subsequent dissolution of secondary minerals is followed by the release of chemical
Bioavailability species that include selenium and arsenic. Release rates decrease after the ultrasound energy elastic limit
for the particles is reached. Selenium and arsenic are bound differently within the rock lattice because no
selenium was detected in the acid soluble fraction and no arsenic was found in the exchangeable fraction.
However, selenium was found in the exchangeable fraction and arsenic was found in the acid soluble
fraction. The characterization of coal associated rocks is essential to the design of methodologies and pro-
cedures that can be used to control the release of arsenic and selenium from valley lls.
Published by Elsevier Ltd.

1. Introduction concentrations of arsenic and selenium that can be released from

Ultrasonic extraction (UE) is an easy, convenient, and fast
(Luque-Garcia and de Castro, 2003) way of desorbing inorganic
(Perez-Cid and Bola, 2001) and organic pollutants (Tse and Lo,
2002) from sediments, soil and biological samples (Arruda et al.,
2003). UE enhances the dissolution process by producing acoustic
cavitations that result in bubble formation and their subsequent
collapse generating high-pressure and temperature gradients
which increase chemical reactivities (Suslick, 1989). It is also an
aseptic method and it has been suggested that high frequency
ultrasound (20 kHz) in low volumes (200 mL) of bacterial suspen-
sions resulted in a continuous reduction of bacterial cell numbers
(Joyce et al., 2003). Thus, microbial sample alteration or modica-
tion after UE is minimized if samples are placed in aseptic contain-
ers. We have applied UE to assess the total bioaccessible
* Corresponding author. Tel.: +1 660 543 4143.
E-mail address: pumure@uemo.edu (I. Pumure).
rocks associated with mountaintopvalley ll (MTR/VF) coal min-
ing and the ultrasonic extractions were compared with those ob-
tained using a sequential extraction technique. In our previous
work, UE was used to study the kinetics of accelerated hysteresis
of arsenic and selenium from coal mining samples (Pumure et al.,
2009).
MTR/VF coal mining is a process that involves the removal of
mountaintops to expose beds of coal. The process begins with
the removal vegetation covering the mountaintop areas of interest.
The waste rock material generated by the mining phase is placed in
adjacent valleys and streams creating valley lls. Hartman et al.
(2005) found reduced densities of microinvertebrates in streams
affected by MTR/VF mining which they proposed was due to the
elevated conductivity caused by increased concentrations of ions
(Na, K, Mn, Mg, Ca, Ni and Fe) and ne sediment compared to ref-
erence streams. Valley lls have been found to leach a variety of
metals that disturb the ecological balance in natural aquatic
systems (Dreher and Finkelman, 1992). Elevated concentrations

0045-6535/$ - see front matter Published by Elsevier Ltd.

