Unformatted Attachment Preview
Environ Earth Sci (2016) 75:306
DOI 10.1007/s12665-015-5203-z
ORIGINAL ARTICLE
Geochemical characterization and modeling of arsenic behavior
in a highly contaminated mining soil
Sara Bisone1 • Vincent Chatain1 • Denise Blanc1 • Mathieu Gautier1
Rémy Bayard1 • Florence Sanchez2 • Rémy Gourdon1
•
Received: 30 July 2015 / Accepted: 27 October 2015 / Published online: 11 February 2016
Ó Springer-Verlag Berlin Heidelberg 2016
Abstract The environmental assessment and management of historical mining sites contaminated with various
inorganic species require a better knowledge of pollutantbearing phases. Among elements present in mining soils,
arsenic is a toxic metalloid with potential high content and
high mobility capacity into the environment. The objective
of this paper was to investigate the mobility and fractionation of arsenic (As) in a highly As contaminated soil (ca. 3
wt%). The soil was collected from an old gold mining site
in France, where mining activities and smelting processes
of gold ores took place. Single and sequential chemical
extraction procedures were firstly conducted. These
leaching tests were used to assess the potential mobility of
As depending on its fractionation in the contaminated soil,
and also on the portion of As sorbed onto soil particles.
Additionally numerical simulations were performed using
the USGS software PHREEQC-3 in order to evaluate the
role of adsorption on As mobilization. This multidisciplinary approach provided information on the nature of As
fixation in this mining soil. Moreover the role of adsorption
in the control of dissolved As was evidenced by geochemical modeling. Results showed that As appeared to be
mainly (ca. 72 wt%) reversibly sorbed to iron (Fe) compounds in the soil, in particular Fe oxyhydroxides. Consequently a potential risk of As mobilization exists
& Vincent Chatain
Vincent.Chatain@insa-lyon.fr
1
LGCIE—DEEP (Déchets Eau Environnement Pollutions),
EA4126, Université de Lyon, INSA Lyon,
69621 Villeurbanne Cedex, France
2
Department of Civil and Environmental Engineering,
Vanderbilt University, Station B-35 1831, Nashville,
TN 37235, USA
especially under acidic and/or reducing conditions, which
frequently occurs in mining environments.
Keywords Arsenic Mining Soil Fe oxyhydroxides
Geochemical modeling Mineral assemblage
Introduction
Arsenic (As) is a toxic metalloid naturally present in the
environment that comes from various sources such as
volcanism and weathering of the bed rock. The As concentration in natural soils typically ranges from 0.1 to
50 mgkg-1 (Baker and Chesnin 1975). Anthropogenic
activities also contribute to the geochemical cycling of As
in a variety of ways (Huang 1994; Smedley and Kinniburgh 2002; Tamaki and Frankenberger 1992). Arsenic is
used in agriculture as a component of pesticides, and in
wood preservation and glassmaking. It is also dispersed
into the environment through the burning of fossil fuels.
Additionally, due to its geochemistry, As is used as an
indicator element in geochemical prospecting for various
types of mineral ores (gold, silver, copper, uranium, etc.)
(Boyle and Jonasson 1973; Reith and McPhail 2007).
Soils near mining sites receive significant inputs of As
and heavy metals during mining operations and minerals
processing (Azcue and Nriagu 1995; Bodénan et al. 2004;
Hudson-Edwards et al. 1999; Navarro et al. 2008; Savage
et al. 2000). Arsenic concentrations in tailings piles and
tailing-contaminated soils can reach up to several thousand
mgkg-1 (Smedley and Kinniburgh 2002). However, unlike
organic contaminants, As in soils cannot be decomposed
chemically or biologically.
It is commonly acknowledged that total As concentration in the soil is not a good indicator of potential mobility
123
306 Page 2 of 9
and leaching. Chemical interactions between soil and As
are important to understand its fate in the environment and
choose the suitable management strategy (Bolan et al.
2014). These last years, part of the research concerned
remediation (Gonzalez-Fernandez et al. 2011; Drouhot
et al. 2014; Jana et al. 2012; Flakova et al. 2012). In this
context, the better understanding of the fractionation and
the potential of mobilization of arsenic is required and
helpful to improve the efficiency of these technics.
Although a wide variety of leaching tests are available
in the literature to determine the potential mobility of
pollutants (Kosson et al. 2002; van der Sloot et al. 1997),
very few have been designed to provide information on the
nature of their fixation processes in studied matrices. On
the other hand, hydrous ferric oxides (HFO) are known to
be important in the retention of inorganic arsenic in soil.
The interaction between HFO and ions can be described by
surface complexation model (SCM) (Dzombak and Morel
1990). This approach has been successfully used to study
As adsorption/desorption onto HFO in sediments, soils and
groundwater (Biswas et al. 2014; Bowell 1994; Jiang et al.
2005; Lumsdon et al. 2001). Used concurrently with geochemical characterization, the modeling has proved to be
useful to describe As behavior (Carrillo-Chávez et al. 2014;
Coussy et al. 2010; Sracek et al. 2004).
In this context, an approach was developed based on the
use of leaching tests (single and sequential chemical
extraction procedures) in conjunction with a mineralogical
study and a geochemical modeling. Indeed, knowledge of
the pollutants operational fractionation in the studied
matrix is required to better understand mechanisms regulating the leaching behavior of inorganic contaminants of
interest.
