The Science of the Total Environment 284 Ž2002. 191᎐203
Binding and mobility of mercury in soils contaminated
by emissions from chlor-alkali plants
H.F. Scholer
H. Biester U , G. Muller,
¨
¨
Institute of En¨ ironmental Geochemistry, Uni¨ ersity of Heidelberg, INF 236, D-69120 Heidelberg, Germany
Received 15 January 2001; accepted 12 May 2001
Abstract
Chlor-alkali plants are known to be an important source of Hg emissions to the atmosphere and related
contamination of soils in their vicinity. In the present study, the results of Hg speciation and mobility of Hg in soils
affected by Hg emissions from three chlor-alkali plants are compared. Solid phase mercury speciation analyses was
carried out using a mercury᎐thermo-desorption technique with the aim of distinguishing elemental Hg wHgŽ0.x from
HgŽII.-binding forms. Mercury species in soil leachates were distinguished using an operationally defined method,
which is based on the reactivity of soluble Hg compounds. Results show that the HgŽ0. emitted from the plants could
not be detected in any of the investigated soils. This indicates quantitative re-emission or oxidation of this Hg species
in the atmosphere or soils. In most soils Hg was predominately bound to organic matter. Only in sandy soils deficient
in organic matter was Hg, to a larger extent, sorbed onto mineral soil components. Leachable Hg in most soils
occurred as non-reactive, soluble organic Hg complexes such as fulvic acid-bound Hg, and reach their highest values
Ž90 ␮g kgy1 . in soils rich in organic matter. Concentrations of reactive, soluble Hg compounds were highest in sandy
soils where the content of organic matter was low. Leachability of Hg was found to be inhibited in soils with a high
content of clayey soil components. The distribution of Hg in soil profiles suggests that migration of Hg to deeper soil
layers Žapprox. 20 cm. is most effective if Hg is bound to soluble organic complexes, whereas reactive Hg or weak Hg
complexes are effectively retained in the uppermost soil layer Ž5 cm. through sorption on mineral surfaces. 䊚 2002
Elsevier Science B.V. All rights reserved.
Keywords: Chlor-alkali plants; Mercury; Soils; Binding forms; Mobility
U
Corresponding author. Tel.: q49-6221-544819; fax: q49-6221-545228.
E-mail address: [email protected] ŽH. Biester..
0048-9697r02r$ - see front matter 䊚 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 8 - 9 6 9 7 Ž 0 1 . 0 0 8 8 5 - 3
192
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
1. Introduction
Chlor alkali plants ŽCAPs. are known to be an
important source of Hg emissions to the atmosphere ŽNriagu and Pacyna, 1988; Maserti, and
Ferrara, 1991., which was estimated to be 18%
Ž1982. in Europe ŽPacyna and Munch,
1991., and
¨
Ž
.
4.5% in the USA EPA 1997 of total gaseous Hg
emitted from anthropogenic sources. Despite the
fact that only a small portion of Hg emitted from
CAPs is deposited near the plant ŽJernelov
¨ and
Wallin, 1973; Wallin, 1976; Hogstrom
¨ ¨ et al., 1979;
Lodenius and Tulisalo, 1983; Lodenius, 1998..
However, Hg concentrations in soils near CAPs
were reported to exceed background concentrations by a factor of up to 75 ŽEPA, 1997..
Mercury deposition in soil is known to occur in
three major forms, namely wet and dry deposition
of HgŽII. and HgŽ0., respectively, and the deposition of particulate Hg. The EPA model calculation of Hg deposition around CAPs using an ISC3
model suggest that within 50 km from the plant,
deposition of HgŽII. exceeds deposition of HgŽ0.
by up to a factor of three ŽEPA 1997..
Mercury deposited to soil is known to be subject to a wide array of chemical and biological
transformation processes such as HgŽ0. oxidation,
and HgŽII. reduction or methylation depending
on soil pH, temperature and soil humic content
Že.g. Weber, 1993.. An overview of Hg binding in
soils is given in Schuster Ž1991. or Stein et al.