doi:10.1016/j.chemosphere.2010.01.020
1296 I. Pumure et al. / Chemosphere 78 (2010) 12951300
of arsenic and selenium have been shown to stunt the growth of
some plants (Lejeune et al., 1996; Terry et al., 2000), a critical pro-
cess required in the remediation of lled sites.
Traditionally, selenium and arsenic in solid samples have been
extracted with water, NaOH, phosphate buffers, oxidizing agents
such as K2S2O8, and a variety of concentrated acids with either
open vessel or microwave digestions (Olivas et al., 1994; Perez-Cid
and Bola, 2001). Microwave digestions with strong acids accom-
plish complete dissolution of the solid samples. This results in
the release of both the loosely and strongly bound metals together
with residual metals that are closely associated with silicate matri-
ces. Such extraction methods, although useful in estimating the
entire amount of available metals in solids, do not provide the
essential information on the readily bioaccessible concentrations
that are immediately leached into aquatic systems soon after val-
leys are lled.
This necessitates the use of analytical methods that extracts
only the loosely bound metals, which are readily mobile. Their con-
centrations can quickly increase under conditions that favor high
residence times. Acid Leaching, Standard Tessier, and the Stan-
dards, Measurements and Testing program (SMT) formerly known
as BCR (Community Bureau of Reference, Commission of the Euro-
pean communities) methods are very useful in estimating the
amount of metals available in different phases (Van Herreweghe
et al., 2003).
The initial stage of the BCR method uses a weak acid (acetic
acid) and in Standard Tessier extraction methods use a highly con-
centrated chloride salt (NaCl, KCl or MgCl2) to estimate both water
soluble and exchangeable metals in solid samples. In the Acid
Leaching method, 0.014 M HCl (Ochsenkuhn-Petropoulou and
Tsopelas, 2002) is used to represent a HCl extractable fraction,
which is often related to the total concentration of selenium that
can be released over an innite period of time. Caution must be
exercised when using acids in acidic-based extractions. High chlo-
ride concentrations used in BCR and HCl based leaching protocols,
although essential to displace the oxoanions from the solid sur-
faces by ion exchange, can also reduce Se (VI) to Se (IV) (Lu and
Yan, 2005) during the course of extraction.
It is well known that the toxicity and the bioavailability of a
chemical species is dependent on both the total concentration
and, often more importantly, on the chemical form. The use of
acidic media, oxidizing, and reducing agents will often change
the oxidation state of metals making the methods only good
for the quantitation of species stable in that particular medium.
There is a general consensus that either hot or cold deionized/
nanopure water should be used as an extracting solvent because
it has the ability to preserve the natural oxidation state of metals
being leached (Zhang and Frankenberger, 2003; Bujdos et al.,
2005).
Attempts have been made to use alkaline phosphate buffers and
sodium hydroxide solutions to leach and preserve the oxidation
states of selenium species (Sharmasarkar and Vance, 2002). Phos-
phate buffer extractions have been found to remove both
exchangeable and strongly bound selenium in soils with insigni-
cant modications of the oxidation states and chemical form (Mar-
tens and Suarez, 1996). The loosely bound portion that comprises
the water soluble and exchangeable fractions is the most
bioavailable.
Once released from valley lls, selenium and arsenic can enter
many different pathways that are dependent on the physicochem-
ical properties of the water and sediment. The combined concen-
trations of: (i) soluble amounts released into the liquid phase
and (ii) amounts retained by the solid phase that can be easily
transferred to the liquid phase is termed bioavailable (Rieuwerts
et al., 1998). Accordingly, the concentrations of bioavailable spe-
cies are made up two fractions the intensity and quantity factors.
The intensity factors represent a fraction that is already available
in solution and therefore can be taken up by living organisms.
The quantity factor represents the concentration of species that
might take time for them to migrate from the solid phase into
the liquid solution and are not of immediate availability. However,
others classify the intensity factors as being bioavailable whilst the
combined intensity and quantity factors as being bioaccessible
(Semple et al., 2004).
The fate of the released selenium and arsenic also depends on
the environmental conditions that include pH, electrochemical
properties, presence or absence of complexing agents such as nat-
ural organic matter (humic and fulvic acids) and other reactive or-
ganic and inorganic materials (Lahermo et al., 1998). These
environmental conditions can result in oxidation state changes
and the sequestration or remobilization of selenium and arsenic,
thereby controlling their bioavailabilities and toxicities.
2. Experimental
2.1. Samples
The rock samples were collected by the West Virginia Geologi-
cal and Economic Survey (WVGES) from Kayford Mountain which
is situated at the intersection of Boone and Kanawha Counties,
West Virginia, USA (81210 3100 W, 37580 4200 ). Mineral analyses
were performed using a Philips PW1200 X-ray diffraction unit,
and elemental analyses were performed using a Philips 1420 X-
ray spectrometer.
All samples had appreciable amounts of Fe2O3 (0.874.82%),
Al2O3 (15.7023.55%), K2O (2.475.75%) and MgO (0.432.28%)
with the sandstone (BT61) having the least amounts. MnO ranged
from 0.020.05%. The levels of SiO2 were very high 6279% and
BT61 had the highest value of 79.51%.
BT61 had the highest amount of quartz (75.3%) compared to
BT241, a shale that had 54.0% and the two claystones (BT572 and
BT701) that varied from 25.227.7%. The variation of 14 clays
(e.g. vermiculite, montmorrilonite, chlorite, etc.) was within +19%
and illite +35% with BT61 having the smallest amounts. Kaolinite
composition in all the four samples were within +5%.
3. Materials and methods
3.1. Ultrasound extraction
Four pulverized rock samples (BT61, BT241, BT572 and BT701)
previously sieved through a 60 lm mesh, were each subjected to
12 parallel batch sonicated extractions in nanopure water at
24 + 1 C. The extractions with 200 W cm2 of continuous sonica-
tion were done in intervals of 5 min up to a total period of
60 min. The mixtures were then centrifuged at 3400g for 20 min
and the supernatant solutions were decanted with a Pasteur pip-
ette and placed in cleaned and previously acid soaked 25 mL poly-
ethylene vials. The individual extracts were ltered and analyzed
for total selenium and arsenic by graphite furnace atomic absorp-
tion spectrophotometry and Se (IV) by hydride generation atomic
absorption spectrometry (Varian Model 55B). Previously, As (III)
was found to be unstable during ultrasound extraction because it
is oxidized to As (V) (Pumure et al., 2009). Total available selenium
and arsenic were obtained by microwave digestion as described
previously (Pumure et al., 2009).
Particle size distribution for each sonicated fraction was deter-
mined using a Shimadzu SALD 3101 laser diffraction particle ana-
lyzer after the particles were freeze dried to preserve their
chemical and physical surface characteristics (He et al., 2005). A
1% suspension of each freeze dried rock samples was used.
 I. Pumure et al. / Chemosphere 78 (2010) 12951300 1297