The work reported in this article focused on the determination of the nature of As distribution and association to
the soil constituents, and the impact of these characteristics
on the leaching behavior observed. The scientific objective
was to better understand the mechanisms regulating the
leaching behavior in order to better assess and model the
potential mobilization of As under the effect of environmental conditions. A further objective was to validate the
experimental methodology developed for that purpose.
The correlation between mineralogy, leachability and
modeling of As was examined in a contaminated soil collected from a French gold mining site. Mineralogical
analyses were carried out following leaching tests to
characterize the studied matrix. The numerical simulation
of batch leaching experiments was performed using
PHREEQC (version 3) (Parkhurst and Appelo 1999).
Our research was developed in order to better understand the distribution of As in a representative sample by
123
Environ Earth Sci (2016) 75:306
combining (i) experimental data obtained from single and
sequential chemical extraction procedures (leaching tests),
(ii) mineralogical characteristics determined with various
analytical tools and (iii) geochemical modeling. This
multidisciplinary approach tends to improve geochemical
assemblage definition and As comportment.
Materials and methods
Sampling and preparation of the soil
An As-contaminated soil collected from a gold mining site
in France, where mining activities and smelting processes
of gold ores took place until 2004, was used for the study
(Chatain 2004). It was already known to present high
concentrations of As (Bayard et al. 2006).
From the soil top surface (sampling depth: 5–35 cm),
four representative soil samples, of about 50 kg were collected from the site. Prior to characterizations and liquid–
solid partitioning experiments, the soil samples were airdried at room temperature (20 ± 3 °C) for 1 day, sieved
through a stainless steel 2 mm mesh sieve to remove coarse
debris and gravel, homogenized, and finally stored at 4 °C
in the dark (Chatain et al. 2005a).
Chemical and mineralogical characterization
The natural pH of the soil was measured in a 10:1 mixture
of soil to deionized water after a contact period of 48 h
(SR003.1 protocol, Kosson et al. 2002) using a WTWÒ
combined glass electrode. The redox potential (ORP) was
measured in the same soil slurry using a Radiometer analyticalÒ platinum-kalomel electrode (Pt–Ag/AgCl,
?197 mV vs. NHE).
After mineralization by acid digestion (AFNOR 1996),
total soil concentration of trace elements and major constituents was determined by Inductively Coupled Plasma
Optical Emission Spectrometry (ICP-OES, Jobin–Yvon
Ultima 2Ò).
X-ray diffraction
Three replicates of the soil fraction \2 mm were ground to
a fine powder (\50 lm) and studied with a SIEMENSÒ
D500 X-ray diffractometer (XRD) equipped with a copper
anticathode. The samples were scanned on a reflection
angle (2h) from 3° to 70° at a scan rate of 0.02° 2h/s.
Results were processed using the DIFFRACPlus EVAÓ
software (BRUKER AXSÒ) and the ICDDÒ database (International Centre for Diffraction Data).
Environ Earth Sci (2016) 75:306
Scanning electron microscopy equipped for energy
dispersive spectroscopy
Mineralogical investigation was also performed on two
replicates of the same fraction (\2 mm) by scanning
electron microscope (SEM, JEOLÒ 840A LGS) coupled
with an X-ray energy dispersive spectrometer (EDS).
Samples were prepared by pasting the soil particles on an
adhesive tape placed on a stub (sample holder of the SEM).
Sample coating with a conductive material was not
required. The backscattered electron (BSE) mode was used
with a voltage of 20 kV.
Leaching tests
A sequential chemical extraction specifically adapted by
Matera et al. (2003) to study As-bearing phases on amorphous and crystal iron was chosen. The seven following
arsenic fractions were extracted: (F1) soluble in MgCl2
(1 molL-1) at pH 7; (F2) bound to carbonates (CH3COONa
(1 molL-1)/CH3COOH pure at pH 5); (F3-Mn) bound to
Mn-oxides (NH2OHHCl (0.04 molL-1) in CH3COOH
25 % at pH 2) (F3-Fe(a)) bound to amorphous Fe oxides
((NH4)2C2O4H2O (0.2 molL-1)/H2C2O4 (0.2 molL-1) at
pH 2); (F3-Fe(c)) bound to crystalline Fe oxides ((NH4)2C2O4H2O (0.2 molL-1)/H2C2O4 (0.2 molL-1)/C6 H8O6
(0.1 molL-1) at pH 2); (F4) bound to organic matter and
sulfides (HNO3 and H2O2 30 % at pH 2; and CH3COONH4
(3.2 molL-1) in HNO3 20 %)); (F5) residual fraction (obtained by difference after total digestion).
Single extraction techniques were also performed to
verify that As adsorption on Fe oxyhydroxides was the
major process of As trapping in the soil (Clozel et al.
2002). These extractions were carried out in triplicate using
1 molL-1 sodium hydroxide (NaOH) solution, and
0.1 molL-1 dipotassium hydrogen orthophosphate
(K2HPO4) solution. A liquid/solid (L/S) ratio of 10 mLg-1
and a contact time of 48 h were used.
