Chemosphere 63 (2006) 16321640

www.elsevier.com/locate/chemosphere

Washing as a remediation technology applicable in

soils heavily polluted by miningmetallurgical activities
A. Moutsatsou *, M. Gregou 1, D. Matsas 2, V. Protonotarios

Laboratory of Inorganic and Analytical Chemistry, Department of Chemical Engineering, National Technical
University of Athens, 9 Iroon Polytechniou street, 157 80 Zografou Campus, Athens, Greece
Received 12 September 2003; received in revised form 20 April 2005; accepted 17 October 2005
Available online 1 December 2005

Abstract
A permanent solution in order to remediate a heavily contaminated soil is washing it utilizing the appropriate
reagents. The study presented in this paper deals with the washing of a soil highly polluted by mining and metallurgical
activities. Pollution consists of slags, sulphur compound waste and various low grade lead condensates, all rich in heavy
metals and metalloids (Pb, As, Cu, Zn, Mn and Fe). Soil was washed with deionized water, 6, 3, 2 and 1 M HCl, 6 M
H2SO4, 6 M HNO3 and 0.1 M Na2EDTA. High extraction eciency was achieved with respect to hydrochloric acid
even at the lowest concentration; the solubility of pollutants in nitric acid was low, while sulphuric acid was proven
to be problematic with respect to lead. Mobilization of metals and metalloids under DI water eect was minimal indicating that the elements fraction that is weakly bound to soil particles is insignicant. Furthermore, a kinetic study was
undertaken, utilizing 1 M HCl and 0.1 M Na2EDTA. Results illustrated that 1 M HCl was more eective for intermediate mixing times (up to 4 h) while 0.1 M Na2EDTA was better applicable to short mixing times (<1 h) and low pH
values (pH < 7.1). 1 M HCl was proven to be very eective for the removal of pollutants from contaminated soil but the
high metal content of the resulting solution necessitates further treatment for the retention of metals.
 2005 Elsevier Ltd. All rights reserved.
Keywords: Soil washing; Heavy metals; HCl; EDTA

1. Introduction
Metals and metalloids such as Pb, As, Cu, Zn, Fe and
Mn have been recognized as an important environmen-

*
Corresponding author. Tel.: +30 210 7723263; fax: +30 210
7723158.
E-mail addresses: angst@central.ntua.gr (A. Moutsatsou),
info@ltp.ntua.gr, btproto@yahoo.com (V. Protonotarios).
1
Tel.: +30 210 7722120.
2
Tel.: +30 210 7722893.
3
Tel.: +30 210 7721907/7722893.

tal hazard, primarily in the area of public health (Kedziorek and Bourg, 1999). Intensive industrial activity has
resulted in the accumulation of high concentrations of
heavy metals and toxic elements in the soil environment
causing serious socio-economic problems (Kaliampakos,
1998; Mavropoulos and Kaliampakos, 1999). These substances, when present in soils, are not biodegradable by
natural process and remain in the ecosystem (Kedziorek
and Bourg, 1999).
Several remediation approaches may be applied to
contaminated soil remediation, namely chemical treatment, soil ushing and washing, electrokinetic methods,
bioremediation and phytoremediaton techniques and

0045-6535/$ - see front matter  2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chemosphere.2005.10.015

A. Moutsatsou et al. / Chemosphere 63 (2006) 16321640

thermal treatment (Mulligan et al., 2001). New technologies include the utilization of industrial by-products
(like y ashes and slags) presenting high adsorption
capacity with respect to metals (Moutsatsou et al.,
2002). These methods may signicantly reduce remediation costs. Nevertheless, there is always a risk of leaching due to dierent environmental conditions (e.g.
changes in pH, acid rain, etc.). One permanent solution
for soil pollution abatement could be the removal of
contaminants from the soil in order to minimize environmental and health risks. Chemical extraction represents a very promising method for that purpose.
Reagents such as Na2EDTA, hydrochloric, sulphuric
and phosphoric acid, isopropyl alcohol, have been successfully utilized for the removal of metals and metalloids from contaminated soil (Rybicka and Jedrzejczyk,
1995; Semer and Reddy, 1995; Sun et al., 2001;
Tokunaga and Toshikatsu, 2002).
Soil washing involves the separation of pollutants
from the soil matrix by solubilizing them in a washing
solution (Peters, 1999), usually implemented as an ex
situ process. This methodology has been successfully
used for the treatment of soils polluted with heavy metals, hydrocarbons and semi-volatile organic compounds,
while it is less eective for treating volatile organic compounds and pesticides (Peters, 1999). Pollutants with
low aqueous solubilities, such as heavy metals, usually
require the presence of an acid or chelating agent for
eective removal. It can be stated that there are four
ways by which metals can be mobilized in soils: (a)
Changes in the acidity, (b) changes in solution ionic
strength, (c) changes in the Redox potential and (d) formation of complexes. The eciency of metal removal
depends mainly on soil and metal characteristics (e.g.
crystalline, solubility), extractant chemistry and processing conditions (Mann, 1999; Peters, 1999).
1.1. Site background information
The Lavrion Technology and Cultural Park (LTCP)
is located at the metallurgy complex of the former
Compagnie Francaise des Mines du Laurium, adjacent to the city of Lavrion, about 55 km from the Athens
metropolis. The site is approximately 250 000 m2 and it
has 41 stone-built buildings, including smelting facilities,
separation plant, storage areas, oces, etc. The broader
area is characterized by a unique historical and industrial heritage, since the mining operations, related to
the exploitation of the silver and lead ores, started by
the year 3000 BC and ended in 1980 (Konofagos,
1980). These ores supported the Golden Age of Athens,
and thousands years later contributed to the development of Greek industry. Nevertheless, the intensive mining and metallurgical development, in and around the
site of the LTCP, has resulted in serious contamination
problems aecting the broader area of Lavrion and the

