Environ. Sci. Technol. 2004, 38, 937-944
Extraction of Heavy Metals from
Soils Using Biodegradable Chelating
Agents
SUSAN TANDY, KARIN BOSSART,
ROLAND MUELLER, JENS RITSCHEL,
LUKAS HAUSER, RAINER SCHULIN, AND
BERND NOWACK*
Institute of Terrestrial Ecology (ITÖ),
Swiss Federal Institute of Technology Zürich (ETH),
Grabenstrasse 3, CH-8952 Schlieren, Switzerland
Metal pollution of soils is widespread across the globe,
and the clean up of these soils is a difficult task. One possible
remediation technique is ex-situ soil washing using
chelating agents. Ethylenediaminetetraacetic acid (EDTA)
is a very effective chelating agent for this purpose but has
the disadvantage that it is quite persistent in the
environment due to its low biodegradability. The aim of
our work was to investigate the biodegradable chelating
agents [S,S]-ethylenediaminedisuccinic acid (EDDS),
iminodisuccinic acid (IDSA), methylglycine diacetic acid
(MGDA), and nitrilotriacetic acid (NTA) as potential alternatives
and compare them with EDTA for effectiveness. Kinetic
experiments showed for all metals and soils that 24 h was
the optimum extraction time. Longer times only gave
minor additional benefits for heavy metal extraction but
an unwanted increase in iron mobilization. For Cu at pH 7,
the order of the extraction efficiency for equimolar
ratios of chelating agent to metal was EDDS > NTA>
IDSA > MGDA > EDTA and for Zn it was NTA > EDDS
> EDTA >MGDA > IDSA. The comparatively low efficiency
of EDTA resulted from competition between the heavy
metals and co-extracted Ca. For Pb the order of extraction
was EDTA > NTA >EDDS due to the much stronger
complexation of Pb by EDTA compared to EDDS. At higher
concentration of complexing agent, less difference
between the agents was found and less pH dependence.
There was an increase in heavy metal extraction with
decreasing pH, but this was offset by an increase in Ca
and Fe extraction. In sequential extractions EDDS extracted
metals almost exclusively from the exchangeable, mobile,
and Mn-oxide fractions. We conclude that the extraction
with EDDS at pH 7 showed the best compromise between
extraction efficiency for Cu, Zn, and Pb and loss of Ca
and Fe from the soil.
Introduction
Metal pollution of soils is widespread across the globe, and
the clean up of these soils is a difficult task. Various in-situ
and ex-situ remediation techniques have been employed,
e.g., solidification, stabilization, flotation, soil washing,
electroremediation, bioleaching, and phytoremediation (1).
Soil washing includes the physical separation of the clay and
* Corresponding author phone: +41 1 633 61 60; fax: +41 1 633
11 23; e-mail: [email protected].
10.1021/es0348750 CCC: $27.50
Published on Web 12/23/2003
 2004 American Chemical Society
silt fraction containing the majority of the metals due to
their high specific adsorption capacity as well as the extraction
of metals by mineral acids or chelating agents. A particularly
promising technique is ex-situ soil washing with chelating
agents (2). The soil is removed from the site, treated in a
closed reactor with the chelating agent, and returned to the
site after separation of the extraction solution that now
contains the extracted heavy metals. The advantage of the
method is the high potential extraction efficiency and the
specificity for heavy metals. To keep treatment costs low, it
is necessary to achieve a cleanup so that the soil can be
reused and it should be possible to recover and reuse the
chelating agent for further extraction cycles.
There are many factors to consider when comparing
studies of chelating agent assisted soil washing in order to
decide whether the chelating agent is suitable for field-scale
decontamination of polluted sites. The ratio of chelating agent
to toxic metals, pH, quantity of major cations extracted, and
source of contamination (artificial or anthropogenic) are the
most important ones. Most studies of chelant-assisted soil
washing have found that a ratio >1 between chelant and
toxic metal is required to give good toxic metal extraction
(3-7). When the ratio is increased, so does the extracted
fraction of metal until the extraction efficiency levels off (47). Many studies however do not give the chelant:metal ratio
used. In some cases this ratio can be calculated from the
concentration and volume of chelating agent and the mass
of soil (8-14), but in others even this information is missing.
One reason an excess of chelating agent is needed is that
major cations in the soil such as Mn, Mg, Fe, and Ca along
with toxic metals present in smaller amounts compete with
the metals being studied for the chelating agent and are
extracted too (7, 14, 15). This is important from the point of
view both that it reduces the extraction of the target metals
and that it extracts major cations which are important as
plant nutrients and for maintaining soil structure in future
reuse of the decontaminated soil.
Another crucial factor to be considered in comparing
studies on chelating agent extraction is the pH of the
extraction solution. While extraction was investigated at
various pH values in some studies (3, 5, 7, 14), some only
stated the pH of the solution (10, 11, 13), while others did
not consider pH at all (6, 8, 9, 12). In general, the lower the
pH of the chelating agent solution, the greater the extraction
efficiency of the toxic metals.