Ž1996.. It is known that Hg mobilisation in soils
through formation of inorganic soluble Hg compounds such as HgCl 2 and HgŽOH. 2 , are of minor importance in the presence of organic matter
as Hg is known to be effectively bound to soil
humic substances Že.g. Kerndorf and Schnitzer,
1980; Weber, 1988.. The formation of organic Hg
ŽII. complexes is known to be the dominating
process, which is due largely to the affinity of
HgŽII. and its inorganic compounds to sulfur-containing functional groups ŽSchuster, 1991; Xia et
al., 1999; Yin et al., 1997.. On the one hand, this
complexing behaviour greatly limits the mobility
of mercury in soil as most of the mercury in soil is
fixed in the bulk organic matter and could therefore, be only mobilised through elution in run-off
by being attached to suspended soil or humus. On
the other hand, HgŽII. will be partly absorbed
onto soluble organic soil components such as
fulvic acids, and may then partition to run-off in
the dissolved phase Že.g. Lindqvist et al., 1991;
Meili, 1991; Mierle and Ingram, 1991.. However,
the atmospheric input of mercury to soil is thought
to greatly exceed the amount leached from soil,
and the amount of mercury partitioning to run-off
is considered to be a small fraction of the amount
of mercury stored in soil. Despite the fact that
mercury in soils is therefore known to have a long
retention time, it may continue to be released to
surface waters for long periods of time, thus
increasing the length of time anthropogenic emissions would impact the environment. Recent estimates indicate that approximately 95% of Hg
emitted from anthropogenic sources resides in
terrestrial soils ŽEPA, 1997.. Appropriate mass
balances were, for example, calculated for Swedish
soils and watersheds Že.g. Lindqvist et al., 1991..
However, until now, little has been reported
about binding forms and mobility of soil Hg derived from emissions of CAPs. This study is a
continuation of a previous work where Hg deposition to soil within 1 km downwind from three
CAPs has been examined ŽBiester et al., 2001.. In
the first part of this study, it was shown that Hg
concentrations in soils from all three sites were
enriched in Hg by factors of between 2 and 5.8
compared to local background levels calculated
from median Hg concentrations.
Here, we present results of soil Hg speciation
analysis in the solid phase and in soil solution
from selected samples from the same three chlor
alkali plant sites. Mercury speciation analysis was
carried out using a mercury᎐thermo-desorption
technique ŽMTD. introduced in previous studies
ŽBiester and Scholz, 1997; Biester and Nehrke,
1997. with the main aim to distinguish metallic
Hg from HgŽII.-binding forms such as humic acid
bound Hg, which might result from HgŽ0. oxidation or deposition of HgŽII.. MTD analysis was
supplemented by Hg analysis in humic acid extracts ŽNaOH.. This method is sometimes used to
determine the amount of organically bound Hg in
soils and sediments Že.g. DiGiulio and Ryan, 1987;
Dmytriw et al., 1995.. Quantification and speciation of soluble mercury in soils was carried out by
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
analysing operationally defined, soluble Hg
species Žreactive- and non-reactive complex-bound
soluble Hg. in soil leachates. Mercury distribution
in soil profiles was analysed to estimate mercury
mobility. Results of solid- and liquid phase Hg
speciation, Hg leachability, and mobility were related to basic soil parameters such as grain size
distribution and organic carbon content.
2. Methods and materials
The surface soil samples investigated in this
second part of the study, were selected from a
total of 56 samples from site S1, 51 samples from
site S2, and 70 samples from site S3. The distribution of Hg concentrations in surface soils within
100᎐1000 m downwind from the sites and sampling locations are given in the first part of this
study ŽBiester et al., 2001.. From each site 20
samples were selected for MTD-measurements.
Most of the surface soil samples were selected to
show Hg concentrations above the median Hg
concentration Ž157 ␮g kgy1 for S1, 541 ␮g kgy1
for S2, and 571 ␮g kgy1 for S3. in soils of each
site, respectively. From those 20 samples, four
were chosen for the determination of humic acid
bound Hg, and 12᎐14 for mercury solubility tests.
All samples were stored in polyethylene bags at
ambient temperature Ž15᎐25⬚C.. We presumed
that volatile Hg compounds such as HgŽ0., which
can exist under open field conditions Žsunlight,
wind, precipitation, etc.. would not be lost during
sample transport and storage at ambient temperatures.
2.1. Soil profiles
Mercury migration to deeper soil layers was
analysed in two soil profiles of each site. Soil
profiles were taken by digging a ditch of approximately 50 cm in depth. Samples were then cut
from the soil column in 3᎐5-cm sections down to
a depth of 20 cm using a plastic spatula. The
distance and direction of sampling points from
the plants are given in Fig. 4a,b,c.
193
2.2. Methods for determination of major soil
characteristics
The total carbon content in soils was determined by photometric ŽIR. detection of CO 2
after combustion of the homogenised dried and
ground sample Ž0.5 g. in a high frequency induction furnace. Inorganic carbon was calculated
from the amount of carbonates in the samples
determined by means of a ‘carbonate bomb’
ŽMuller
and Gastner, 1971. and subtracted from
¨
total carbon, which gives the amount of organic
carbon ŽC org ..