1633

local community. Contamination mainly consists of

by-products of the industrial activity deposited uncontrollably on soil, such as slags, various sulphur compounds waste, smelting waste, etc., all rich in hazardous metals and metalloids. Five years ago, National
Technical University of Athens conducted a large
research project for the environmental characterization of the broader area of LTCP (NTUA, 2000). The
site investigation included soil sampling from 15 boreholes to depths up to 8 m, a local groundwater investigation with boreholes to depths up to 40 m, and airborne
dust sampling using a PM-10 high-volume sampler
(Protonotarios et al., 2002). Approximately 200 samples, solid, water and dust, were collected and analyzed.
The current state of art is targeted in the following
topics: (a) extraction of metals from polluted soil utilizing various common and relatively low cost solvents and
chemical agents, (b) comparison of reagents performances in terms of their metal and metalloid extraction
yield and (c) for the most benign of the aforementioned
solvents (namely hydrochloric acid 1 M and Na2EDTA)
which are the most applicable under real conditions, a
kinetic study was undertaken in order to nd the optimum conditions for the ecient metals and metalloids
removal.

2. Materials and methods

2.1. Soil samples
Surface samples were collected from the topsoil (0
0.25 m) for further investigation due to the following
facts (a) it is considered as the most representative of
the nature of the pollution present in the area, since it
consists of metallurgical slags and sulphur compounds
waste, the by-products of the past industrial activity
and (b) it is exposed to physical and chemical eects,
resulting to signicant hazards with regard to human
health and the ecosystem. The samples contain extremely high metal and metalloid content (Table 1), in a
variety of compounds and oxidizing states (Table 2).
Thus, the implementation of a remediation scheme is
imperative and complicate.
In order to both clarify and quantify the eectiveness
of the dierent extraction mechanisms, apart from pollutants concentrations already determined by NTUA project, the determination of additional physicochemical
and mineralogical characteristics of the soil was necessary. In view of that, pH (ISO 6588), chloride content
(EN 196-21) and specic gravity (ASTM C642-90) of
soil were specied (Table 1). Furthermore, X-ray diraction analysis (XRD, Siemens D-500) and thermogravimetric analysis (Mettler TGA/STDA 851C) were also
conducted for purposes of identifying the structure of

1634

A. Moutsatsou et al. / Chemosphere 63 (2006) 16321640

Table 1
Chemical Analysis and additional physicochemical characteristics of soil sample
Element/parameter

Concentration

Pb (mg kg 1)a
As (mg kg 1)a
Cu (mg kg 1)a
Zn (mg kg 1)a
Mn (mg kg 1)a
Fe (mg kg 1)a
Ca (mg kg 1)a
Mg (mg kg 1)a
Na (mg kg 1)a
S (mg kg 1)a
pH
Cl (mg kg 1)
Specic gravity (g cm 3)

64 195
7540
4100
55 900
6500
223 600
61 900
3900
1030
109 100
7.0
5.0
3.31

NTUA (2000).

EDTA solution and leached under stirring for 1 h. The

concentration of each metal and metalloid in the leachate is considered as the respective bioavailable fraction.
2.2. Soil washing
Deionized water, 6 M, 3 M, 2 M, 1 M HCl, 6 M
H2SO4, 6 M HNO3, and 0.1 M Na2EDTA were mixed
with soil in solid/liquid = 30 g l 1 and the suspension
was mechanically agitated for 1 h at 150 rpm and room
temperature and then ltered with a 0.45 lm paper. The
concentrations of metals under investigation were measured in the ltrate by AAS and ICP-AES. Each measurement was replicated twice, for dierent samples
and the dierence between the two samples was less than
5%.
2.3. Kinetic study for Na2EDTA 0.1 M and HCl 1 M

Table 2
Pollutants compounds detected by XRD-analysis
Name

Chemical form

Lead oxide
Massicot
Anglesite
Galena
Lead carbonate
Lead oxide carbonate
Lead oxide carbonate
Plumbojarocite
Mimetite
Beudantite
Adamite
Aimataite
Wuestite
Pyrite
Sphalerite
Zinc oxide
Pyrolousite
Hauerite
Coronadite
Chalcopyrite
Calsite
Quartz
Anhydrite