The source of the metal and its form can also affect the
extraction efficiency, especially for Pb(16). Furthermore, it
has been found that significantly larger extraction efficiencies
are obtained when chelating agents are applied to artificially
contaminated soils than to soils with field contamination (7,
17). Added soluble metals quickly transfer to the exchangeable
fraction and more slowly but still within the first hour to the
carbonate and organic matter fractions (18). Inclusion in
precipitates and diffusion into micropores takes place at
much slower rates (19). In recently contaminated soils the
metals are generally in a much more accessible form than
in soils that have been contaminated years ago. This
important factor needs to be considered when chelating
agents are evaluated for field use.
Many studies have compared EDTA to other chelating
agents, acids, and surfactants and found it better suited than
or equal to its competitors for the extraction of toxic metals
from soils (3-5, 9, 11). Between 45% and 100% Pb, 54% and
100% Zn, and 47% and 98% Cu were extracted from various
contaminated soils by EDTA (3-11, 14). The maximum
extraction efficiencies of other chelating agents for Pb were
VOL. 38, NO. 3, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
937
TABLE 1. Soil Properties
soil
pH
Dornach 1 7.0
Dornach 2 6.2
Rafz
5.5
sum of
Ca
Zn
Cu
Pb
Ni
heavy metals
-1
-1
-1
-1
-1
CaCO3 (%) Corg (%) clay (%) silt (%) sand (%) (g kg ) (mg kg ) (mg kg ) (mg kg ) (mg kg )
(µmol g-1)
13.8
0.6
0.5
5.7
7.3
3.4
33
29
16
39
56
30
28
15
54
found to be the following: 63% for ADA (N-2-(acetamido)iminodiacetic acid), 60% for NTA, 58% for PDA (pyridine2,6-dicarboxylic acid), 11% for DPTA (diethylenetriamine
pentaacetic acid), and 0% for citric acid (3-5, 9). For Zn
maximum extraction efficiencies were reported as 41% with
DPTA and 12% with citric acid (8, 9, 14), and for Cu they were
32% with DPTA and 11% with citric acid (9, 12, 14).
Thus far, EDTA has been the most widely used chelating
agent for these processes due to its high extraction efficiency.
It is very effective in mobilizing metals, but unfortunately,
due to its low biodegradability (20), it is also very persistent
in the environment. This can cause a rather high risk of metal
leaching to the groundwater (21). NTA, which is easily
biodegradable, is under scrutiny due to possible adverse
health effects (22). Recently the easily biodegradable chelating
agent SS-EDDS (S,S-ethylenediaminedisuccinic acid) (23) has
been proposed as a safe and environmentally benign
replacement for EDTA in soil washing (24) and for chelantenhanced phytoremediation (25).
The aim of this study was to investigate SS-EDDS and
other chelating agents that can be potentially used in soil
washing and that are less persistent in the environment than
EDTA and to compare them with commonly used chelating
agents for effectiveness. Extraction experiments were carried
out under controlled conditions (pH and chelant: toxic metal
ratio) using soil from contaminated sites requiring cleanup
or other treatment according to Swiss legislation.
Materials and Methods
Soils. The soil materials used in this study were taken from
two polluted sites in Switzerland. Two soils were taken from
an area in Dornach which has been heavily contaminated
with Cu, Zn, and Cd for about a century by particulate
emissions from an adjacent brass smelter (26). The two soils
differed in their carbonate content: one was a Calcaric
Regosol (Dornach 1) and the other a non-calcareous Regosol
(Dornach 2) (27). Both soils had very similar total metal
contents. The third soil, (Rafz) a Haplic Luvisol, originated
from an agricultural field in Northern Switzerland, which
had been contaminated with Zn, Pb, and Cd due to sewage
sludge applications (27).
The soil samples were taken from the top 20 cm, dried at
40 °C, and sieved to <2 mm. Soil properties are displayed in
Table 1. The two Dornach soils contained almost the same
Cu and Zn concentrations but differed in pH according to
their difference in carbonate content. The molar sum of the
heavy metals Zn, Cu, Pb, Cd, and Ni in all three soils was
between 18.6 and 20.1 mmol kg-1. The total metal content
of the three soils was therefore almost the same.
Chelating Agents. EDTA was used as a reference compound to which the other chelating agents were compared.
It was used as the Na2-EDTA salt (Merck). NTA was used as
Na3NTA (Fluka). S,S-Ethylenediaminedisuccinic acid (S,SEDDS, in this article referred to as EDDS), was obtained from
Procter & Gamble. EDDS is used as a replacement for EDTA
in laundry detergents (23). The S,S-isomer is easily biodegradable, while the R,R- and S,R-isomers are not (28). We
therefore used the pure S,S′-enantiomer. Iminodisuccinic
acid (IDSA) was obtained from Bayer, Leverkusen, Germany.