Grain size distribution in surface soils was determined by wet sieving of subsamples Ž100 g. of
two soil samples from each site, respectively,
which represent the predominant soil types in the
study area Žsee Fig. 1..
For soil pH measurement, 50 ml of 0.01 M
CaCl 2 solution was mixed with 20 g of fresh
sample material. Soil pH was determined using a
glass electrode after equilibrating for 1 h. Soil pH
was determined from the same samples selected
for determination of Hg solubility.
2.3. Mercury thermo-desorption analysis
Determination of Hg phases by solid
phase᎐Hg᎐thermo-desorption is based on the
specific thermal desorption or decomposition of
Hg compounds from solids at different temperatures. Mercury thermo-desorption characteristics
were determined using an in-house apparatus,
consisting of an electronically controlled heating
unit and a Hg detection unit. For Hg detection, a
quartz cuvette where the thermally released Hg is
purged through, is placed into the optical system
of an atomic absorption spectrometer ŽPerkin
Elmer AAS 3030., and absorption of Hg is detected at 253.7 nm at continuous detection mode
Ž1-s intervals.. Interferences, mainly from pyrolytic products of organic matter, were compensated by continuous deuterium background correction. Analysis was carried out at a heating rate
of 0.5⬚C sy1 and an N2 gas flow of 300 ml miny1 .
194
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
The lowest level of detection under such conditions was in the range of 40᎐50 ng if all Hg is
released within a single peak ŽBiester and Nehrke,
1997.. Results are depicted as Hg thermo desorption curves ŽMTDC., which show the release of
HgŽ0. vs. temperature. Depending on the total Hg
content of the samples, 20᎐250 mg of the fresh
material or approximately 2 mg of the quartzdiluted standard Hg compounds were used for
MTD measurements. MTDCs of standard materials are illustrated in Fig. 2.
Metacinnabar ŽHgS., HgŽ0., humic acid bound
Hg, and Hg sorbed to Fe-oxyhydrates are Hg
compounds known to be formed in soils. Solid
phase standards of these Hg compounds were
obtained as follows. Metacinnabar was obtained
through precipitation of HgS from a HgCl 2 solution using Na 2 S solution. For MTD measurements, approximately 0.001 M of black cinnabar
was mixed with 20 g of quartz powder for dilution.
Carboniferous schists from the Idrija Hg mine,
Slovenia, solely bearing metallic Hg Žvisible
droplets., were used as a standard of unbound
elemental Hg. HgŽ0.-incubated Fe-oxyhydrates
were prepared by incubating dry Fe-oxyhydrates
in a sealed container for 14 daysr40⬚C in a HgŽ0.
saturated atmosphere. Mercury bearing humic
acids were extracted with 0.1 M NaOH from a
forest soil Žnear the cinnabar smelter of the Idrija
mercury mine, Slovenia. contaminated by atmospheric Hg deposition. Humic acids were precipitated from the supernatant at pH 2 ŽHCl., washed
twice and freeze-dried.
2.4. Extraction of Hg bound to humic and ful¨ ic
acids from soils
Humic and fulvic acids were extracted twice
from 20 g of a fresh soil sample material by
shaking end-to-end overnight with 60 ml of 0.1 M
NaOH. After centrifugation, the supernatant containing the humic and fulvic acid fraction was
decanted. Humic and fulvic acids were digested
by heating 5 ml of the extracts with 10-ml concentrated nitric acid for 3 h at 120⬚C. Mercury was
then detected by CVAAS after reduction of HgŽII.
with stannous chloride Žsee Table 1..
2.5. Leaching tests
Soil leachates were obtained by shaking 20 g of
fresh sample material for 24 h end-to-end in
250-ml polyethylene centrifuge bottles with 200
ml of demineralized water. After centrifugation
for 2 h Ž5000 rpm., the supernatant was decanted
in 50-ml PE bottles for further treatment. The
mercury species in solution were determined similar to the method used by Meili et al. Ž1991..
They distinguished reactive, soluble Hg which
could be reduced by SnCl 2 solution Žeasily reducible., non-reactive Hg, and Hg weakly bound
to complexes which could be reduced by NaBH 4
ŽNaBH 4-reducible Hg., and strongly bound, nonreactive soluble Hg which is bound to soluble
organic complexes Žcomplex bound Hg. Žsee Table
2.. In our samples, we could not easily determine
reducible or reactive soluble Hg in soil solutions.