Pb3O4
PbO
PbSO4
PbS
PbCO3
Pb3C2O7
Pb3O2CO3
PbFe6(SO4)4(OH)12
Pb(AsO4)Cl
PbFe(AsO4)(SO4)(OH)6
Zn2(AsO4)(OH)
Fe2O3
FeO
FeS2
ZnS
ZnO
MnO2
MnS2
PbMn8O16
CuFeS2
CaCO3
SiO2
CaSO4

pollutant compounds (Table 2). Metal concentrations

were determined by atomic absorption spectrometry
(Perkin Elmer AAS-3300) and inductively coupled
plasma emission spectrometry (ICP-AES Perkin Elmer
Optima 3000 Excel). Finally, in order to assess the bioavailable (phytoavailable) fraction of pollutants contained in soil samples, the EDTA leaching test
(Vangronsveld and Clijsters, 1992) was used: 10 g of soil
is mixed with 100 ml of 1 N CH3COONH4 and 0.02 M

As described above, soil samples were mixed with

0.1 M Na2EDTA and 1 M HCl in a solid/liquid =
30 g l 1 and the suspension was stirred at 150 rpm. At
1, 2, 4, 8 and 24 h of stirring, the ltrate was collected
and analyzed for Pb, As, Zn, Cu, Fe and Mn. As it
was mentioned above, measurements were replicated
twice, for dierent samples and the dierence between
the two samples was less than 5%.

3. Results and discussion

3.1. Soil samples
Factors aecting metal retention and mobility in soils
include pH, initial metal concentration, soil sorption
capacity, organic matter content, age of contamination
and the presence of dierent inorganic contaminants in
soil (Reed et al., 1996). Metal concentrations in our samples are extremely high (Table 1), posing a signicant
hazard for human health and the environment. Soil
under consideration is characterized as unusually polluted according to Interdepartmental Committee for
the Redevelopment of Contaminated Land (ICRCL)
classication of soils (ICRCL, 1987). Even for Mn and
Fe, which are not considered toxic, their high concentrations may signicantly deteriorate groundwater quality
(Mulligan et al., 2001). On the other hand, the age of
the contamination renders soilmetal and metalmetal
bounds less mobile due to the formation of numerous
surface complexes of solids. Moreover, the metal sorption capacity of soil is exceeded; conclusively, pollution
is present as discrete metalmineral phases and is not
bound to specic soil components (Peters, 1999). Soil
pH was found to be neutral and the high value of specic gravity of soil samples (Table 1), concerns the presence of numerous metallurgical phases. The thermal

A. Moutsatsou et al. / Chemosphere 63 (2006) 16321640

3.2. Soil washing

In Table 3, the percentage of metals extracted from
soil sample is illustrated. The importance of washing
with DI water is to reveal the metal fraction that is
weakly bound to the soil particles or sorbed at easily
coordinated sites in the outer surfaces of soil particles
(Mann, 1999). Soil solubility in DI water was below
50 g kg 1 and below 1% for all tested metals (Table 3).
Conclusively, the bulk of the metals under investigation
is very tightly bound and probably belongs to one or
more of the following fractions: ion exchangeable,
adsorbed, precipitated, organically bound or trapped
in an insoluble form in the soil matrix (Mann, 1999).
Lead is better extracted under 6 and 3 M HCl eect
(almost 80%), while poor results are presented for 6 M
H2SO4. The latter is probably attributable to the pres-

100
90
80

% Removal

behavior revealed a low organic content and a signicant mass reduction after 750 C, which is probably
attributable to the decomposition of CaCO3 and sulphur
compounds. The bioavailable (phytoavailable) fraction
of pollutants exceeds the respective international standards (European Environmental Agency, 1999) in the
case of zinc, arsenic, copper and cadmium and found
to be extremely high in the case of lead.
Mineralogical analysis of the soil sample (XRD
Table 2), demonstrated the presence of many dierent
lead compounds, such as galena (PbS), anglesite
(PbSO4), oxides, carbonates and complexed leadiron
and leadarsenic compounds. Zinc appears mainly as
sphalerite (ZnS) and secondarily as oxide and adamite.
Copper is detected only in complexed ferric compounds,
such as chalcopyrite, while manganese is primarily present as hauerite (MnS), oxide and in leadmanganese
compounds. The predominant iron form is that of pyrite
(FeS), while many dierent iron oxides are also present.
Finally, arsenic usually accompanies lead, iron and zinc
in dierent mineralogical phases. This variety of pollutant forms was expected due to the disposal of dierent
types of waste over the years and the continuous eect of
air and rainfall.