Batch leaching experiments were conducted following
the SR002.1 protocol (Kosson et al. 2002). Under ambient
conditions, 10 g of soil sample and 100 ml of prepared
solutions were mixed for 48 h (L/S ratio of 10 mL/g of
dried soil). HCl and NaOH solutions of varying concentrations were used as leachants in order to obtain a range of
pH varying from 1 to 13.
Leachate analyses
All leachates were filtered through 0.45 lm pore size
acetate-cellulose WhatmanÒ membranes. They were subsequently analyzed for concentrations of the constituents of
interest using ICP-OES. Sulfates and chlorides were
Page 3 of 9 306
analyzed using IC (ion chromatography, Dionex DX320
HPICÒ). Total organic carbon (TOC) in solution was
measured using a TOC analyzer (Total Organic Carbon
analyzer, Shimadzu TOC-5000AÒ).
Geochemical modeling
The PHREEQC program is based on the calculation of
equilibrium between aqueous solutions and minerals,
gases, solid solutions, exchangers, and sorption surfaces.
As suggested by Peyronnard et al. (2009), a simplified
mineral assemblage was defined on the basis of the mineralogical study (XRD and SEM analysis) and chemical
characterization. The acid attack was simulated on the
assemblage by adding nitric acid or sodium hydroxide.
Finally, quantification of minerals was optimized by
comparing simulated curves to experimental data in a
reiterated process.
Simulation was performed on a 1 L volume of liquid,
which means that the solid assemblage should represent
100 g of soil to comply with the L/S ratio of the leaching
experiment. During simulation, the equilibrium between
liquid and solid was reached and no gases were considered.
The geochemical model was developed to evaluate the
role of adsorption in As solubilization. The assemblage
which best fit buffer capacity and solution concentration
was used to compare two models: with and without considering adsorption. The adsorption phenomena is modelled with a surface complexation model: the DDL
(diffuse-double layer) (Dzombak and Morel 1990), already
incorporated in PHREEQC, was used. From the mineralogical analysis data, it was assumed that the dominant
adsorption/desorption phase was hydrous-ferric oxide
(HFO). In the DDL, the major surface complexes for
As(III) and As(V) are :FeH2AsO3 and :FeHAsO4-,
respectively. With the decrease in pH the formation of
:FeHAsO4- is favored, whereas at higher pH :FeH2AsO3 is prevalent.
The Lawrence Livermore National Library (llnl) thermodynamic database supplied with PHREEQC was used
for solubility products and dissolution reactions of mineral
phases. Surface complexation constants from Dzombak
and Morel (1990) are already implemented in the llnl
database for reactions between ferrihydrite and major
cations (Ca, S, Ba, Cd, Zn, Cu, Pb, Mg, Mn, Fe) and
anions. The database was modified by adding surface
complexation constants for arsenate and arsenite on HFO
from the minteq.v4 database. Fe(OH)3(am) phase (Hummel
et al. 2002), reactions involving H3AsO3 and their corresponding equilibrium constants were included as well.
Table 1 reports the surface complexation constants for
sorption of As on HFO.
123
306 Page 4 of 9
Environ Earth Sci (2016) 75:306
Table 1 Adsorption reactions and equilibrium constant of surface
complexation of As with HFO
Adsorption reaction
log K
:FeOH ? H3AsO4 $ :FeH2AsO4 ? H2O
8.67
:FeOH ? H3AsO4 $ :FeHAsO4- ? H2O ? H?
2.99
:FeOH ? H3AsO4 $ :FeOHAsO43- ? 3H?
-10.15
:FeOH ? H3AsO3 $ :FeH2AsO3 ? H2O
5.41
:FeOH ? H3AsO4 $ :FeAsO42- ? H2O ? 2H?
Table 2 Physicochemical
characteristics (L/S ratio:
10 mLg-1; contact time: 48 h)
and concentration of trace
elements as determined by acid
digestion of the studied soil
Table 3 Mineral composition
by XRD analysis (relative
abundance: ???? very abundant; ??? abundant;
?? present; ? traces)
-4.70
Water content (wt%)
10
pH (H2O)
6.5
ORP (mV vs. NHE)
400
Si (wt%)
Fe (wt%)
20.3
8.9
Al (wt%)
4.3
Ca (wt%)
3.7
As (mgkg-1)
-1
Cu (mgkg )
27,700
1700
Pb (mgkg-1)
800
Zn (mgkg-1)
400
Mineral
Abundance
Quartz
????
Micas
??
Feldspar
??
Chlorite
?
Kaolinite
?
Calcite
??
Dolomite
??
Gypsum
???
Jarosite
??
Hematite
?
Scorodite
?
sulfide phases in the soil could be due to the weathering
and oxidation of pyritic minerals during the post-mining
period.
Microscopy analyses performed using the BSE images
by SEM coupled with EDS microanalysis can be used to
complement the XRD analysis. The major minerals identified were quartz, gypsum, and particles containing Fe and
As, with a composition in accordance with XRD results. A
close similarity between the arsenic and iron cartography
was observed in SEM results (Fig. 1).
Results and discussion
Leaching behavior
Chemical characterization
The major physicochemical characteristics and the elemental total concentrations as determined by acidic
digestion of the studied soil are summarized in Table 2.