Addition of the acid stannous chloride, caused
precipitation of humic acid in most samples, coprecipitating Hg. Therefore, we could only determine the NaBH 4 reducible and the total soluble Hg fraction. As NaBH 4 is a stronger reductant than stannous chloride, the NaBH 4 reducible Hg fraction in this study includes the
easily reducible, reactive Hg fraction. NaBH 4 reducible Hg was obtained by adding 1 ml of 3%
NaBH 4 in 1.5% KOH, to 5 ml of the leachates.
Total Hg in the leachates was determined after
oxidation of all Hg compounds using 500 ␮l of a
0.05 mol ly1 BrCl solution for 100 ml of the soil
leachate, followed by stannous chloride HgŽII.
reduction and CV-AAS Hg detection ŽTSP-Hg
Analyser., similar to the method of Szakacs et al.
Ž1980.. The amount of Hg bound in non-reactive
organic complexes was calculated by subtracting
the NaBH 4 reducible Hg fraction from total soluble Hg.
3. Results and discussion
3.1. Soil composition
Site S1 is dominated by peat soils in forest
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
areas ŽS1-31. and organic rich soils at grassland,
which show a slightly higher amount of clayey
components ŽS1-11. ŽFig. 1.. Soils of site S2 show
lower concentrations of organic matter ŽTable 3.,
but much higher amounts of clayey components
reaching up to 75% of all soil components ŽFig.
1.. Soils from site S3 consist of fine to medium
grained sands with a very low content of organic
matter ŽTable 4., and clayey components ŽFig. 1..
The Ah soil layer in these soils is mostly absent or
very thin. The marine origin of the sands is indicated by the occurrence of shell debris.
3.2. Mercury᎐thermal-desorption (MTD)
characteristics
The most important finding of MTD-analysis
was that no free or weakly adsorbed HgŽ0., which
is the dominant Hg species emitted from the
plants, could be detected in any of the soil samples. All samples showed a higher Hg release
temperature compared with the HgŽ0. standard
sample, which already starts degassing from the
samples at ambient temperatures ŽFig. 2.. However, HgŽ0. in soil gas, which is known to be
attributed to natural reduction processes of HgŽII.
ŽAlberts et al., 1974; Rogers and McFarlane,
195
1979., could not be detected by MTD analysis, as
it usually occurs in the pg ly1 range, which is
below the detection limit of the MTD system.
Soils from sites S1 and S2 show Hg desorption
between 180 and 300⬚C with the main Hg release
between 200 and 250⬚C ŽFig. 3.. Lower Hg desorption temperatures were found for samples from
site S3, which release Hg between 150 and 280⬚C
with a Hg release maximum at 200⬚C indicating
lower thermal stability of Hg binding in these
soils. These results indicate that Hg in all samples
exists in a matrix bound form, which means that
mercury might be physically adsorbed to mineral
surfaces or chemically bound to functional groups
of the organic matter as HgŽII.. MTD measurements indicate that curves of S1 and S2 organic
rich soils ŽFig. 3. fit better with that of the Hg in
the humic acids standard ŽFig. 2. than with that of
Hg sorbed to mineral compounds such as Hg
spiked Fe-oxihydrates, which show lower Hg desorption temperatures ŽFig. 2.. In contrast, curves
of S3 soils ŽFig. 3. match best with the second
peak of the HgŽ0. spiked Fe-oxihydrates, which
could be interpreted as HgŽ0. bound to innersphere sorption sites. The lower Hg desorption
temperatures of S3 samples indicate physical
sorption rather than chemical binding of Hg in
these soils. As clay minerals in soils of site S3 are
Fig. 1. Grainsize distribution in selected soils taken in the vicinity of three chlor-alkali plants ŽS1, S2, S3..
196
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
Fig. 2. Comparison of thermo-desorption curves of standard mercury compounds or mercury binding forms, which might potentially
occur in soils around chlor-alkali plants.
mostly absent ŽFig. 1.. Fe-oxihydrates or other
sesqui oxides are supposed to be potential mineral components for Hg sorption in these soils.
The deposition of Hg derived from emissions of a
coal-fired power plant located nearby, as assumed
in a previous study ŽBiester et al., 2001., suggests
that at least parts of the Hg in these soils might
be associated with particles emitted from this
coal-fired plant, which does not have any filtering
systems.