3.2. ATR-FTIR
Residues from the UE samples were ltered and freeze dried fol-
lowed by recording Attenuated Total Reectance-Fourier Trans-
form Infrared (ATR-FTIR) spectra from 630 cm1 to 4000 cm1
using a Bruker Hyperion 3000 spectrometer. The normalized
absorption-reectance signals at 696.61 cm1 (OH from trioctahe-
drals), 795.33 cm1 (SiO stretch) and 910.81 cm1 (Al2OH bend-
ing) (Bishop et al., 2002; Madejova, 2003; Klinkenberg et al.,
2006) were monitored as a function of continuous sonication time
for a period of 90 min.
titate the reducible diselenides, the other half of the residue was
reacted with CrCl2 (CrCl3 passed through a Jones reactor a col-
umn consisting of 2% Hg and 98% Zn) in 5 mL concentrated HCl
(Manjanna et al., 2003). The mixtures were heated at 90 C for
2 h while bubbling a stream of N2 into the mixture to allow the re-
lease of H2Se (and possibly H2S) gas. The remaining solid was ana-
lyzed for selenium and arsenic after microwave digestion. The
difference in Se concentrations between the CrCl2 treated and un-
treated residues was attributed to the concentration of diselenides.
4. Results and discussion

3.3. Sequential extraction 4.1. Batch sonicated extracts

A combined modication of sequential extraction methods was The sonication of 1 g samples in 10 mL nanopure water resulted
used in the sequential extraction procedure (Tessier, 1979; Kulp in the slow initial release of both arsenic and selenium as shown in
and Pratt, 2004). The extractions were done in triplicate. Figs. 13. The concentrations peaked at around 20 min followed by
small decreases thereafter. Continuous UE for 20 and 25 min was
3.3.1. Water soluble the optimum time period required for the release of selenium
A 1.0 g sample and 10 mL nanopure water were mixed in a and arsenic, respectively. The decrease in concentration after 20
20 mL polyethylene vial on an end over mechanical shaker for and 25 min is attributed to the resorption of the previously des-
2 h spinning at 90 rpm. The mixture was centrifuged for 20 min orbed species and agglomeration of ne particles (He et al.,
at 3400g. The clear supernatant liquid was siphoned with a Pasteur 2007). This resorption is facilitated by the formation of smaller
pipette without disturbing the settled residue into a cleaned and particles with a large specic surface area available for chemisorp-
previously acid soaked polyethylene vial. The residue was washed tion. Since selenium and arsenic require different sonication times
with 1 mL nanopure water and the washings were decanted and for optimal extraction they may occupy different sites in the rock
disregarded (Pueyo et al., 2003). matrix and that arsenic could be strongly bound compared to
selenium.
3.3.2. Exchangeable
The residue from the water soluble fraction was transferred
4.2. Particle size
quantitatively into a 20 mL polyethylene vial using 10 mL of
0.1 M Na2PO4NaHPO4 at pH 7. The resulting supernatant and
The particle size distributions of the freeze dried residues were
the residue were treated the same way as in step (i).
measured and maximum particle disintegrations were observed to
occur between 0 and 25 min of continuous UE. The mean particle
3.3.3. Base soluble
size decreased and remained fairly constant after 20 min of contin-
Residues from the exchangeable fraction were quantitatively
uous UE as shown in Fig. 4. Hence, the elastic limit of the rock
transferred into a 20 mL polyethylene vial using 10 mL 0.1 M
particles was achieved within 2025 min of continuous UE where
NaOH. The vials were placed in a water bath maintained at 90 C
no further signicant particle disintegrations occur. The elastic
for 2 h followed by centrifugation at 3400g for 20 min and the
limit corresponds to a point where the incident ultrasound energy
resulting supernatant and the residue were treated the same way
absorbed by the particles will be dissipated as heat without caus-
as in step (i).
ing any considerable particle damage (Tavares and King, 2002;
Tavares, 2009). It was also observed that 2025 min of continuous
3.3.4. Elemental Se
UE coincided with the highest amounts of arsenic and selenium re-
Elemental selenium dissolves in sulte solution to form soluble
leased with 20 and 25 min being optimal for selenium and arsenic
thioselenate. A 10-mL of aliquot of 0.5 M Na2SO3 (adjusted to pH 7)
extractions, respectively. The ratio between the mean particle size
was used to dissolve elemental selenium in the rock residues ob-
and the modal particle size for each sample also remained constant
tained from the step (iii). The mixtures were sonicated at
throughout the sonication process. This suggests a uniformity in
200 W cm2 for 20 min, centrifuged and the supernatant liquid
rock composition with a similar disintegration pattern regardless
was separated and the remaining residue was used in the following
of particle size. Further, the ATR-FTIR data at 696.61 cm1 (OH
stage.
absorption from trioctahedrals), 795.33 cm1 (SiO stretch) and
910.81 cm1 (Al2OH bending) shown in Fig. 5 suggest the forma-
3.3.5. Acid soluble
The residue from step (iv) was quantitatively transferred into a
20 mL vial using 10 mL 15% acetic acid solution. The solutions were
allowed to stand for at least 15 min until any effervescence sub-
sided. The vials were then placed on an end over mechanical sha-
ker for 2 h to ensure complete dissolution of acid soluble material
in the sample. The mixtures were centrifuged followed by separa-
tion of the supernatant. The residue was dried at 80 C and the
supernatant was analyzed for arsenic and selenium.

3.3.6. Residual
Half of the dried residue from the acid soluble extraction was
microwave digested using a HF/HNO3/HCl (5:5:3) acid combina-
tion. This fraction is composed of the inert material that is closely
associated with siliceous material in rocks. In an attempt to quan- Fig. 1. Selenium (IV) concentration released in batch sonicated extracts.
1298 I. Pumure et al. / Chemosphere 78 (2010) 12951300

were subsequently dissolved and simultaneously released most

ions including arsenic and selenium. Referring to Figs. 15, the
majority of selenium and arsenic containing secondary minerals
were formed and dissolved during the rst 2025 min of continu-
ous UE.