1635

Fe
Cu
Zn
Mn
Pb
As
Total

70
60
50
40
30

20
10
0

HCl Concentration (moles/l)

Fig. 1. Mobilization of metals as a function of HCl concentration.

ence of PbSO4 in soil waste, leading to precipitation of

the insoluble respective salt (Ksp PbSO4 = 1.82 10 8).
Lead mobilization (MPbFig. 1) increases almost logarithmically with the increase of HCl concentration
(CHCl). Arsenic is better removed by 2 M HCl, while satisfactory removal is also presented for 3 and 6 M HCl
and 6 M H2SO4. Arsenic mobilization increases when
CHCl rises from 1 to 2 M, reaching a maximum and then
decreases for further CHCl increase. The decrease of As
dissolution when CHCl > 2 M is probably an outcome
of the presence of S2 anions: the increase of CHCl and
consequently of the [H+], leads to the precipitation of
As2S3, which is an insoluble salt under strong acidic conditions (very well known method for the separation of
As from Sb and SnCotton et al., 1995). On the other
hand, very poor results are presented for 6 M HNO3 and
0.1 M Na2EDTA. The ineciency of Na2EDTA was
rather expected since it does not form stable complexes
with Arsenic (Tokunaga and Toshikatsu, 2002).
Hydrochloric acid is also proved eective in the case
of Zinc, since 6 M HCl dissolves almost 97% of Zn while
satisfactory extraction is achieved for lower HCl concentrations and 6 M H2SO4. The mobilization of Zn
remains constant for 1 6 CHCl 6 3 M and then increases
to a maximum for CHCl = 6 M. On the other hand, low
extraction is presented for 0.1 M Na2EDTA and 6 M

Table 3
Actual mass of metals mobilized under the dierent solvent eect1 h of mixing (% in parenthesis)

Fe
Cu
Zn
Mn
Pb
As
Total
PH (eq.)

mg g 1 in soil
sample

HCl 6 M

HCl 3 M

HCl 2 M

HCl 1 M

EDTA 0.1 M

H2SO4 6 M

HNO3 6 M

DI

223
4
56
6
64
8
362
7.0

160 (72)
1.8 (46)
54 (97)
5.0 (80)
53 (83)
5.8 (73)
280 (78)
0.8

142 (63)
2 (51)
36 (65)
4.5 (75)
50 (79)
6.2 (77)
241 (67)
0.9

123 (55)
1.7 (42)
38 (67)
4.2 (70)
37 (57)
7.4 (92)
211 (58)
0.9

83 (37)
1.6 (41)
38 (67)
6.1 (68)
23 (35)
5.0 (64)
157 (43)
1.1

10 (14)
1.6 (41)
21 (38)
2.5 (42)
29 (44)
1.0 (13)
65 (18)
7.1

104 (47)
1.2 (30)
44 (78)
5.6 (93)
4 (6)
6.4 (80)
165 (45)
0.8

31 (14)
0.4 (10)
25 (45)
2.5 (42)
29 (44)
0.9 (11)
89 (25)
0.9

<2 (<1)
<0.04 (<1)
<0.5 (<1)
<0.05 (<1)
<0.6 (<1)
<0.1 (<1)
<4 (<1)
6.9

A. Moutsatsou et al. / Chemosphere 63 (2006) 16321640

HNO3. In the case of Na2EDTA, there is a rst indication of a slow complexation kinetics, while, under HNO3
eect, zinc probably forms insoluble salts. The extraction percentage of Cu is lower than 51% for every solvent used, with HCl in all concentrations having
demonstrated the highest yields, while satisfactory
results are obtained for 0.1 M Na2EDTA. As shown in
Table 3, copper mobilization is not essentially aected
by CHCl due to the precipitation of CuCl2. Coppers
problematic dissolution is also attributable to the presence of the metal in the matrix of a sulphate mineral
(CuFeS2), which is sparingly soluble. Iron is also better
extracted with HCl and its mobilization increases almost
logarithmically with CHCl (Fig. 1), while, very poor
results are presented for Na2EDTA and HNO3, attributable, probably, to a slow complexation kinetics and the
strong oxidizing nature of the solvent respectively. Manganese is the only metal presenting the highest extraction
under H2SO4 eect, due to the high solubility of MnSO4
formed as H2SO4 percolates the soil. Nevertheless, the
extraction capacities of HCl, in all concentrations used,
are also signicant (>70%), while Mn mobilization is
slightly improved when CHCl increases.
Considering the mobilization of the total metal content of soil sample (MT), HCl in all concentrations and
6 M H2SO4 present the optimum results. Total metal
removal follows Fe and Pb, presenting a logarithmic
increase with regard to CHCl (Fig. 1). The latter was
rather expected, since the two metals are the most abundant in soil sample.
In consideration of the aforementioned results, it can
be stated that, due to the dierent metal concentrations
in soil, the percentage of metal extracted under the solvent eect is not always a clear indication for the extraction capacity of each solvent. Supposing that a solvent
extracts 70% of Cu and 20% of Fe; that means 50 mg
of Cu and 270 mg of Fe i.e. the percentage alone may
lead to an underestimation of the dissolution capacity
of the solvent with respect to Fe. In Table 3, the actual
mass of metals extracted is also presented.
A critical parameter, aecting the yield of pollutants
mobilization, is proven to be the nature of waste present
in soil and primarily the dierent forms of metals and
metalloids compounds. Moreover, results indicate that
strong acids do not always extract great portion of the
soil metal content. It is obvious, that metal extraction
involves complicated chemical interactions between the
solvent and the dierent metal compounds in soil. In
case of HCl, for instance, it is known that the excess
of chloride ions may form a protective layer around
solid particles, thus constricting further dissolution or
cause dissolution of the precipitate forming complex
ions.
In the case of HNO3, its ineciency of achieving high
extraction capacities may be attributable to the strong
oxidizing nature of the particular agent resulting to the