These results reflect the composition of the
soils at the three sites. Soils of S1 and S2 show
large amounts of organic matter, which explains
the predominant binding of Hg to humic substances, whereas the sandy soils of S3 with a very
low content of organic matter mainly offer sorption sites on mineral surfaces. However, any
metal᎐organic complex may be specifically adsorbed onto mineral surfaces ŽSchuster, 1991.; therefore, it is reasonable to suggest that Hg binding
occurs with both types of components in all kinds
of soils, but to a different extent.
The absence of elemental Hg in the soils might
be primarily due to re-emission. Hogstrom
¨ ¨ et al.
Table 1
Mercury in extracted humic- and fulvic acids ŽHArFA.;
percentage of humic acid bound Hg of total Hg and content of
organic carbon in selected soil samples of the chlor alkali
plant sites S1, S2 and S3
Sample
Total Hg
Ž␮g kgy1 .
Hg in HArFA
Ž␮g kgy1 .
S1-06
S1-07
S1-31
S1-48
Mean
1102
761
4188
467
1629
679
806
2165
450
1025
61.6
106
51.7
96.5
78.95
18.6
3.2
16.4
22.2
15.1
S2-03
S2-06
S2-40
S2-45
Mean
1265
3005
1123
1505
1724
1153
2204
403
1414
1293
91.2
73.3
35.9
93.9
73.6
14.2
12.1
8.5
2.3
9.3
S3-19
S3-43
S3-48
S3-61
Mean
624
2296
1033
1131
1271
515
501
682
151
462.3
%
82.5
21.8
66.1
13.4
46
Corg%
0.28
0.69
0.63
0.33
0.48
Relative S.D. of total Hg analyses was - 8 Ž n s 4. and
21% Ž n s 3. for Hg in HArFA-analyses, and 4% Ž n s 3. for
organic carbon determination.
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
197
Table 2
Total Hg, NaBH 4 -reducible, complex bound and total soluble Hg in soil leachates and organic carbon in soil samples of site S1
Sample
no.
Total Hg
wŽd.w..,
␮g kgy1 x
Total
sol.Hg
Ž␮g kgy1 .
Total
sol. Hg Ž%.
NaŽBH4 .
red. Hg Ž%.
Complexbound Hg Ž%.
Corg.
Ž%.
Soil
pH
S1-6
S1-7
S1-11
S1-19
S1-30
S1-31
S1-33
S1-34
S1-37
S1-45
S1-48
S1-49
S1-51
Mean
Median
1102
761
767
897
3331
4188
612
753
632
390
467
398
486
1137
757
63
21
17
27
63
90
26
47
16
18
15
14
10
33.1
24
5.8
2.9
2.3
3.1
1.9
2.2
4.3
7.6
2.5
4.5
3.2
3.5
2.1
3.51
3.14
0.72
0.47
0.33
0.22
0.57
0.27
0.47
0.52
0.41
n.da
n.d.a
n.d.a
0.54
0.35
0.38
5.03
2.39
1.92
2.85
1.34
1.88
3.79
7.08
2.11
4.50
3.20
3.50
1.52
3.16
3.01
18.6
3.2
3.7
8.30
31.6
16.4
16.8
9.3
7.3
25.1
22.2
29.6
27.0
16.9
16.8
4.14
3.80
6.79
6.04
3.94
3.96
4.11
4.39
3.99
3.64
3.8
3.92
3.28
4.29
3.98
Relative S.D. of total soluble Hg and NaŽBH 4 .-reducible Hg was in the range of 13᎐29% Ž n s 3..
a
Not detected Ž- 1 ␮g kgy1 ..
Ž1979. suggested that up to 25% of Hg deposited
in soils around CAPs, are re-emitted within a
short time. However, model calculations of the
US-EPA using the ISC3 dispersion model for
gaseous compounds ŽEPA, 1997., indicate that
dry deposition of HgŽII. is the predominant
Table 3
Total Hg, NaBH 4 -reducible, complex bound and total soluble Hg in soil leachates and organic carbon in soil samples of site S2
Sample
no.
Total Hg
wŽd.w..
␮g kgy1 x
Total
sol. Hg Ž␮g kgy1 .
Total
soluble Hg Ž%.
NaŽBH4 .
red. Hg Ž%.
Complexbound Hg Ž%.
Corg.
Ž%.