4.3. Effect of extracting solvent on UE

Various solvents (4 M HCl, 0.1 M NaOH, and 0.1 M Na2PO4

Fig. 2. Total selenium concentration released in batch sonicated extracts.
NaHPO4 pH 7 buffer) were compared by measurement of the con-
centration of selenium and arsenic released after 20 and 25 min of
continuous UE, respectively. These solvents were chosen because
they are commonly used in the approximation of extractable and
bioavailable metals from rocks solids (Krysiak and Karczewska,
2007). The extracts were analyzed for Se (IV) concentration, total
Se and total As and the results are shown in Table 1. Comparable
concentrations of total selenium were extracted using ultrasound
assisted nanopure water, 4 M HCl or 0.1 M Na2PO4NaHPO4 pH 7
buffer. From 10 to 80% more Se (IV) and total selenium were ex-
tracted using 0.1 M NaOH compared to the other solvents. This is
because of the high alkaline pH, NaOH can dissolve natural organic
matter which can contain appreciable amounts of bound selenium.
In BT241, BT572 and BT701 shale and claystone samples, the total
arsenic concentrations in the various extracts was approximately
10% of the total available arsenic as obtained by microwave diges-
tions. However, in BT61, a sandstone, the extractable As was
around 50%.

Fig. 3. Total arsenic released in batch sonicated extracts.

4.4. Sequential extractions

Results from the sequential extraction method are summarized

in Figs. 68. Elemental selenium was found to be below the detec-
tion limit of 0.5 ng mL1 in all samples analyzed. No signicant dif-
ferences in concentrations were found between the CrCl2 treated
and untreated residues and it was concluded that there was no
measurable diselenide in the rocks analyzed.
Total extracted selenium was found to be primarily contained in
the water soluble, exchangeable, base soluble and the residual frac-

Table 1
Fig. 4. Variation of mean particle size with sonication time. Concentrations of As and Se extracted after 20 min (Se species) and 25 min (As
species) of continuous sonication in different extraction media.