formation of insoluble metal compounds. The latter procedure involves the oxidation of the central atom of
complexed compounds (pyrometallurgical phases) leading, not to soluble nitric salts, but to insoluble complexes. Furthermore, the common ion eect, in soil
and in the solvent, may signicantly aect solvent
extraction eciency.
Although some of the strong reagents studied (e.g.
6 M HCl) may extract the bulk of the contaminants
from soil, their in situ implementation is problematic.
The extremely low pH of the washing solution, necessitates further treatment, while, the toxicity of the fumes,
rich in H2S and other toxic substances, produced as
acids percolate the soil, is high. Finally, the basic nature
of soil may be damaged, thus prohibiting any future use.
3.3. Kinetic studies
Na2EDTA is the most benign (and the most expensive) among the solvents under investigation. It appears
to be eective in mobilizing lead, zinc, copper and manganese from the specic soil sample. Due to its mild nature and its eectiveness of removing specic toxic
metals, it is important to further investigate its eect
on contaminated soil. 1 M HCl is also regarded as relatively mild acid, compared to those utilized in this study.
Its eect on soil samples presents critical advantages,
such as highest mobilization yields than those obtained
for 6 M H2SO4 and 6 M HNO3 and comparable, for
all metals, with yields obtained for the denser hydrochloric acid solutions.
For the reasons stated above, both 1 M HCl and
0.1 M Na2EDTA were further investigated as soil-washing agents. In order to comprehend the washing process
and determine the optimum conditions for contaminants
removal, a kinetic study was undertaken by washing
soil with 1 M HCl and 0.1 M Na2 EDTA (solid/
liquid = 30 g l 1) and the extraction rates for metals
were measured. Results are presented in Figs. 2 and 3.
Hydrochloric acid, in various portions and concentrations, has already been utilized by many researchers

% Removal

1636

100
90
80
70
60
50
40
30
20
10
0

Fe
Cu
Zn
Mn
Pb
As

Time (h)

Fig. 2. Mobilization of metals under 1 M HCl as a function of

time.

A. Moutsatsou et al. / Chemosphere 63 (2006) 16321640

70

% Removal

60

Fe

50

Cu

40

Zn
Mn

30

Pb

20

As

10
0
0

10 12 14 16 18 20 22 24

Time (h)

Fig. 3. Mobilization of metals under 0.1 M EDTA as a

function of time.

for purposes of extracting metals contained in soil

(Davis and Singh, 1995; Reed et al., 1996; Nael et al.,
1997; van Benschoten et al., 1997; Tampouris et al.,
2001; Tokunaga and Toshikatsu, 2002). The main mechanisms for metal removal under HCl eect are the
desorption enhanced by low pH, the dissolution of discrete metal compounds and the dissolution of specic
soil components contained metals (e.g. FeMn oxides).
Literature results showed an excellent performance
for Pb and Zn, intermediate for Cu and rather low for
As. In most studies, the yield of extraction increased
with the density of HCl solution or by importing fresh
HCl solution. Nevertheless, there are several factors that
dierentiate the case of the polluted soil coming from
Lavrion: (a) the majority of the researchers deal with
articially polluted samples. That implies a selective
adsorption of pollutants at ion-exchange positions on
soil matrix, facilitating their extraction and transportation into the liquid phase. On the contrary, real waste
is not expected to have metals in loosely bound forms
(Peters, 1999) and pollutants are probably situated in
positions that do not favor their transportation into
the liquid phase. (b) Metal formations in synthetic soil
samples are usually simple i.e. single metal compounds.
On the contrary, in real miningmetallurgical waste
there is always a complicated metal speciation (Davis
and Singh, 1995). Therefore, the interaction between
the complicated metal forms and the solvent, will probably dierentiate metal mobilization rate and (c) most of
the studies deal with a restricted number of metals (maximum 3). Thus, 1 h of mixing is usually enough for pollutants extraction, independently of their positions in
soil matrix and the selectivity of the solvent with respect
to specic elements and their compounds. Results presented below, support the aforementioned statements
for the dierentiation of Lavrion samples and render
the kinetic studies crucial for the optimization of pollutants removal.
Under 1 M HCl eect, Pb solubility decreases significantly between 1 and 2 h of mixing. The reduction is