Soil
pH
S2-1
S2-3
S2-6
S2-7
S2-11
S2-14
S2-17
S2-28
S2-37
S2-39
S2-40
S2-45
S2-49
S2-51
Mean
Median
2181
1265
3005
2215
2267
826
154
208
1041
455
1123
1505
535
1062
1274
1092
14.1
11.3
10.2
18.7
18.2
8.29
n.d.a
1.26
15.3
3.2
6.04
8.60
1.54
12.80
9.25
9.4
0.64
0.89
0.34
0.85
0.80
1.00
n.d.a
0.60
1.47
0.71
0.54
0.57
0.28
1.20
0.71
0.68
0.20
0.40
0.12
n.d.a
n.d.a
n.d.a
n.d.a
n.d.a
n.d.a
0.31
0.20
0.10
n.d.a
n.d.a
0.10
0
0.44
0.49
0.26
0.85
0.80
1.00
n.d.
0.60
1.47
0.40
0.34
0.47
0.28
1.20
0.61
0.48
3.10
10.90
12.07
8.13
10.70
6.12
3.25
6.45
6.87
7.51
8.48
2.33
6.65
6.87
7.10
6.87
6.34
6.94
5.92
6.73
7.21
6.20
6.43
6.60
6.47
6.12
7.39
7.51
6.37
6.89
6.65
6.6
Relative S.D. of total soluble Hg and NaŽBH 4 .-reducible Hg was in the range of 16᎐27% Ž n s 3..
a
Not detected Ž- 1 ␮g kgy1 ..
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
198
Table 4
Total Hg, NaBH 4 -reducible, complex bound and total soluble Hg in soil leachates and organic carbon in soil samples of site S3
Sample
no.
Total
Hg
wŽd.w.. ␮g kgy1 x
Total
sol
Hg Ž␮g kgy1 .
Total
sol. Hg Ž%.
NaŽBH4 .
red. Hg Ž%.
Complexbound Hg Ž%.
Corg
Ž%.
Soil
pH
S3-6
S3-9
S3-14
S3-19
S3-23
S3-37
S3-38
S3-43
S3-48
S3-57
S3-61
S3-67
Mean
Median
1554
1050
782
624
888
969
1440
2296
1033
1530
1131
460
1146
1042
11.0
10.9
9.24
31.9
23.3
7.8
5.47
47.2
8.9
12.9
17.4
2.9
15.7
11
0.71
1.0
1.2
5.1
2.6
0.80
0.38
2.1
0.86
0.84
1.5
0.63
1.5
0.95
0.24
0.51
0.48
1.78
1.05
0.15
n.d.a
1.21
n.d.a
0.51
1.11
n.d.a
0.59
0.50
0.47
0.53
0.70
3.33
1.58
0.65
0.38
0.85
0.86
0.33
0.43
0.63
0.90
0.64
1.33
0.47
0.53
0.28
0.44
0.90
0.44
0.69
0.63
1.6
0.33
2.6
0.85
0.58
6.72
6.31
6.16
6.81
5.90
6.12
6.24
5.82
6.89
6.76
6.52
5.87
6.34
6.32
Relative S.D. of total soluble Hg and NaŽBH 4 .-reducible Hg was in the range of 19᎐26% Ž n s 3..
a
Not detected Ž- 1 ␮g kgy1 ..
process of soil contamination within 50 km from a
hypothetical chlor-alkali plant emitting 30 and
70% HgŽII. and HgŽ0., respectively. They calculated that 14.7% of total emitted Hg at a humid
site, and 13.3% at an arid site is deposited as
HgŽII. within this distance. In contrast, only 8 and
4.4% of total emitted Hg was deposited as HgŽ0.
at the humid and arid site, respectively. Deposi-
Fig. 3. Comparison of thermo-desorption curves of mercury in soils in the vicinity of three chlor-alkali plants compared to standard
Hg compounds. The small figures provide an overview of the initial temperature Žlines. of Hg release of all analysed samples of
each site, respectively.
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
tion of HgŽII., which is easily sorbed in soil,
probably explains most of the Hg enrichment
observed in the soils near the plants. Oxidation of
HgŽ0. to HgŽII. in soils is assumed to be generally
slow, so that most HgŽ0. might be re-emitted
before oxidation.
3.3. Mercury in extracted humic and ful¨ ic acids
Analysis of Hg in humic and fulvic acid extracts
ŽTable 1. confirm the results of MTD analysis
that binding of Hg to humic substances is the
predominant process of Hg sorption in soils of S1
and S2. However, extraction rates of Hg in these
samples vary considerably, and show no correlation with the amount of organic carbon in the
soil. Values above 90% were, in most cases, obtained from forest soils, whereas samples taken
from grassland showed lower values. Despite this,
the content of organic carbon in all samples of
site S3 was below 1%; extraction rates of Hg
bound to humic and fulvic acid still reached 46%
on average ŽTable 1.. These data support the
assumption derived from MTD measurements
that binding of Hg to organic matter in S6 soils
occurs to a lesser extent than in S1 and S3 soils.