ng g1Se (IV) ng g1 Total Se lg g1 Total As

BT61
Nanopure water 4.6 0.2 17.2 0.1 2.88 0.15
4 M HCl 4.0 0.3 20.2 0.9 2.99 0.03
P-Buffer 3.6 0.5 19.2 2.5 2.74 0.09
0.1 M NaOH 4.9 0.3 86.3 10 2.78 0.06
Microwave digestion 90.4 4.5 5.44 0.21
BT241
Nanopure water 70.0 1.4 169 15 5.99 0.25
4 M HCl 41.9 2.8 180 14 6.05 0.14
P-Buffer 68.6 1.6 192 22 4.24 0.04
0.1 M NaOH 95.0 3.8 265 16 4.46 0.01
Microwave digestion 717 7 36.7 0.5
BT572
Fig. 5. Variation in attenuated total reectance at 696.61 cm1, 795.33 cm1 and Nanopure water 82.1 2.3 236 13 4.48 0.15
910.81 cm1. 4 M HCl 63.4 4.3 267 28 5.55 0.38
P-Buffer 85.6 1.3 213 15 4.44 0.02
0.1 M NaOH 130 5 286 25 4.28 0.03
tion of intermediates on the surfaces of the particles that disappear Microwave digestion 632 10 34.4 1.8
with prolonged sonication. The normalized peak intensities also
BT701
reached their respective maxima within 2030 min of continuous Nanopure water 92.8 4.0 239 17 5.29 0.43
UE followed by a gradual decrease. These peaks correspond to 4 M HCl 69.0 2.7 272 15 4.88 0.13
reection bands of secondary minerals being formed during the P-Buffer 83.9 1.2 253 22 5.05 0.07
accelerated weathering process. The characteristic transient peaks 0.1 M NaOH 214 4 350 28 5.28 0.06
Microwave digestion 675 7 40.9 1.3
could be attributed to the formation of secondary minerals that
 I. Pumure et al. / Chemosphere 78 (2010) 12951300
Fig. 6. Comparison of Se (IV) obtained using UE and sequential extraction
techniques. Combined water, exchangeable and NaOH soluble extracts are consid-
ered bioaccessible.
Fig. 7. Comparison of total Se obtained using UE and sequential extraction
techniques. Combined water, exchangeable and NaOH soluble extracts are consid-
ered bioaccesible.
Fig. 8. Comparison of total As obtained using UE and sequential extraction
techniques. Combined water, exchangeable and NaOH soluble extracts are consid-
ered bioaccessible.
tions. Since selenium concentrations in the acid soluble fraction
were found to be below the detection limit, this observation could
suggest that selenium compounds in the core samples are not
associated with carbonate-containing rock species such as siderite,
calcium and magnesium carbonates. Selenite was found to be lar-
gely associated with three fractions; water soluble, exchangeable
and base soluble. No Se (IV) concentrations were detected in the
microwave digests due to the presence of acid aided oxidation.
The highest amounts of selenite were found in the NaOH fraction
as shown in Fig. 6. Also, the NaOH soluble fraction for BT701 had
a higher total selenium concentration compared to the residual
fraction. This observation suggests that selenium in the BT701
sample is closely associated with organic material that had not
yet been mineralized. Samples BT241 and BT571 samples had total
residual selenium concentrations that are higher than NaOH frac-
tions suggesting that a greater percentage (6080%) of the total
available selenium had already been mineralized and lithied.
As shown in Fig. 8, 1013% of the available arsenic was released
by the combined three extractants; water soluble, base soluble and
acid soluble with 8790% remaining in the residual matrix. No
detectable arsenic was found in the exchangeable fraction, which
1299
suggests that selenium and arsenic occupy different sites in the
coal associated rocks. Although only 1013% of the arsenic was
extractable, the amount extracted was about three orders of mag-
nitude greater than selenium. The acid soluble fraction contained
substantial arsenic, which also suggests that some arsenic species