1637

probably attributable to PbCl2 precipitation, as it has

been detected by conducted XRD analysis on the solid
residue. The latter salt is re-dissolved with further mixing because of excessive Cl , reaching equilibrium after
4 h. It should be noted that results obtained by previous
studies (Cline and Reed, 1995; van Benschoten et al.,
1997), illustrated a rapid release for lead, reaching equilibrium after approximately 1 h of mixing. In case of
Lavrion soil samples, the presence of lead as sulphide,
carbonate or in Fe-compounds, signicantly retards its
dissolution rate (van Benschoten et al., 1997).
Arsenic extraction increases constantly with mixing
time, reaching a maximum after 8 h (84%). That is a very
satisfactory result, bearing in mind that As is usually
problematic under HCl eect, even in articially polluted soils (Tampouris et al., 2001). In the case of Lavrion samples, the presence of arsenic in the matrix of
easily dissolved mineralogical and pyrometallurgical
phases (Table 2) permits its ecient mobilization under
the eect of a relatively mild acid. Zinc solubility
decreases between 1 and 2 h (similarly to Pb), increases
to a maximum at 4 h (82%) and then decreases signicantly (56% after 8 h). As in the case of lead, Zn follows
a dissolutionprecipitation process as mixing time
increases. The latter fact is conrmed by the augmented
presence of ZnCl2 in the solid residue after 8 h of mixing.
The rate of copper mobilization is similar to that of Zn,
with the dierence of being almost constant up to 2 h.
Coppers maximum solubility is 61% after approximately 4 h mixing while reducing to 46% at 8 h. Finally
Fe and Mn do not vary signicantly with respect to their
dissolution rate, uctuating between 37% and 48% and
45% and 51% correspondingly, implying a rapid equilibrium.
Considering the above results, in conjunction with
the total metal content removal, it is concluded that
the optimum time of mixing for the simultaneous extraction of the major portion of Cu, Pb, Zn and Mn can be
marked approximately at 4 h, while the maximum
removal for Fe and As is presented after 8 h. Total metal
removal is almost the same for the two intervals
(48%).
Comparing the 1 M HCl best performance after 4 h of
mixing with that of strong acids after 1 h, it is observed
that 1 M HCl achieves better Cu removal (61% vs 51%
obtained for 3 M HCl), comparable removal for As
(77% vs 91% for 2 M HCl), Zn (82% vs 97% obtained
for 3 M HCl) and Mn (80% vs 92% obtained for 6 M
H2SO4), while is still less eective for Pb (44% vs 83%
for 6 M HCl) and Fe (45% vs 72% for 6 M HCl).
Solution pH is not aected by mixing time, remaining constant at the value of 1. It is obvious that pH,
although being very important for promoting metal solubilization, does not inuence the increase or the
decrease of metal extraction with time, since it remains
almost constant for the entire period of mixing.

A. Moutsatsou et al. / Chemosphere 63 (2006) 16321640

Na2EDTA use as an extracting agent presents three

important advantages: (a) pH of the nal solution is
usually alkaline and facilitates the development of further mechanisms for metal retention (e.g. precipitation),
(b) Na2EDTA is slowly biodegraded in groundwater
and (c) Na2EDTA possesses a high complexation capacity. Nevertheless, metal complexation is dicult when
these species are not close to solid surface, requiring a
large contact-surface and large amounts of the agent
for ecient metal removal (Kedziorek and Bourg,
1999). In the case of Lavrion samples, there is an important advantage regarding the EDTA amount needed for
ecient pollutants removal: the low concentrations of
Ca, Mg and Na (compared to those of natural soils),
reduce the compexation competition between metals
and the aforementioned cations.
Studying metal extraction under 0.1 M Na2EDTA
eect (Fig. 3), Pb solubility reaches an early maximum
after 1 h (44%) and then slightly decreases uctuating
between 32% and 38%. As solubility reaches some 13%
solubility after 1 h, decreases to 9% at 2 h and then
increases constantly up to 24 h of mixing (28%). Zinc
mobilization presents little alteration with time, with
better results at 1 and 4 h (maximum solubility 38%).
Copper solubility is presented high up to 1 h (41%) but
decreases signicantly with further mixing. Manganese
extraction rate increases up to 4 h (maximum 61%)
and then slightly decreases reaching equilibrium. Finally
Fe solubility is rather low for all times, reaching equilibrium after 8 h (maximum 7%). As it is shown in Fig. 3,
considering the overall pollutants removal, the best mixing time is around 1 h.
According to the relevant literature Na2EDTA usually interacts rapidly with Cu, Pb and Zn in soils, reaching equilibrium at 1 h mixing (Semer and Reddy, 1995)
and after that period of time, no signicant alteration
in metal solubility is observed. In our case, it can be
shown that Cu, Pb and Zn present a rather rapid initial
release, reaching a maximum after 1 h. Nevertheless,
further mixing causes a reduction in the solubility of
Cu and Pb, while Mn and Fe solubility increases, especially after 4 h. Considering the solubility constants of
metalNa2EDTA complexes, Cu, Pb and Zn should be
solubilized rst followed by Mn. Further mixing causes
a pH increase, making Cu, Pb and Zn complexes less
stable. That is probably the reason for solubility reduction with respect to Cu and Pb after 1 h (for 1 < t < 4 h,
a 75% of the dissolved Cu and 21% of the dissolved Pb
returns to the insoluble state). As far as Fe is concerned,
the low solubility could be attributable to several factors
(a) iron is mostly present as Fe(II) (low stability constant), (b) exchange of Na contained in Na2EDTA with
the surface Fe(II) and Fe(III) ions is taking place, (c)
assuming the existence of Fe(III), regardless that the
respective stability constant is high, slow complexation
kinetics may be involved and (d) Fe(III) is not stable