The extracted, organically bound Hg determined here does not include other organic
fractions such as organo-mineral compounds
Žhumin fraction. or non-humified organic matter,
where Hg could also be bound. Moreover, the
efficiency of humic acid extraction using NaOH
may depend on soil properties. Accordingly, the
results presented in this study give only an estimation of the predominant Hg binding form in
the investigated soils
3.4. Solubility of mercury in soils
Average mercury concentrations in soil
leachates ranged from 9.25 ŽS2. to 31.7 ␮g kgy1
ŽS1., which corresponds to 0.76᎐3.69% of total
Hg ŽTables 2᎐4.. As HgŽ0. and HgŽII. are the
main Hg species derived from the emissions of
chlor-alkali plants, it could be assumed that all
Hg deposited on the soils is initially reactive
ŽSnCl 2 reducible.. The data of Hg species analysis in S1 and S2 soil leachates ŽTables 2 and 3.
199
indicate that, on average, more than 90% of
soluble Hg occurs in a non-reactive complex
bound form Žorganic complexes. confirming the
preferentially binding of Hg to the organic matter
in those soils. The amount of reactive or weakly
complexed, non-reactive Hg ŽNaBH 4 reducible
Hg. in these soils is therefore, generally low.
Mercury in leachates of S3 samples also mainly
occurs in a strong, complex-bound form, but to a
minor extent, compared with S1 and S2 soils.
Here, reactive or weakly complexed, non-reactive
Hg reaches 35% on average, and exceeds 50% of
total soluble Hg in some of the samples ŽTable 4..
Comparing the data on Hg in soil leachates
with the content of organic matter ŽTables 2᎐4.
and the grain size distribution in the soils ŽFig. 1.
suggests that Hg mobility and the Hg species in
soil solution strongly depend on soil composition.
The results show that Hg solubility is highest in
soils containing large amounts of organic matter
and low amounts of clayey components Žgrain
sizes - 2 ␮m.. Coupling of Hg to fulvic acid, and
the important role of these soluble humic sub
stances as carriers of soil derived Hg into freshwater systems has been demonstrated in numerous studies Že.g. Lindqvist et al., 1991; Meili,
1991; Mierle and Ingram, 1991; Driscoll et al.,
1995.. The data also indicate that clayey soil
components can inhibit Hg mobility through adsorption of soluble Hg bearing humic and fulvic
acids, even when the organic matter content in
the soil is high. Therefore, retention of Hg in soils
is not determined solely by the amount of organic
matter as suggested in other studies ŽYin et al.,
1997.. One mechanism of the formation of such
organo᎐mineral complexes is adsorption of the
positively charged metal cations of organic complexes to the negatively charged surface of clay
minerals, which leads to immobilisation of the
metal.
According to this general trend, the highest Hg
concentrations in leachates were found in the
peat soils of site S1, which show high amounts of
organic matter ŽTable 2. and low amounts of
clayey components ŽFig. 1.. Here, the median
percentage of total soluble Hg in soils was approximately 3% exceeding that of S2 and S3 soils
by a factor of 4.6 and approximately 2, respec-
200
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
tively. The formation of strong complexes of Hg
with organic ligands Žstability constants between
18.4 and 21.1; Stein et al., 1996. also explains that
the amount of NaBH 4 reducible, soluble Hg in
those soils is relatively low Žmean s 12.5% of
total soluble Hg. ŽTable 2..
Median mercury concentrations in soil
leachates of S2 do not exceed 10 ␮g kgy1 which
is 0.68% of median total Hg concentrations and
show, therefore, the lowest values of all sites
ŽTable 3.. Although total Hg concentrations in S2
soils are in the same range as those of S1 soils,
and the content of organic carbon is also high,
leachable Hg in S2 soils is significantly lower than
in soils of S1 due to the higher clay content in S2
soils ŽFig. 1., which favours the formation of
insoluble organo᎐mineral complexes. The amount
of reactive Hg and weakly complexed soluble Hg
in soils of this site varies to a wide extend. In
seven out of 13 samples, this Hg fraction was
below the detection limit Ž- 1 ␮g kgy1 .. Considerable amounts of this fraction were again
mostly found in samples with a comparatively
high content of organic matter ŽTable 3., which
Fig. 4. Distribution of mercury ŽHg. and organic carbon ŽC org . in profiles taken from soil in the vicinity of three chlor-alkali plants
wS1ŽA., S2ŽB., and S3ŽC.x. S1 profiles were taken 240 m NNE ŽP1., and 300 m E ŽP2. of the plant. S2 profiles were sampled 420 m
SSW ŽP1., and 930 SE ŽP2. from the plant. S3 profiles were collected 250 m S ŽP1., and 600 m NNW ŽP2. from the plant.