Manufactured in a green process avoiding toxic chemicals
938
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 3, 2004
62
9.9
10.6
653
654
983
515
473
76
57
74
723
56
46
23
19.3
18.6
20.1
TABLE 2. Abbreviation, Structures, and log K Values for
Metal Complexation of the Used Chelating Agentsa
a log K values are for 0.1 M ionic strength from ref 34.
BASF, personal communication. c From ref 30.
b
A. Mitschker,
and traded under the name Baypure CX 100, it is used in
detergents (29). Methylglycinediacetic acid (MGDA) was
obtained from BASF, Ludwigshafen, Germany. MGDA, which
is traded under the name Trilon M, is easily biodegradable
(30). The structures of the chelating agents used in this study
are shown in Table 2.
Metal Extractions. Chelating agents were applied in two
concentrations: 0.4 and 4 mM. Metal extractions were carried
out in batch experiments at a solid:solution ratio of 1:50.
Thus, two ratios between chelant and the sum of the total
concentrations of Cu, Zn, Pb, and Ni in the soil were
achieved: equimolar (20 µmol g-1 of soil) and 10 times of
that (10:1) (200 µmol g-1 soil). For kinetic experiments 8 g
of soil was suspended in 400 mL of 0.01 M NaNO3 as a
background electrolyte. In the case of the non-calcareous
soils (soil Dornach 2 and Rafz), the pH was adjusted with 1
M HNO3 or 1 M NaOH 2 days before the addition of chelating
agent. With the calcareous soil Dornach 1, the experiments
were carried out at the pH that was established in suspension
(pH 7.5-8.0) without any adjustment. The suspensions were
shaken at room temperature and sampled after various time
intervals. Samples used for metal analysis were centrifuged
for 15 min at 3000 rpm and passed through 0.45 µm filters
(Sartorius). Samples used for iron analysis were additionally
filtered through 0.05 µm filters. The pH was measured in
suspension and readjusted if necessary. To inhibit microbial
degradation, the biocide sodium azide (NaN3) was added in
some experiments (concentration 1 g L-1).
Variation of pH was only studied with the noncalcareous
soils. The experiments were carried out in the same way as
described above, except that for each batch only 0.8 g of soil
TABLE 3. Chemical Interpretation and Operational Definition of the Fractions within the Sequential Extraction Scheme
(for 2 g of soil and 50 mL of solution) (31, 32)
fraction
chemical interpretation
extraction conditions
duration
F1
F2
F3
F4
F5
F6
F7
mobile
easily mobilizable
bound to Mn oxides
bound to organic matter
bound to amorphous Fe oxides
bound to crystalline Fe oxides
residual fraction
1 M NH4NO3
1 M NH4-acetate (pH 6)
0.1 M NH2OH-HCl + 1 M NH4-acetate (pH 6)
0.025 M NH4EDTA (pH 4.6)
0.2 M NH4-oxalate (pH 3.25)
0.1 M ascorbic acid + 0.2 M NH4-oxalate (pH 3.25)
X-ray fluorescence analysis
24 h
24 h
30 min
90 min
4 h in the dark
30 min at 96 °C
was suspended in 40 mL of solution and the reaction time
was fixed at 24 h. Chelating agents were added at two
concentrations: 0.4 and 4 mM.
Subsamples of 2 g of soil from the treatments with 20
µmol g-1 EDDS for 24 h at pH 7 were also analyzed in duplicate
by sequential extraction following the scheme of Zeien and
Brümmer before and after the treatment (31, 32). The
operational definition of the fractions obtained with this
scheme and the compositions of the extraction solutions are
given in Table 3. All suspensions were centrifuged at 2500
rpm for 10 min and passed through Schleicher & Schüll 790
1/2 paper filters before metal analysis.
To evaluate the reproducibility of the results, triplicate
extractions with and without EDDS have been performed
with all three soils at pH 4 and 7. The average standard
deviation for Cu was (4.9%, for Pb (3.2%, for Zn (5.4% for
the EDDS extractions and (8.3% without EDDS, for Fe (4.0%,
and for Ca (9.9%. We can therefore assign an average error
of about 5% to all data except for Ca, where the error is 10%.
Analytical Methods. The carbonate content was measured
by volumetric analysis of the evolved CO2 after addition of
HCl to the soil. The pH was measured at a solid:water ratio
of 1:2.5 in 0.01 M CaCl2; the organic matter content was
determined gravimetrically following oxidation with H2O2,
the grain size fractions by the pipetting method, and the
total metal contents of the soil by X-ray fluorescence (Bruker
D4 Endeavor/Germany). The pH of the metal extraction
solutions was measured with pH electrodes in the suspension.
The dissolved Cu, Zn, Pb, Fe, Mn, Mg, and Ca concentrations
in the extraction solutions were measured with a flame-AAS
(VarianAA400).
Results and Discussion
Effect of pH on Extraction. In the absence of chelating agents,
less than 20% Cu and Pb and about 40% of Zn were extracted
at pH 3 (Figures 1-3). At pH 7 the extraction of Cu, Zn, and
Pb amounted to less than 1% of their total concentration.