at high pH values (Kedziorek and Bourg, 1999). The relatively low solubility of As is attributable to the fact that
AsNa2EDTA complexes are rather unstable, independently of the pH value (Tokunaga and Toshikatsu,
2002). Total metal removal uctuates between 14%
and 18%, reaching equilibrium after 4 h. It can be stated
that for t 6 1 h, the highest removal of Cu and Pb can be
achieved, while for t P 4 h the removal of Fe, Mn and
As is favored. Zn presents the same solubility (almost
38%) for both 1 and 4 h mixing.
According to the results obtained for DI water (trivial
solubility for all metals), it is concluded that soil does not
possess a weakly bound fraction of metals. Conclusively,
the rapid release of Cu, Pb and Zn under Na2EDTA
eect is not attributable to that fraction, in contradiction
of what has been indicated by many researchers (Mann,
1999; Peters, 1999; Sun et al., 2001). In other words,
Na2EDTA manages to mobilize less soluble metal and
metalloid fractions, contained in soil (carbonates, oxides,
mixed metal compounds).
pH increases constantly with mixing time in the presence of Na2EDTA, following a logarithmic function
(R2 = 0.99). Considering metal mobilization under
Na2EDTA as a function of pH (Fig. 4), Pb solubility
reaches a maximum at low pH values (7.1) and then
decreases presenting insignicant alteration up to pH =
8.9. Zinc mobilization does not vary signicantly with
pH increase, while Cu presents the highest mobilization
at pH = 7.1, but decreases dramatically with pH evolution, starting to increase when pH approaches 9. The
behavior of Cu and Pb is rather normal, since the
Na2EDTA complexes with these metals are not stable for
pH P 7. On the contrary, Zn complex with Na2EDTA
continues to be almost stable even when pH increases.
On the same basis, Mn was expected to form stable complex with Na2EDTA at low pH, but in fact, it is less soluble at low pH values and its solubility increases
constantly up to pH = 8.7. The increase in Mn solubility

100

Fe
Cu
Zn
Mn
Pb
As

90
80
70

% Removal

1638

60
50
40
30
20
10
0
7.0

7.5

8.0

8.5

9.0

pH

Fig. 4. Metals mobilization under 0.1 M EDTA as a function

pH (equilibrium) for a 24 h of mixing.

A. Moutsatsou et al. / Chemosphere 63 (2006) 16321640

90

1 M HCL

80

0.1 M EDTA

70
60
50
40
30
20
10
0
Fe

Cu

Zn

Mn

Pb

As

Total

Fig. 5. Comparison between 1 M HCl (4 h mixing) and 0.1 M

EDTA (1 h mixing) eciencies for metal removal.

may be attributed to the oxidation state in which Mn

appears (IV) as it shown in Table 2 (only Mn(II) complexes are unstable in higher pH values). Iron and arsenic
are also less soluble at low pH presenting a minimum at
pH = 7.5. For highest pH values, Fe solubility increases
up to pH = 8.9. It can be stated that removal of Cu and
Pb is favored at an almost neutral pH (7.1), while Mn, Fe
and As are better dissolved for pH P 8.1.
3.4. Comparison of 1 M HCl and 0.1 M Na2EDTA
performances for metal removal
In Fig. 5, a comparison of the optimum results
obtained after 1 M HCl and 0.1 M Na2EDTA eect on
contaminated soil is illustrated. 1 M HCl presents better
results for all metals, while the two reagents remove
almost equal portion of Pb. Na2EDTA is more eective
for low mixing times (best performance after 1 h of mixing18% of total metals removal), meaning reduced
energy costs, while HCl presents better results after 4 h
of mixing (47% total metal removal).

4. Conclusions
A highly metal-contaminated soil, containing slags
and sulphur compound waste as a result of mining and
metallurgical activities, was washed with DI water, 6, 3,
2 and 1 M HCl, 6 M H2SO4, 6 M HNO3, and 0.1 M
Na2EDTA in order to examine the removal of Cu, Pb,
As, Zn, Mn and Fe. Results for DI water proved that
the metal fraction that is weakly bound to the soil particles is minimal. From the other reagents used, high
extraction eciency was achieved with respect to hydrochloric acid even at the lowest concentration. The solubility of pollutants in nitric acid was low, while
sulphuric acid was proven to be problematic with respect
to lead. Strong acids did not always extract high percentages of individual metals, indicating that complicated

1639

chemical actions are involved in metals removal. The factor with the major eect on pollutants mobilization yield
found to be the speciation of metals in soil samples.
Moreover, a kinetic study was undertaken utilizing 1 M
HCl and 0.1 M Na2EDTA in order to establish the optimum conditions for metal removal. Results showed that
Na2EDTA is more eective at low mixing times (t 6 1 h)
and low pH values (pH 6 7.1) when 42% of Pb, 38% of
Zn, 41% of Cu and 18% of total metal content is
removed. With further mixing, the solubility of Cu and
Pb is reduced, while Mn, As and Fe become more soluble. 1 M HCl was found to be more eective after 4 h
mixing when 44% of Pb, 82% of Zn, 61% of Cu, 80%
of Mn and 47% of total metal content is removed.
Conclusively, 1 M HCl was proven more eective for
metals removal but the low pH of the nal solution
requires further treatment, increasing total remediation
costs.