The soil was mainly characterized by a water content of
10, 3 wt% of As, 9 wt% of Fe and 1.9 wt% of total organic
carbon. Its natural pH and ORP measured after 48 h of
contact with deionized water at a L/S ratio of 10 mL/g,
were 6.5 and ?400 mV vs. NHE, respectively. Moreover, a
very low soluble (0.016 wt%) fraction of As was obtained
from the single extraction procedure using deionized water
as the extractant.
Mineralogical characterization
This characterization was achieved by using scanning
electron microscopy coupled with an EDS analyzer and
XRD analysis. Results from the XRD analysis (Table 3)
indicated that the soil was mainly composed of quartz,
gypsum, feldspars, micas (as muscovite), calcite, and
dolomite.
XRD analyses indicated the presence of traces of
hematite, jarosite and scorodite, which are the common
pyrite weathering products with which As can be typically
bound (Clozel et al. 2002). The absence of residual mixed
123
The results of sequential extractions are summarized in
Fig. 2. The largest portion of As (about 66 wt%) was
extracted as fraction F3-Fe(a) (bound to amorphous Fe
oxides). Globally the first six fractions represented about
73 wt% of the total As and only 27 wt% was in the residual
fraction, and then strongly bound to the soil matrix. A very
low fraction of As and Fe (i.e., \1 wt%) was extracted as
exchangeable and bound-to-carbonates fractions. These
results confirmed that As was mainly bound on soil
amorphous oxyhydroxide particles, as previously shown by
the results from XRD and SEM.
The results obtained from single extractions, using
1 molL-1 NaOH solution, and 0.1 molL-1 K2HPO4
solution, are summarized in Table 4. They are compared
with those acquired during single extraction using deionized water (natural).
Concerning NaOH extraction, about 74 wt% of the total
As was extracted within these extreme conditions of pH
(ca. 14). A mechanism of As desorption from the Fe
oxyhydroxide surface by competition with OH- ions might
explain the abrupt increase in As extractability (CarbonellBarrachina et al. 1999; Yang et al. 2002). Fe solubility
remained very low (ca. 0.2 wt% of the total Fe content). Fe
solubility might have been controlled by Fe oxyhydroxide
precipitates.
Environ Earth Sci (2016) 75:306
Page 5 of 9 306
Fig. 1 SEM-EDX mapping of the 0.2–2 mm fraction of the soil (12X)
100%
90%
F5
80%
70%
F4
60%
F3-Fe(c)
50%
40%
F3-Fe(a)
30%
20%
10%
F3 - Mn
F2
0%
F1
As
Fe
Fig. 2 Sequential chemical extractions performed on the \2 mm
fraction of the soil (conc. in wt%)
These results suggest that most of the As present in the
soil is sorbed onto Fe oxyhydroxides, which is consistent
with the mineralogical analysis and sequential extraction.
The remaining fraction (ca. 26 % of the total As content)
can be considered to represent the part co-precipitated with
Fe and/or bound to resistant compounds (i.e., silicates or
sulfides).
Table 4 Results from single
extractions (LS ratio:
10 mLg-1; contact time: 48 h)
The results of As extraction from the contaminated soil
by K2HPO4 solution are also shown in Table 4. The
addition of this solution was found to have no significant
effect on pH and ORP conditions and Fe release, compared
to those obtained with deionized water (natural conditions).
However, a drastic increase in As solubilization (as much
as 400-fold at a pH value about 7), and in phosphate
adsorption onto Fe oxyhydroxides (ca. 76 %) was evidenced. These results are consistent with studies performed
by Alam et al. (2001) and Clozel et al. (2002), confirming
that As is for the most part reversibly sorbed onto Fe
oxyhydroxides.
Extracting solution
Deionized water
1 molL
-1
NaOH solution
0.1 molL-1 K2HPO4 solution
Contaminant mobility as a function of pH
and geochemical modeling
Batch leaching tests showed a low solubility of As between
pH 3 and 7.5. Mobility increased with acid and alkaline
conditions with a maximum for pH below 1.
All minerals observed in the XRD analysis (Table 3)
were included in the model. The presence of amorphous
hydroxide was evidenced by previous studies (Bayard et al.
2006; Chatain et al. 2003, 2005b; Clozel et al. 2002) and
confirmed by the sequential extraction presented in section ‘‘Leaching behavior’’. It was therefore introduced in
the assemblage as amorphous Fe(OH)3. Arsenic was
As (conc. in wt%)
-3
Fe (conc. in wt%)
pH
ORP (mV vs. NHE)
-4
6.5
?400
73.51
-1
1.7 9 10
13.8
?60
4.89
3.40 9 10-3
7.5
?430
1.60 9 10
1.0 9 10
123
306 Page 6 of 9
Environ Earth Sci (2016) 75:306
Table 5 Reactions, equilibrium constant and initial mass of assemblage minerals
Mineral
Dissolution reaction
log Ksp
Quartz
SiO2 = SiO2
Kaolinite
Al2Si2O5(OH)4 ? 6 H? = 2 Al3? ? 2 SiO2 ? 5 H2O
K-Feldspar
KAlSi3O8 ? 4 H? = ?Al3? ? K? ? 2 H2O ? 3 SiO2
?
?
3?