are associated with carbonate-containing rock constituents such
as siderite or calcium and magnesium carbonates.
4.5. Comparison of bioaccessible fractions determined by sequential
and ultrasound extractions
The comparison between the bioaccessible fraction (water solu-
ble + exchangeable + base soluble + acid soluble) and the UE for
selenite, total selenium (20 min), and arsenic (25 min) are also
summarized in Figs. 68. The application of a two-way ANOVA be-
tween the total combined bioaccessible As and Se concentrations
against their corresponding concentrations from UE concluded that
there is no signicant difference between the two data sets
(p 6 0.001, n = 12). Single-laboratory validation of an analytical
method can also be assessed using the Horwitzs criterion (Horwitz
and Albert, 1995; Thompson et al., 2002) acceptable %
RSD = 2[C]0.1505, where C is the concentration factor. The ratio of
the % RSD between the two different extraction methods to the
critical % RSD obtained by the Horwitz formula is known as the
Horwitz ratio (HorRat). Acceptable HorRat values for single labora-
tory method validation must be less than 1.3 (Horwitz and Albert,
2006). The % RSD values between the averaged bioaccessible and
sonication fractions for total selenium analysis were all less than
corresponding Horwitzs precision criterion, with HorRat values
ranging from 0.27 to 0.92. The HorRat values for total arsenic
determination in samples BT61, BT241 and BT571 ranged from
0.61 to 1.2 with the exception of sample BT701 which was 3.8.
The large difference in arsenic concentrations (>40% RSD) obtained
by the two different extraction methods for sample BT701 can be
explained in terms of the contributions from the acid soluble frac-
tion which was small for the other three samples. The acid soluble
fraction could not be extracted by UE because of the inherent
highly alkaline pH values in the extraction cell. Further, the pH of
MTR/VF associated water ranges from 6.5 to 9.0 (Bryant et al.,
2002) which are not acidic enough to decompose carbonate-con-
taining rock constituents. This fraction will likely not be of imme-
diate environmental signicance unless there are incidences of
acidic drainages. Not including the contribution from the acidic
fraction in BT701, gives a HorRat value of 1.2 and the 20 and
25 min UE can be used to estimate the bioaccessible selenium
and arsenic in the core samples.
5. Conclusions
UE is a reliable method for the aqueous extraction of bioacces-
sible selenium and arsenic from coal associated rocks. The solution
composition during UE will reect that of the rock sample matrix
and create pH and redox environments that closely resemble nat-
ural aquatic systems at the solid/solution interface. Twenty min-
utes sonication and 25 min sonication in nanopure water at
200 W cm2 are the optimal conditions for the extraction of bioac-
cessible selenium and arsenic from pulverized rock samples. Statis-
tical analysis suggests that UE is as efcient as the sequential
extraction method for the determination of bioaccessible selenium
and arsenic fractions. Six hours were required to obtain the com-
bined bioaccessible fraction using the sequential extraction meth-
od, whereas only 2025 min was required for UE. Therefore, it
offers a faster direct means of assessing the readily bioaccessible
concentration of arsenic and selenium in coal associated rocks.
The liberation of selenium and arsenic during UE is thought to be
1300 I. Pumure et al. / Chemosphere 78 (2010) 12951300