References
Cline, R.S., Reed, E.B., 1995. Lead removal from soils via
bench-scale soil washing techniques. J. Environ. Eng. 121,
700705.
Cotton, A.F., Wilkinson, W.G., Gauss, L.P., 1995. Basic
Inorganic Chemistry, third ed. John Wiley and Sons Inc.,
New York, USA.
Davis, A.P., Singh, I., 1995. Washing of zinc(II) from contaminated soil column. J. Environ. Eng. 121, 174185.
EN 196-21, 1994. Methods of testing cement.
European Environmental Agency, 1999. Environment in
the European Union at the Turn of the Century. Soil
Degradation, Chapter 3.6, European Environmental
Agency, 192.
ICRCL (Interdepartmental Committee for the Redevelopment
of Contaminated Land) Guidance Note 59/83, ICRCL,
1987.
ISO 65881: 2005. Determination of pH of aqueous extracts
Part 1.
Kaliampakos, D.C., 1998. Rehabilitation of an abandoned
quarry used as uncontrolled landll. Int. J. Surf. Min.
Reclam. Environ. 12, 6165.
Kedziorek, M.A.M., Bourg, C.M.A., 1999. Solubilization of
lead and cadmium during the percolation of EDTA through
a soil polluted by smelting activities. J. Contamin. Hydrol.
40, 381392.
Konofagos, K., 1980. The Ancient Lavrion and the Greek
Technique for Silver Production. Publishing of Greece S.A.,
Athens, Greece.
Mann, J.M., 1999. Full-scale and pilot-scale soil washing. J.
Hazard. Mater. 66, 119136.
Mavropoulos, A., Kaliampakos, D., 1999. International report:
uncontrolled landll investigationa case study in Athens.
Waste Manage. Res. 2, 159164.
Moutsatsou, A., Matsas, D., Protonotarios, V., 2002. Application of a calcareous y ash on soil remediation. In: R2002
Recovery, Recycling and Re-integration Congress, Geneva,
Switzerland.

1640

A. Moutsatsou et al. / Chemosphere 63 (2006) 16321640

Mulligan, C.N., Yong, R.N., Gibbs, B.F., 2001. Remediation

technologies for metal-contaminated soils and groundwater:
an evaluation. Eng. Geol. 69, 193207.
Nael, C.N., Bricka, M.R., Chao, A.C., 1997. Evaluating acids
and chelating agents for removing heavy metals from
contaminated soils. Environ. Prog. 16, 274280.
N.T.U.A., 2000. Remediation of Lavrion Technology Park.
Research Project Report. National Technical University of
Athens, Greece.
Peters, W.R., 1999. Chelant extraction of heavy metals from
contaminated soil. J. Hazard. Mater. 66, 151210.
Protonotarios, V., Petsas, N., Moutsatsou, A., 2002. Levels and
composition of atmospheric particulates derived from
surface depositions contaminated by heavy metals and from
adjacent industrial plants. Air Pollut. Waste Manage.
Assoc. 52, 174185.
Reed, B.E., Carriere, P.C., Moore, R., 1996. Flushing of a
Pb(II) contaminated soil using HCl, EDTA and CaCl2. J.
Environ. Eng. 121, 4850.
Rybicka, E.H., Jedrzejczyk, B., 1995. Preliminary studies on
mobilization of copper and lead from contaminated soils
and readsorption on competing sorbents. Appl. Clay Sci. 10,
259268.

Semer, R., Reddy, R.K., 1995. Evaluation of soil washing

process to remove mixed contaminants from a sandy loam.
J. Hazard. Mater. 45, 4557.
Sun, B., Zhao, J.F., Lombi, E., McGrath, P.S., 2001. Leaching
of heavy metals from contaminated soil using Na2EDTA.
Environ. Pollut. 113, 111120.
Tampouris, S., Papasiopi, N., Paspaliaris, I., 2001. Removal of
contaminant metals from ne-grained soils, using agglomeration, chloride solutions and pile leaching techniques. J.
Hazard. Mater. B84, 297319.
Tokunaga, S., Toshikatsu, H., 2002. Acid washing and stabilization of an articial arsenic-contaminated soil. Chemosphere 46, 3138.
van Benschoten, J.E., Matsumoto, R.M., Young, H.W., 1997.
Evaluation and analysis of soil washing for seven lead
contaminated soils. J. Environ. Eng. 123, 217224.
Vangronsveld, J., Clijsters, H., 1992. A biological test system
for the evaluation of metal phytotoxicity and immobilization by additives in metal contaminated soils. In: Merian,
E., Haerdi, W. (Eds.), Metal Compounds in the Environment and Life 4. Science Reviews Inc, USA, pp. 117
125.