Already at an equimolar ratio between chelant and metal,
Cu extraction was greatly enhanced by all tested chelating
agents. At pH 4 the order of the extraction efficiency was
EDDS > EDTA > MGDA > NTA > IDSA, and at pH 7 it was
EDDS > NTA> IDSA > MGDA > EDTA. At pH 4 EDDS and
IDSA did not enhance Zn extraction compared to the controls
without chelating agent, due to their very weak complexes
at this pH, but EDTA, NTA, and MGDA were able to mobilize
Zn. At pH 7 the order of extraction efficiency was NTA >
EDDS> EDTA > MGDA >IDSA. Pb was analyzed only for
treatments of Rafz soil with EDTA, NTA, and EDDS. At both
pH 4 and 7 the extraction efficiency was EDTA > NTA >
EDDS, although at pH 7 the extraction efficiency of EDDS
was much closer to that of NTA and EDTA than at pH 4
(Figure 3).
At a chelant:metal ratio of 10, the pH dependence of the
extraction and the differences between the compounds was
much less pronounced (Figures 1-3). At this concentration
EDTA was the most effective compound for Cu, Zn, and Pb
over the whole pH range studied.
FIGURE 1. Extraction of Cu from the non-calcareous soil Dornach
2 as a function of pH with 0.4 mM (ratio chelant:metal 1) and 4 mM
chelant (ratio chelant:metal 10). Conditions: 20 g L-1 soil, 0.01 M
NaNO3.
At the low chelant:metal ratio, neither EDDS or EDTA
affected the Ca concentration in solution compared to the
treatment without chelant (Figure 4). The Ca concentration
was determined by ion exchange in the 0.01 M NaNO3
medium. At the high chelant:metal ratio of 10 significantly
more Ca was extracted than in the absence of chelant. About
20% of the total Ca in the soil was extracted by EDTA and
EDDS at pH 7 at the high ratio compared to about 10% at
the low ratio.
The Fe concentration was very low in the absence of
chelant. Both EDTA and EDDS mobilized significant concentrations of Fe (Figure 4). The extracted iron was in the
form of an Fe(III) complex because any extracted Fe(II)
complex would have been immediately oxidized to the Fe(III)
complex by oxygen (33). Whereas the EDTA extractable Fe
dropped to very low levels at pH 7 and above, EDDS was still
able to extract relatively high concentrations of Fe. At the
high chelant:metal ratio EDTA extracted about 2% of the
iron in the soil at pH 7 and EDDS only 0.25%. Fe extraction
is important for chelate speciation, although it affected only
a small fraction of the total iron in the soil.
At pH 4 the order of Cu extraction roughly followed the
decrease in the stability constants of the Cu complex but not
at pH 7 (Table 2). The order of extraction efficiency of Cu and
Zn at pH 7 was more closely related to the stability constant
of the Ca complex. To understand this behavior, we
performed speciation calculations. Reactions considered
were complexation with Ca, Mg, Fe, Mn, Zn, and Cu. The
VOL. 38, NO. 3, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
939
FIGURE 2. Extraction of Zn from soil Dornach 2 as a function of pH
with 0.4 mM (ratio chelant:metal 1) and 4 mM chelant (ratio chelant:
metal 10). Conditions: 20 g L-1 soil, 0.01 M NaNO3.
FIGURE 4. Concentration of Ca and Fe in the extraction solution
from soil Dornach 2 with 0.4 (ratio chelant:metal 1) and 4 mM chelant.
Conditions: 20 g L-1 soil, 0.01 M NaNO3.
FIGURE 3. Extraction of Pb from soil Rafz as a function of pH with
0.4 mM (ratio chelant:metal 1) and 4 mM EDTA, EDDS, and NTA
(ratio chelant:metal 10). Conditions: 20 g L-1 soil, 0.01 M NaNO3.
FIGURE 5. Calculated speciation of EDTA and EDDS in the extraction
solution from soil Dornach 2 at a chelant:metal ratio of 1.
results are presented in Figure 5. Two important differences
are obvious: whereas CaEDTA resulted as the major species
at pH 8, CaEDDS was found to be only a marginal species
at that pH. Competition between heavy metals and Ca thus
appears to be an important factor for extraction with EDTA
but not with EDDS, and this results in a decrease in extraction
efficiency for EDTA compared to EDDS. A very small fraction
of EDTA and about 20-35% of EDDS were predicted to be
in the free, uncomplexed form. Part of this fraction could
also be complexed to metals not included in the calculations
(e.g., Ni or Pb). The Fe complex is the major species for both
EDTA and EDDS at low pH. Due to the Ca competition,
FeEDTA decreases with increasing pH while FeEDDS is also
a relevant species at neutral pH. These results contradict the
statement of Vandevivere et al. (24) that Fe may be neglected
when speciating EDDS in soil suspension. These authors
based their statement on a comparison of the log K values
for heavy metals and Fe(III) but did not provide measurements of dissolved Fe. Our results show that Ca and Fe have
to be taken into account in chelant-assisted extraction of
heavy metals from soils. The competition between the two
metals and the target pollutant metals for the available
chelating agents is particular important at the low chelant:
metal ratio of 1.