Initial mass (g)
-3.9993
12.0168
6.8101
0.2582
-0.2753
4.1745
13.5858
0.0398
Muscovite
KAl3Si3O10(OH)2 ? 10 H = K ? 3 Al
Chamosite-7A
Fe2Al2SiO5(OH)4 ? 10 H? = SiO2 ? 2 Al3? ? 2 Fe2? ? 7 H2O
32.8416
0.0342
Clinochlore-7A
Mg5Al2Si3O10(OH)8 ? 16 H? = 2 Al3? ? 3 SiO2 ? 5 Mg2? ? 12 H2O
70.6124
0.2779
Cronstedtite-7A
Fe2Fe2SiO5(OH)4 ? 10 H? = SiO2 ? 2 Fe2? ? 2 Fe3? ? 7 H2O
16.2603
0.0400
Ripidolite-14A
Mg3Fe2Al2Si3O10(OH)8 ? 16 H? = 2 Al3? ? 2 Fe2? ? 3 Mg2? ? 3 SiO2 ? 12 H2O
60.9638
0.0619
Gypsum
CaSO42H2O = Ca2? ? SO42- ? 2 H2O
-4.4823
10.3320
?
?
SO42-
? 3 SiO2 ? 6 H2O
3?
Jarosite
KFe3(SO4)2(OH)6 ? 6 H = K ? 2
Calcite
Dolomite
CaCO3 ? H? = Ca2? ? HCO3CaMg(CO3)2 ?2 H? = Ca2? ? Mg2? ? 2 HCO3-
Tenorite
CuO ? 2 H? = Cu2? ? H2O
7.6560
0.0716
Hematite
Fe2O3 ? 6 H? = 2 Fe3? ? 3 H2O
0.1086
3.1940
Fe(OH)3 (amorphous)
Fe(OH)3 ? 3 H? = Fe3? ? 3 H2O
5.0000
6.4128
Pyrite
FeS2 ? H2O = ?0.25 H? ? 0.25 SO42- ? Fe2? ? 1.75 HS-
-24.6534
0.0120
Chalcopyrite
CuFeS2 ? 2 H? = Cu2? ? Fe2? ? 2 HS-
-32.5638
0.0220
Arsenopyrite
FeAsS ? 1.5 H2O ? 0.5 H? = 0.5 AsH3 ? 0.5 H2AsO3- ? Fe2? ? HS-
-14.4453
0.0163
Scorodite
FeAsO42H2O = Fe3? ? AsO43- ? 2 H2O
-20.249
2.3078
? 3 Fe
? 6 H2O
-9.3706
0.0501
1.8487
2.5135
1.0009
0.9220
Except for Fe(OH)3(am) (Hummel et al. 2002), all log Ksp values come from the Lawrence Livermore National Laboratory (llnl) database
supplied with PhreeqCÒ
introduced as arsenopyrite (present in the original minerals) and scorodite, a secondary mineral which currently
occurs in waste rocks rich in arsenopyrite and/or arsenian
pyrite as a weathering product (Nordstrom and Parks 1987;
Paktunc and Bruggeman 2010). Lastly, the model was set
so that hematite and iron(III) hydroxide could dissolve but
not precipitate. Since the time contact of the experiment
was relatively short (48 h), the equilibrium condition of
PHREEQC calculations would overestimate the precipitation of these phases. For the same reason, the assemblage
does not represent the entire soil mineralogy but the reactive fraction, bringing the total mass of the mineral
assemblage to 41.2/100 g. Table 5 presents reactions,
equilibrium constants and initial mass of minerals used to
represent the leaching behavior of the soil.
As recommended by Dzombak and Morel (1990) a
specific surface area of 600 m2g-1 was defined, while
surface density was adjusted to match batch experiments. A
density of 0.09 sitesmol-1 was chosen for strong sites and
0.5 sitesmol-1 for the weak sites. In this respect, Sracek
et al. (2004) pointed out that in many cases, the quantity of
Fe(III) oxides and hydroxides indicated by mineralogical
methods poorly matches modeled data and the amount of
adsorbent has to be adjusted accordingly.
The best fitting assemblage was used to evaluate the role
of adsorption in As release. In Fig. 3, S, Fe and As
experimental solubility (Exp.) as a function of pH was
123
compared with their solubility estimated by two models:
without (Model 1) and with (Model 2) adsorption.
The behavior of S (Fig. 3a) and Fe (Fig. 3b) was well
correlated to experimental data whether adsorption was
used or not. For As (Fig. 3c), between pH 1 and 3, the two
models identically matched experimental data. Then, when
adsorption was neglected (Model 1), soluble As increased
substantially from pH 3–5 because of the total dissolution
of scorodite (Fig. 4a) and arsenopyrite (Fig. 4b), which SI
values became negative. Consequently, As was overestimated in Model 1 for pH values greater than 3.
In PHREEQC modeling, scorodite was the only phase
containing As that could possibly precipitate at low pH. As
stated by Paktunc and Bruggeman (2010), scorodite has its
lowest solubility around pH 3. In contrast, for pH less than
2 and greater than 6, its solubility is quite high. It is
reported in the literature that As may precipitate as secondary sulfides (orpiment, arsenic trisulfite or pyrite) in a
reduced environment that causes the reduction of SO4
(Sracek et al. 2004). Given the oxidant conditions of the
leaching test in the present study, SO4 reduction and consequently secondary sulfide precipitation can be excluded.