through the formation of secondary minerals that eventually dis- Lejeune, K., Galbraith, H., Lipton, J., Kapustka, L.A., 1996. Effects of metals and
arsenic on riparian communities in southwest Montana. Ecotoxicology 297,
solve with continued application of ultrasound energy.
312.
Lu, C.Y., Yan, X.P., 2005. Capillary electrophoresis on-line coupled with hydride
generation-atomic uorescence spectrometry for speciation analysis of
Acknowledgements selenium. Electrophoresis 26, 155160.
Luque-Garcia, J.L., de Castro, L., 2003. Ultrasound: a powerful tool for leaching. Trac-
We wish to express our sincere gratitude to the West Virginia Trend Anal. Chem. 22, 4147.
Madejova, J., 2003. FTIR techniques in clay mineral studies. Vib. Spectrosc. 31, 110.
Geological and Economic Survey for providing the pulverized coal Manjanna, J., Venkateswaran, G., Kumbhar, A.G., 2003. Electrolysis and
associated rock samples and the United States Geological Survey electrochemical methods of Cr(II) preparation in bulk amounts: a feasibility
State Water Resources Research Institute Program for providing study. Bull. Electrochem. 19, 541546.
Martens, D.A., Suarez, D.L., 1996. Selenium speciation of soil/sediment determined
nancial assistance for this research.
with sequential extractions and hydride generation atomic absorption
spectrophotometry. Environ. Sci. Technol. 31, 133139.
Ochsenkuhn-Petropoulou, M., Tsopelas, F., 2002. Speciation analysis of selenium
References using voltammetric techniques. Anal. Chim. Acta 467, 167178.
Olivas, R.M., Donard, O.F.X., Camara, C., Quevauviller, P., 1994. Analytical techniques
applied to the speciation of selenium in environmental matrices. Anal. Chim.
Arruda, S.C.C., Rodriguez, A.P.M., Arruda, M.A.Z., 2003. Ultrasound-assisted
Acta 286, 357370.
extraction of Ca, K and Mg from in vitro citrus culture. J. Brazil Chem. Soc. 14,
Perez-Cid, B., Bola, C., 2001. Comparison between total determination and
470474.
extractable heavy metals from river sediments using conventional and
Bishop, J., Murad, E., Dyar, M.D., 2002. The inuence of octahedral and tetrahedral
microwave accelerated leaching tests. Int. J. Environ. Anal. Chem. 81, 101115.
cation substitution on the structure of smectites and serpentines as observed
Pueyo, M., Sastre, J., Hernandez, E., Vidal, M., Lopez-Sanchez, J.F., Rauret, G., 2003.
through infrared spectroscopy. Clay Miner. 37, 617628.
Prediction of trace element mobility in contaminated soils by sequential
Bryant, G., Mcphilliamy, S., Childers, H., 2002. A Survey of the Water Quality of
extraction. J. Environ. Qual. 32, 20542066.
Streams in the Primary Region of Mountaingtop/Valley Fill Coal Mining.
Pumure, I., Renton, J.J., Smart, R.B., 2009. Accelerated aqueous leaching of selenium
WVGES.
and arsenic from coal associated rock samples with selenium speciation using
Bujdos, M., Mulova, A., Kubova, J., Medved, J., 2005. Selenium fractionation and
ultrasound extraction. Environ. Geol. 56, 985991.
speciation in rocks, soils, waters and plants in polluted surface mine
Rieuwerts, J.S., Thornton, I., Farago, M.E., Ashmore, M.R., 1998. Factors inuencing
environment. Environ. Geol. 47, 353360.
metal bioavailability in soils: preliminary investigations for the development of
Dreher, G.B., Finkelman, R.B., 1992. Selenium mobilization in a surface coal-mine,
a critical loads approach for metals. Chem. Spec. Bioavailab. 10, 6175.