The extraction of Pb at the low chelant:metal ratio seems
to depend mainly on the stability constants of the Pb
complexes (log K EDTA 17.9, EDDS 12.7, and NTA 11.3) apart
from Ca competition in the case of EDTA at high pH (Figure
5). At low concentrations of chelating agent, Pb extraction
showed a very strong dependence on pH (Figure 3). EDTA
shows very strong extraction of Pb up to pH 6 due to its high
log K value; after this however the extraction efficiency
940
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 3, 2004
reduces by about 50% due to competition for EDTA from Ca.
Although NTA has a lower log K value than EDDS, it extracts
much greater amounts of Pb than EDDS at pHs below pH
7. This is due to the much higher conditional log K value of
NTA at low pH values, which is due to the much smaller
second pKa value of NTA (2.48) compared to EDDS (6.84)
(34). NTA also suffers due to Ca competition at pHs above
7, although to a much smaller extent than EDTA, showing
a reduction in extraction efficiency. Thus, at pH values below
7, EDDS was not suitable for Pb extraction, but above pH 7,
it performed as well as EDTA and better than NTA due to the
above-mentioned factors.
At a chelant:metal ratio of 10, most of the EDTA and EDDS
were present in uncomplexed form according to the speciation calculations. About 25% of EDTA was present as
CaEDTA. Uncomplexed EDDS always accounted for about
90% of total EDDS. Thus, there was always enough free EDDS
to extract further metals. For EDTA, this was only the case
at the high ratio. However, metals are not only extracted by
the free ligand but also by metal complexes. A metal in a
chelate may not be at equilibrium and can be exchanged
with another metal. CuEDTA, for example, was able to extract
Zn efficiently from a river sediment, although the rate was
only 20% of that of CaEDTA (35).
Our results suggest that EDDS is a better metal extractant
for Cu and Zn than EDTA at pH values above 6 at low chelant:
metal ratios because it forms only a weak Ca complex. In
addition, it has the advantage that it is readily biodegradable
as classified by the modified Sturm test OECD 301B (36). Any
residual EDDS that remains in the soil after extraction will
rapidly be degraded and poses little risk with respect to
leaching of metals to the groundwater. Chelates are only
weakly adsorbed at neutral pH (37, 38) and are therefore
easily leached. NTA extracted a similar amount of Zn as EDDS
at neutral pH but significantly less Cu. IDSA and MGDA were
less efficient at low concentrations but performed more or
less as well as EDDS at high chelant:metal ratios.
We conclude that the extraction with EDDS at pH 7 gave
the best compromise between extraction efficiency for Cu,
Zn, and Pb and loss of Ca and Fe from the soil. To extract
polluting metals from soil with near-neutral pH, no acidification is necessary and unwanted extraction of major ions
is minimal.
Extraction Kinetics. The results presented so far refer to
24 h extraction time. Figure 6 shows the kinetics of Cu, Zn,
and Pb extraction from the three soils using EDDS at pH 7
up to a reaction time of 48 h. Zn and Pb extraction exhibited
a fast initial step followed by a much slower release of metals.
Extraction of Cu was slower than that of Pb or Zn. The
extraction kinetics of MGDA and EDTA were comparable to
EDDS; only the extracted amounts were smaller (data not
shown). EDDS extracted different amounts from the three
soils, a fact that will be discussed later in more detail.
IDSA first increased the dissolved Zn, but after 30 h the
concentrations decreased again (Figure 7). The same trend
could be seen for Cu. The hypothesis that this was caused
by microbial degradation of the ligand and subsequent
immobilization of the released metals was tested by conducting experiments in the presence of the biocide sodium
azide. Indeed, no re-immobilization of solubilized Cu and
Zn was observed with IDSA in the presence of the biocide.
For the efficient use of IDSA and any other rapidly biodegradable chelating agent, therefore, the extraction time must
be optimized in order to gain the maximum extraction before
the start of biodegradation.
At pH 4.5, Zn extraction by EDDS also showed a very
rapid initial increase and then a gradual decrease in dissolved
Zn concentration with time (Figure 7). This decrease, which
was not observed at pH 7, also occurred in the presence of
sodium azide. It cannot be explained, therefore, by biodeg-
FIGURE 6. Extraction kinetics of Cu, Zn, and Pb by EDDS from the
three soils at pH 7. Conditions: chelant:metal ratio 1, 20 g L-1 soil,
0.01 M NaNO3. The “no chelant” data are for soil Dornach 1 (Cu and
Zn) and for soil Rafz (Pb).