In the experimental data, As concentration in solution
decreased until pH 6. Since scorodite precipitation alone
cannot explain As behavior for pH higher than three,
adsorption was taken into account. In fact, the model with
adsorption (Model 2) was better correlated with As
Environ Earth Sci (2016) 75:306
(a)
Page 7 of 9 306
Exp.
Model 1
(a)
Model 2
S mg/L
SI Scorodite
10000
1000
5
0
-5
-10
-15
-20
0
2
4
6
8
10
12
14
8
10
12
14
pH
100
0
2
4
6
8
10
12
14
pH
(b)
Exp.
Model 1
Model 2
10000
Fe mg/L
-140
-160
-180
-200
-220
0
2
4
6
pH
100
Fig. 4 Saturation indices of scorodite (a) and arsenopyrite (b) as a
function of pH (L/S ratio, 10 mLg-1) calculated for model with
surface complexation (Model 2)
10
1
0.1
0
2
4
6
8
10
12
14
pH
(c)
Exp.
Model 1
Model 2
1000
100
As mg/L
-120
-240
1000
0.01
SI Arsenopyrite
(b) -100
dissolved and that scorodite precipitation would remain the
governing process.
In conclusion, modeling confirmed that As release is
mainly controlled by adsorption on ferric hydroxide. High
concentrations of As were attributed to the dissolution of
iron oxyhydroxides, causing the release of arsenic initially
associated with this phase. In the absence of SCM, As
concentration in solution at neutral pH would be three
orders of magnitude higher. Modeling also revealed the
contribution of scorodite to explain As behavior at low pH
(\3).
10
Conclusions and perspectives
1
0.1
0
2
4
6
8
10
12
14
pH
Fig. 3 S (a), Fe (a) and As (a) solubility as a function of pH (L/S
ratio, 10 mLg-1). Experimental (Exp.) and modeled data: without
surface complexation (Model 1) and with surface complexation
(Model 2)
behavior, especially at neutral (6–7) and alkaline pH. For
pH lower than 3, modeling suggests that amorphous iron(III) hydroxide (Fe(OH)3(am)) may be completely
An As-contaminated soil collected from a gold mining site
in France has been studied. In order to better understand
the parameters controlling the potential release of As from
contaminated soils into surface or ground-waters, an
experimental methodology was developed that combined
the use of leaching tests (single and sequential chemical
extraction procedures) in conjunction with a mineralogical
study. This experimental procedure was applied to contaminated soil samples collected from a gold mining site.
The results obtained indicated first that although As
concentration in the considered soil was quite high, the
release of As into deionized water was very limited in
leaching tests. This was attributed primarily to the low
123
306 Page 8 of 9
solubility and the stability of the solid-bearing phases of As
in the soil under slightly acidic natural conditions.
The multidisciplinary approach (leaching and mineralogy) developed in this work provided information on the
nature of As fixation processes that may control As reactivity
in the subsurface environment. It was observed that most of
the As (72 %) appeared to be reversibly sorbed onto Fe
phases in the soil particles, in particular Fe oxyhydroxides.
These results indicated a potential risk of As mobilization
over the long term under specific leaching conditions (i.e.,
pH or ORP gradient, chelation effect, etc.) which frequently
occur in mining environments. Indeed, many factors can
significantly affect the geochemical status of contaminated
soils such as climate, fluctuating groundwater levels, periodic inundation, activity of microorganisms, vegetation, or
deterioration of the physical properties of the substrate.
The role of adsorption in the control of dissolved As was
supported by geochemical modeling. The retention of As at
neutral pH seems to be governed by HFO adsorption,
indicating that Fe oxyhydroxides dissolution is responsible
for As release.
Acknowledgments The authors are grateful to the Région RhôneAlpes and the Association RE.CO.R.D. (Waste Research Cooperative
Network, France) for financial support.
References
AFNOR (1996) Qualité des sols—Sols, sédiments—Mise en solution
totale par attaque acide (NF X31-147). AFNOR, Paris
Alam MGM, Tokunaga S, Maekawa T (2001) Extraction of arsenic in
a synthetic arsenic-contaminated soil using phosphate. Chemosphere 43(8):1035–1041. doi:10.1016/S0045-6535(00)00205-8
Azcue JM, Nriagu JO (1995) Impact of abandoned mine tailings on
the arsenic concentrations in Moira Lake, Ontario. J Geochem
Explor 52(1):81–89. doi:10.1016/0375-6742(94)00032-7
Baker DE, Chesnin L (1975) Chemical monitoring of soils for
environmental quality and animal and human health. Adv Agro
27:305–375. doi:10.1016/S0065-2113(08)70013-0
Bayard R, Chatain V, Gachet C, Troadec A, Gourdon R (2006)
Mobilisation of arsenic from a mining soil in batch slurry
experiments under bio-oxidative conditions. Water Res
40(6):1240–1248. doi:10.1016/j.watres.2006.01.025
Biswas A, Gustafsson JP, Neidhardt H, Halder D, Kundu AK,
Chatterjee D, Berner Z, Bhattacharya P (2014) Role of
competing ions in the mobilization of arsenic in groundwater
of Bengal Basin: Insight from surface complexation modeling.