powder river basin, Wyoming, USA. Environ. Geol. Water Sci. 19, 155167.
Semple, K.T., Doick, K.J., Jones, K.C., Burauel, P., Craven, A., Harms, H., 2004. Dening
Hartman, K.J., Kaller, M.D., Howell, J.W., Sweka, J.A., 2005. How much do valley lls
bioavailability and bioaccessibility of contaminated soil and sediment is
inuence headwater streams? Hydrobiologia 532, 91102.
complicated. Environ. Sci. Technol. 38, 228A231A.
He, Z.Q., Traina, S.J., Bigham, J.M., Weavers, L.K., 2005. Sonolytic desorption of
Sharmasarkar, S., Vance, G.F., 2002. Soil and plant selenium at a reclaimed uranium
mercury from aluminum oxide. Environ. Sci. Technol. 39, 10371044.
mine. J. Environ. Qual. 31, 15161521.
He, Z.Q., Traina, S.J., Weavers, L.K., 2007. Sonolytic desorption of mercury from
Suslick, K.S., 1989. The chemical effects of ultrasound. Sci. Am. 260, 8086.
aluminum oxide: effects of pH, chloride, and organic matter. Environ. Sci.
Tavares, L.M., 2009. Analysis of particle fracture by repeated stressing as damage
Technol. 41, 779784.
accumulation. Powder Technol. 190, 327339.
Horwitz, W., Albert, R., 1995. Precision in analytical measurements: expected values
Tavares, L.M., King, P., 2002. Modeling of particle fracture by repeated impacts using
and consequences in geochemical analyses. Fresen. J. Anal. Chem. 351, 507513.
continuum damage mechanics. Powder Technol. 123, 138146.
Horwitz, W., Albert, R., 2006. The Horwitz ratio: a useful index of method
Terry, N., Zayed, A.M., DE Souza, M.P., Tarun, A.S., 2000. Selenium in higher plants.
performance with respect to precision. J. Assoc. Off. Anal. Chem. 89, 10951109.
Annu. Rev. Plant Phys. 51, 401432.
Joyce, E., Phull, S.S., Lorimer, J.P., Mason, T.J., 2003. The development and evaluation
Tessier, A.C.P., 1979. Sequential extraction procedure for the speciation of
of ultrasound for the treatment of bacterial suspensions. A study of frequency,
particulate trace metals. Anal. Chem. 51, 844851.
power and sonication time on cultured Bacillus species. Ultrason. Sonochem.
Thompson, M., Ellison, S.L.R., Wood, R., 2002. Harmonized guidelines for single-
10, 315318.
laboratory validation of methods of analysis (IUPAC technical report). Pure
Klinkenberg, M., Dohrmanna, R., Kaufholda, S., Stanjek, H., 2006. A new method for
Appl. Chem. 74, 835855.
identifying Wyoming bentonite by ATR-FTIR. Appl Clay Sci. 33, 195206.
Tse, K.K.C., Lo, S.L., 2002. Desorption kinetics of PCP-contaminated soil: effect of
Krysiak, A., Karczewska, A., 2007. Arsenic extractability in soils in the areas of
temperature. Water Res. 36, 284290.
former arsenic mining and smelting, SW Poland. Sci. Total Environ. 379, 190
Van Herreweghe, S., Vandecasteele, C., Cappuyns, V., 2003. Solid phase speciation of
200.
arsenic by sequential extraction in standard reference materials and
Kulp, T.R., Pratt, L.M., 2004. Speciation and weathering of selenium in upper
industrially contaminated soil samples. Environ. Pollut. 122, 323342.
cretaceous chalk and shale from South Dakota and Wyoming, USA. Geochim.
Zhang, Y.Q., Frankenberger, W.T., 2003. Determination of selenium fractionation
Cosmochim. Acta 68, 36873701.
and speciation in wetland sediments by parallel extraction. Int. J. Environ. Anal.
Lahermo, P., Alfthan, G., Wang, D., 1998. Selenium and arsenic in the environment in
Chem. 83, 315326.
Finland. J. Environ. Pathol. Tox. 17, 205216.