FIGURE 7. Extraction of Zn from soil Dornach 2 with EDDS at pH
4.5 and with IDSA at pH 7 in the absence and presence of the
biocide sodium azide. Conditions: chelant:metal ratio 1, 20 g L-1
soil, 0.01 M NaNO3.
radation. We suspect that it was caused by the dissolution
of iron oxides by ZnEDDS. Metal-ligand complexes are able
to dissolve oxides in a similar manner as the free ligands,
although at a slower rate (39). The mobilization of Zn by free
EDDS at pH 4 is very rapid, showing an exponential increase
before the first sampling at 1 h (Figure 7). Subsequently, the
newly formed ZnEDDS slowly dissolves Fe oxides. Equilibrium calculations show that about 70% of the EDDS in
solution could have been present as Fe complex if the system
would have been at equilibrium with hydrous ferric oxide.
VOL. 38, NO. 3, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
941
TABLE 4. Percentages of Heavy Metals Extracted from
Contaminated Soils by Applying Chelating Agents for 24 h at
pH 7 at Various Ratios between Chelant and Metals
FIGURE 8. Extraction of Fe from Dornach soil 2 as a function of time
at pH 4.5 and 7 with EDDS and EDTA. Conditions: chelant:metal
ratio 1, 20 g L-1 soil, 0.01 M NaNO3.
The observed FeEDDS concentration (as a percentage of the
total chelating agent concentration) was 28% after 8 h and
36% after 24 h, only gradually increasing up to 48 h (Figure
8). Dissolution of iron oxides either by the free ligand or
metal complexes therefore continued during the whole period
of the experiment.
The kinetic experiments showed for all metals and soils
that 24 h was the optimum extraction time with the time
period from 24 to 48 h only giving minor additional benefits.
Comparison to Other Studies. The factor of primary
importance in chelant-assisted metal extraction is the ratio
of chelant to metal. The higher the ratio, the more uncomplexed ligand is present in the extraction solution and the
faster and more complete is the extraction. This ratio is used
in only a few investigations to guide the design of experiments.
The concentration of the applied extractant solution, which
is usually specified, is of little use if the metal concentrations
of the soils and the solid:solution ratio are not known.
To compare our results with those of other authors, we
compiled data obtained under the following conditions: The
pollution occurred in the field and not by fresh addition of
metals in the laboratory, the pH during extraction was around
7, the extraction time was 24 h, and the ratio chelant:metal
was at least 1. Table 4 summarizes the results of this
compilation. Extraction of Zn spans a range from 17% to
63%, with most data ranging between 30% and 35%. Thus,
Zn extraction was quite weak for most soils, and an increased
chelant:metal ratio did not result in a significant increase in
extraction efficiency. Cu extraction varied between 28% and
100%, with most data ranging between 40% and 80%, and
Pb between 23% and 96%, with most data ranging between
30% and 89%. It is obvious that a higher ratio resulted in a
more effective extraction of these two metals and that without
consideration of this parameter, comparison of data from
different studies has little meaning.
The extraction efficiencies obtained with EDDS in this
study for Cu are among the best observed so far at a low
chelating agent:metal ratio. Zn extraction at pH 7 was about
the same as that observed for EDTA in our study and
comparable to other efficiencies reported in the literature
for low chelating agent:metal ratios. Pb extraction by EDDS
was lower than that for EDTA in most cases, but at the high
ratio it exceeded some extraction efficiencies given in the
literature for high EDTA concentrations. The use of a low
ratio of EDDS at neutral pH is therefore very promising, and
we can expect results comparable to or better than other
yields reported in the literature. This is especially good
considering that SS-EDDS was found to be readily biodegradable (40) and EDTA was not (20).
Influence of Solid-Phase Speciation on Extraction Yield.
Sequential extractions can give the information needed to
explain different extraction efficiencies for different metals.
942
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 3, 2004
chelant
ratio
Cu
Zn
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
1.0
1.0
1.0
1.0
1.2
1.8
1.8
2
3.6
6
8.9
9
10
10
20
33
50
89
300
1000
29
36
17
32
46
NTA
NTA
NTA
NTA
NTA
NTA
1.0
1.0
1.7
10
10
17
EDDS
EDDS
EDDS
EDDS
EDDS
EDDS
1.0
1.0
1.0
4.9
10
10
37
36
60
56
44
55
36
84
80
35
35
37
38
60
43
32
13
48
63
54
14
60
57
100
37
50
54
46
39
58
33
50
63
40
41
66
61
53
61
28
67
19
34
48
32
29
60
Pb
29
50
65
80
83
41
91
88
95
70
33
37
96
88
23
65
16
50
67
ref
this study
this study
this study
7
41
14
14
14
42
14
43
42
this study
this study
14
44
44
14
14
14
this study
this study
45
this study
this study
45
this study
this study
this study
24
this study
this study
Figure 9 shows the distribution of Cu and Zn fractions in soil
Dornach 2 before and after the extraction with EDDS. Cu
was mainly found in the “organic” fraction, while about 60%
of the Zn was found in iron oxide and residual fractions. The
applied EDDS solution extracted these metals mainly from
the first four fractions, the exchangeable, mobile, Mn oxides,
and organic fraction. Table 5 shows the relative reduction of
the respective metal fractions by EDDS. For Cu and Zn, the
first three fractions were reduced by 71%, 80%, and 75% on
average, the organic Zn fraction by 36%, and the organic Cu
fraction by 58%. The iron oxide and residual fractions were
not or only marginally reduced. The much better extractability
of Cu compared to Zn by chelating agents can therefore be
explained by the presence of larger weakly bound fractions.