Water Res 55:30–39. doi:10.1016/j.watres.2014.02.002
Bodénan F, Baranger P, Piantone P, Lassin A, Azaroual M, Gaucher
E, Braibant G (2004) Arsenic behaviour in gold-ore mill tailings,
Massif Central, France: hydrogeochemical study and investigation of in situ redox signatures. Appl Geochem
19(11):1785–1800. doi:10.1016/j.apgeochem.2004.03.012
Bolan N, Kunhikrishnan A, Thangarajan R, Kumpiene J, Park JE,
Makino T, Kirkham MB, Scheckel K (2014) Remediation of
heavy metal(loid)s contaminated soils—To mobilize or to
immobilize? J Hazard Mater 266:141–166. doi:10.1016/j.jhaz
mat.2013.12.018
123
Environ Earth Sci (2016) 75:306
Bowell R (1994) Sorption of arsenic by iron oxides and oxyhydroxides in soils. Appl Geochem 9(3):279–286. doi:10.1016/08832927(94)90038-8
Boyle RW, Jonasson IR (1973) The geochemistry of arsenic and its
use as an indicator element in geochemical prospecting.
J
Geochem
Explor
2(3):251–296.
doi:10.1016/03756742(73)90003-4
Carbonell-Barrachina A, Jugsujinda A, DeLaune RD, Patrick WH,
Burló F, Sirisukhodom S, Anurakpongsatorn P (1999) The
influence of redox chemistry and pH on chemically active forms
of arsenic in sewage sludge-amended soil. Environ Int
25(5):613–618. doi:10.1016/S0160-4120(99)00027-6
Carrillo-Chávez A, Salas-Megchún E, Levresse G, Muñoz-Torres C,
Pérez-Arvizu O, Gerke T (2014) Geochemistry and mineralogy
of mine-waste material from a ‘‘skarn-type’’ deposit in central
Mexico: modeling geochemical controls of metals in the surface
environment. J Geochem Explor 144(Part A):28–36. doi:10.
1016/j.gexplo.2014.03.017
Chatain V (2004) Characterization of arsenic and other inorganic
constituents potential mobilization from soils collected from a
gold mining site. Doctoral dissertation, INSA-Lyon, Lyon
University, France
Chatain V, Sanchez F, Bayard R, Moszkowicz P (2003) Arsenic
behavior in mining soil. J Phys-Paris IV 107(1):289–292. doi:10.
1051/jp4:20030298
Chatain V, Bayard R, Sanchez F, Moszkowicz P, Gourdon R (2005a)
Effect of indigenous bacterial activity on arsenic mobilization
under anaerobic conditions. Environ Int 31(2):221–226. doi:10.
1016/j.envint.2004.09.019
Chatain V, Sanchez F, Bayard R, Moszkowicz P, Gourdon R (2005b)
Effect of experimentally induced reducing conditions on the
mobility of arsenic from a mining soil. J Hazard Mater
122(1):119–128. doi:10.1016/j.jhazmat.2005.03.026
Clozel B, Battaglia F, Conil P, Ignatiadis I (2002) Physical, chemical,
and biological treatability of contaminated soils (Final report—
BRGM/RP-52065-FR). BRGM, Orléans
Coussy S, Benzaazoua M, Bussière B, Peyronnard O, Blanc D,
Moszkowicz P, Malchère A (2010) Stabilization/solidification of
arsenic in cemented paste backfill: geochemical modeling as a
mineralogical characterization tool. In: Proceedings of the first
International Stabilization/Solidification Technology forum,
pp. 161–170
Drouhot S, Raoul F, Crini N, Crini N, Tougard, Druart C, Rieffel D,
Lambert JC, Tête N, Giraudoux P, Scheifler R (2014) Responses
of wild small mammals to arsenic pollution at a partialy
remediated mining site in Southern France. Sci Total Environ
470–471:1012–1022
Dzombak DA, Morel FM (1990) Surface complexation modeling:
hydrous ferric oxide. Wiley, New York
Flakova R, Zenisova Z, Sracek O, Kremar D, Ondrejkova I, Chovan
M, Lalinska B, Fendekova M (2012) The behavior of arsenic and
antimony at Pezinok mining site, southwestern part of the Slovak
Republic. Environ Earth Sci 66:1043–1057
Gonzalez-Fernandez O, Queralt I, Manteca JI, Garcia G, Carvalho
ML (2011) Distribution of metals in soils and plants around
mineralized zones at Cartagena-La Unión mining district (SE,
Spain). Environ Earth Sci 63(6):1227–1237
Huang YC (1994) Arsenic distribution in soils. In: Nriagu JO (ed)
Arsenic in the environment—Part 1. Wiley, New York, pp 17–49
Hudson-Edwards KA, Schell C, Macklin MG (1999) Mineralogy and
geochemistry of alluvium contaminated by metal mining in the
Rio Tinto area, southwest Spain. Appl Geochem
14(8):1015–1030. doi:10.1016/S0883-2927(99)00008-6
Hummel W, Berner U, Curti E, Pearson F, Thoenen T (2002) Nagra/
PSI chemical thermodynamic data base 01/01. Nagra NTB
Environ Earth Sci (2016) 75:306
02-16, Nagra, Wettingen, Switzerland. http://www.nagra.ch/g3.
cms/s_page/77900/s_name/shopengl/S_NAME/shopde/lang/EN
Jana U, Chassany V, Bertrand G, Castrec-Rouelle M, Aubry E,
Boudsocq S, Laffray D, Repellin A (2012) Analysis of arsenic
and antimony distribution within plants growing at an old mine
site in Ouche (Cantal, France) and identification of species
suitable for site revegetation. J Environ Manage 110:188–193
Jiang W, Zhang S, Shan XQ et al (2005) Adsorption of arsenate on
soils. Part 2: modeling the relationship between adsorption
capacity and soil physiochemical properties using 16 Chinese
soils. Environ Pollut 138(2):285–289. doi:10.1016/j.envpol.