The maximum extraction efficiency (pH 7) at high chelant:
metal ratio for soil Dornach 2 was 84% for Cu and 48% for
Zn (EDTA), while the first four fractions of the sequential
extraction amounted to 79% for Cu and 41% for Zn. A similar
result also holds for Pb in soil Rafz. The maximum extraction
efficiency (pH 7) is 95% (EDTA), which is also the percentage
of Pb in the first four fractions of the sequential extraction.
Van Benschoten et al. (13) found similar results for seven
soils from sites polluted with Pb. The Pb not removed was
found in the Fe oxide, sulfide, and residual fractions. When
looking at the extractants used in the sequential extraction,
this result is not surprising since fraction 4 is determined by
extraction with 25 mM EDTA (31, 32).
Table 1 shows the properties of each soil. These properties
influence the distribution of the metals between the different
fractions as classified by sequential extraction. For example,
soil Dornach 1 is a calcareous soil with a high pH, which is
likely to make the metals less labile and less easily extractable.
Soil Dornach 2 has little carbonate and a neutral pH, while
the soil Rafz soil has a acidic pH with a low clay content,
TABLE 5. Reduction of Metal Fractions in Soils by Extraction with EDDS (in %), According to the Zeien and Bru1 mmer Scheme
(31, 32)
metal
soil
exchangeable
mobile
Mn oxide
organic
amorphous Fe
crystalline Fe
residual
Zn
Dornach 1
Dornach 2
Rafz
73
86
92
69
83
80
65
68
64
26
44
40
2
20
8
-5
1
-23
-32
-33
-25
Cu
Dornach 1
Dornach 2
63
40
84
86
87
91
46
69
12
33
-14
16
-32
-32
Pb
Rafz
16
37
36
17
0
-
-11
Literature Cited
FIGURE 9. Sequential extraction of Cu and Zn in soil Dornach 2
before and after extraction with EDDS.
making the metals more easily available. This can be seen
in the distribution of Zn in these three soils: 31% Zn in soil
Dornach 1 is found in the first four fractions, while 41% is
found in soil Dornach 2 and 57% in soil Rafz. The different
extraction efficiencies for different metal obtained by the
chelating agents are therefore simply due to the different
distribution of the metal fractions. Extraction by chelating
agents is therefore feasible for soils containing a large fraction
of metals in the first four fractions of the sequential extraction
procedure applied here but will not be an efficient treatment
for soils containing a high percentage of metals in strongly
bound fractions.
Acknowledgments
We thank Werner Attinger and Anna Grünwald for help with
the soil sampling and the chemical analyses and Diederik
Schowanek from Procter & Gamble for providing S,S-EDDS.
This work was funded in part by the Federal Office for
Education and Science within COST Action 837 and the Swiss
National Science Foundation in the framework of the Swiss
Priority Program Environment.
(1) Mulligan, C. N.; Yong, R. N.; Gibbs, B. F. Eng. Geol. 2001, 60,
193-207.
(2) Peters, R. W. J. Hazard. Mater. 1999, 66, 151-210.
(3) Elliott, H. A.; Brown, G. A. Water, Air, Soil Pollut. 1989, 45, 361369.
(4) Steele, M. C.; Pichtel, J. J. Environ. Eng. 1998, 124, 639-645.
(5) Pichtel, J.; Pichtel, T. M. Environ. Eng. Sci. 1997, 14, 97-104.
(6) Pichtel, J.; Vine, B.; Kuula-Vaisanen, P.; Niskanen, P. Environ.
Eng. Sci. 2001, 18, 91-98.
(7) Kim, C.; Lee, Y.; Ong, S. K. Chemosphere 2003, 51, 845-853.
(8) Papassiopi, N.; Tambouris, S.; Kontopoulos, A. Water, Air, Soil.
Pollut. 1999, 109, 1-15.
(9) Xie, T.; Marshall, W. D. J. Environ. Monit. 2001, 3, 411-416.
(10) Reed, B. E.; Carriere, P. C.; Moore, R. J. Environ. Eng. 1996, 122,
48-50.
(11) Cline, S. R.; Reed, B. E. J. Environ. Eng. 1995, 121, 700-705.
(12) Di Palma, L.; Medici, F. Waste Management 2002, 22, 883-886.