2005.03.008
Kosson DS, van der Sloot HA, Sanchez F, Garrabrants AC (2002) An
integrated framework for evaluating leaching in waste management and utilization of secondary materials. Environ Eng Sci
19(3):159–204. doi:10.1089/109287502760079188
Lumsdon DG, Meeussen JCL, Paterson E, Garden LM, Anderson P
(2001) Use of solid phase characterisation and chemical modelling
for assessing the behaviour of arsenic in contaminated soils. Appl
Geochem 16(6):571–581. doi:10.1016/S0883-2927(00)00063-9
Matera V, Le Hecho I, Laboudigue A, Thomas P, Tellier S, Astruc M
(2003) A methodological approach for the identification of
arsenic bearing phases in polluted soils. Environ Pollut
126(1):51–64. doi:10.1016/S0269-7491(03)00146-5
Navarro MC, Pérez-Sirvent C, Martı́nez-Sánchez MJ, Vidal J, Tovar
PJ, Bech J (2008) Abandoned mine sites as a source of
contamination by heavy metals: a case study in a semi-arid
zone. J Geochem Explor 96(2–3):183–193. doi:10.1016/j.gexplo.
2007.04.011
Nordstrom DK, Parks GA (1987) Solubility and stability of scorodite,
FeAsO42H2O: discussion. Am Mineral 72(7–8):849–851
Paktunc D, Bruggeman K (2010) Solubility of nanocrystalline
scorodite and amorphous ferric arsenate: implications for
stabilization of arsenic in mine wastes. Appl Geochem
25(5):674–683. doi:10.1016/j.apgeochem.2010.01.021
Page 9 of 9 306
Parkhurst DL, Appelo C (1999) User’s guide to PHREEQC (Version
2): a computer program for speciation, batch-reaction, onedimensional transport, and inverse geochemical calculations
Peyronnard O, Blanc D, Benzaazoua M, Moszkowicz P (2009) Study
of mineralogy and leaching behavior of stabilized/solidified
sludge using differential acid neutralization analysis: Part II: use
of numerical simulation as an aid tool for cementitious hydrates
identification. Cement Concrete Res 39(6):501–509. doi:10.
1016/j.cemconres.2009.03.012
Reith F, McPhail D (2007) Mobility and microbially mediated
mobilization of gold and arsenic in soils from two gold mines in
semi-arid and tropical Australia. Geochim Cosmochimi Ac
71(5):1183–1196. doi:10.1016/j.gca.2006.11.014
Savage KS, Tingle TN, O’Day PA, Waychunas GA, Bird DK (2000)
Arsenic speciation in pyrite and secondary weathering phases,
Mother Lode gold district, Tuolumne County, California. Appl
Geochem 15(8):1219–1244. doi:10.1016/S0883-2927(99)00115-8
Smedley P, Kinniburgh D (2002) A review of the source, behavior
and distribution of arsenic in natural waters. Appl Geochem
17(5):517–568. doi:10.1016/S0883-2927(02)00018-5
Sracek O, Bhattacharya P, Jacks G, Gustafsson JP, von Brömssen M
(2004) Behavior of arsenic and geochemical modeling of arsenic
enrichment in aqueous environments. Appl Geochem
19(2):169–180. doi:10.1016/j.apgeochem.2003.09.005
Tamaki S, Frankenberger WT Jr (1992) Environmental biochemistry
of arsenic. Rev Environ Contam T 124:79–110
van der Sloot H, Heasman L, Quevauviller P (1997) Harmonization of
leaching/extraction tests. Studies in Environmental Science.
Elsevier, Amsterdam
Yang JK, Barnett MO, Jardine PM, Basta NT, Casteel SW (2002)
Adsorption, sequestration, and bioaccessibility of As (V) in soils.
Environ Sci Technol 36(21):4562–4569. doi:10.1021/es011507s
123
(a)
-
NO
1000
Phreeąc modeling project, pt2
100
2
10
12
14
Cross Mark
pH
Environ Bar S (2016 7:26
DOI 10.100716-01-2012
ORIGINAL ARTICLE
(b)
10000
Model
-Model
1000
100
Geochemical characterization and modeling of arsenic behavior
in a highly contaminated mining soil
Sara Bisone'. Vincent Chatain'. Denise Blanc". Mathieu Gautier'.
Rémy Bayard' - Florence Sanchez?- Remy Gourdon'
Famil
10
01
12
14
pH
(c)
1000
-Model 1 —Modela
Study the above paper and model the
measured concentrations of S, Fe, and
As concentration in soil solution
reported in Fig. 3 (right).
As mot
12
14
pl
Fe
... As all as a fun of HLAS
na 10 Experimental (Explod dalk with
sure completo Mold 1) and with surface completion
Model