(13) Van Benschoten, J. E.; Matsumoto, M. R.; Young, W. H. J. Environ.
Eng. 1997, 127, 217-224.
(14) Ghestem, J. P.; Bermond, A. Environ. Technol. 1998, 19, 409416.
(15) Elliott, H. A.; Linn, J. H.; Shields, G. A. Hazard. Waste Hazard.
Mater. 1989, 6, 223-229.
(16) Elless, M. P.; Blaylock, M. J. Int. J. Phytorem. 2000, 2, 75-89.
(17) Hessling, J. L.; Esposito, M. P.; Traver, R. P.; Snow, R. H. In
Metals speciation, separation and recovery; Patterson, J. W.,
Passino, R., Eds.; Lewis Publishers: Chelsea, MI, 1989; Vol. 2.
(18) Han, F. X.; Banin, A. Water, Air Soil Pollut. 1999, 114, 221-250.
(19) Scheidegger, A. M.; Sparks, D. L. Soil Sci. 1996, 161, 813-831.
(20) Bucheli-Witschel, M.; Egli, T. FEMS Microbiol. Rev. 2001, 25,
69-106.
(21) Nowack, B. Environ. Sci. Technol. 2002, 36, 4009-4016.
(22) Ebina, Y.; Okada, S.; Hamazaki, S.; Ogino, F.; Li, J. L.; Midorikawa,
O. J. Natl. Cancer Inst. 1986, 76, 107-113.
(23) Schowanek, D.; Feijtel, T. C. J.; Perkins, C. M.; Hartman, F. A.;
Federle, T. W.; Larson, R. J. Chemosphere 1997, 34, 23752391.
(24) Vandevivere, P.; Hammes, F.; Verstraete, W.; Feijtel, T.;
Schowanek, D. J. Environ. Eng. 2001, 127, 802-811.
(25) Grcman, H.; Vodnik, D.; Velikonja-Bolta, S.; Lestan, D. J. Environ.
Qual. 2003, 32, 500-506.
(26) Geiger, G.; Federer, P.; Sticher, H. J. Environ. Qual. 1992, 22,
201-207.
(27) Kayser, A.; Schröder, T. J.; Grünwald, A.; Schulin, R. Int. J.
Phytorem. 2001, 3, 381-400.
(28) Vandevivere, P. C.; Saveyn, H.; Verstreate, W.; Feijtel, T. C.;
Schowanek, D. R. Environ. Sci. Technol. 2001, 35, 1765-1770.
(29) Ritter, S. K. Chem. Eng. News 2001, 79/27, 24-28.
(30) Potthoff-Karl, B.; Greindl, T.; Oftring, A. Seifen-Oele-FetteWachse-J. 1996, 6, 392-397.
(31) Zeien, H.; Brümmer, G. W. Mitt. Dtsch. Bodenkundl. Ges 1989,
59, 505-510.
(32) Zeien, H. Ph.D. Dissertation, Friedrich-Wilhelms-Universität,
Bonn, 1995.
(33) Zang, V.; Evan Eldik, R. Inorg. Chem. 1990, 29, 1705-1711.
(34) Martell, A. E.; Smith, R. M.; Motekaitis, R. J. NIST critically selected
stability constants of metal complexes, version 6.0; NIST:
Gaithersburg, MD, 2001.
(35) Nowack, B.; Kari, F. G.; Krüger, H. G. Water, Air, Soil Pollut.
2001, 125, 243-257.
(36) Jaworska, J. S.; Schowanek, D.; Feijtel, T. C. J. Chemosphere
1999, 38, 3597-3625.
(37) Nowack, B.; Lützenkirchen, J.; Behra, P.; Sigg, L. Environ. Sci.
Technol. 1996, 30, 2397-2405.
(38) Nowack, B.; Sigg, L. J. Colloid Interface Sci. 1996, 177, 106-121.
VOL. 38, NO. 3, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
943
(39) Nowack, B.; Sigg, L. Geochim. Cosmochim. Acta 1997, 61, 951963.
(40) Takahashi, R.; Fujimoto, N.; Suzuki, M.; Endo, T. Biosci.
Biotechnol. Biochem. 1997, 61, 1957-1959.
(41) Theodoratus, P.; Papassiopi, N.; Georgoudis, T.; Kontopoulos,
A. Water, Air, Soil Pollut. 2000, 122, 351-368.
(42) Hong, P. K. A.; Li, C.; Banerji, S. K.; Regmi, T. J. Soil Contam.
1999, 8, 81-103.
(43) Yu, J.; Klarup, D. Water, Air, Soil Pollut. 1994, 75, 205-225.
944
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 3, 2004
(44) Barona, A.; Aranguiz, I.; Elias, A. Environ. Pollut. 2001, 113,
79-85.
(45) Linn, J. H.; Elliott, H. A. Water, Air, Soil Pollut. 1988, 37, 449458.
Received for review August 7, 2003. Revised manuscript
received November 3, 2003. Accepted November 4, 2003.
ES0348750