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
are mostly forest soils where the clay content in
the uppermost soil layer is low. Although not
having determined the clay content in all samples,
we assume that reactive Hg compounds and weak
soluble Hg complexes are also effectively sorbed
and immobilised by clayey soil components. Accordingly, it seems that the mobility of reactive
Hg compounds and weakly soluble, Hg complexes
in soils depends on the presence of clayey components rather than organic matter.
The median percentage of soluble Hg in S6
soils is 0.95% of total Hg, which is approximately
40% higher than in S2 soils, and by a factor of
approximately 3.3 lower than in S1 soils ŽTable 4..
In contrast to soils of the other sites, coupling of
Hg to soluble organic complexes is of minor importance. Here, the median amount of NaBH 4
reducible Hg in S3 soil leachates was 50% of total
soluble Hg and even exceeds the amount of complex bound Hg in some samples reaching up to
70% of total soluble Hg ŽTable 4.. We conclude
that the reason for the low amount of organic᎐Hg
complexes in leachates of S3 samples is because
of the low amount of organic matter in these
soils. Moreover, clayey soil components, which
could potentially immobilise NaBH 4 ᎐Hg compounds through sorption are also almost absent
in these sandy soils ŽFig. 1.. The predominance of
easily reducible and weakly complex bound Hg in
the soil leachates are in accordance with the
results found for solid phase Hg binding in these
soils which indicated predominately weak sorption of Hg to mineral soil components.
3.5. Mercury distribution in soil profiles
In profiles from all three sites, the highest Hg
concentrations were found in the uppermost Ž5
cm. soil layer ŽFig. 4a,b,c.. In profiles of sites S1
and S2, Hg concentrations show a gradual decrease with depth which, in most cases, follows
strongly the decrease of organic matter content
ŽFig. 4a,b.. This typical behaviour, which was
observed in several studies Že.g. Andersson, 1979.
indicates, on one hand, that retention of Hg is
strongly related to coupling to organic matter. On
the other hand, this implies that Hg is effectively
transported to deeper soil layers as soluble or-
201
ganic complexes as expected from the results of
Hg speciation in the soil leachates. At both sites,
Hg in soil profiles reach values near the local
background of the surface soils ŽS1s 75 " 25 ␮g
kgy1 ; S2s 150 " 50 ␮g kgy1 . at approximately 20
cm. In contrast to S1 and S2, profiles from S3
soils show a sharp decrease in Hg concentrations
from contaminated levels to background concentrations Ž58 " 35 ␮g kgy1 . within the uppermost 5
cm ŽFig. 4c.. In both S3 soil profiles, the decrease
in Hg concentrations also follow a sharp decrease
in organic carbon, which is more pronounced in
S3-P1 ŽFig. 4c.. These profiles imply that Hg in
these soils is strongly retained by organic matter,
which is highest in the uppermost soil layer. However, solid phase Hg speciation as well as Hg
speciation in the soil leachates, indicate that adsorption of Hg to mineral soil components is, in
this case, more important than binding of Hg to
organic matter.
As these sandy soils are highly permeable, we
presume that Hg can easily migrate to deeper soil
layers. Due to this assumption, we expected to
find a more gradual decrease of Hg concentrations with soil depth. However, the results indicate that under the given pH conditions, Hg bound
to mineral surfaces is, in fact, less mobile than Hg
bound to soluble organic complexes.
Our results have shown that binding and mobility of Hg in the investigated soils are ᎏ at least
in the case of soils from the sites S1 and S2 ᎏ
strongly determined by Hg᎐humic acid-complexes. Due to the known stability and solubility
of these organic metal complexes in a wide pH
range ŽAndersson, 1979; Xu and Allard, 1991., we
conclude that differences in pH in the soils of the
various sites ŽTables 2᎐4., play a minor role on
Hg mobility compared to that of, for example,
clay content. Speculating on changes of Hg mobility due to soil acidification, we assume that Hg
mobility in S1 and S2 soils might be even reduced
as the solubility of humic acid decreases with
decreasing pH ŽSchuster, 1991.. Moreover, it has
been recently reported that humic acids are increasingly sorbed to mineral surfaces such as iron
oxides with decreasing pH ŽAvena et al., 1999..
However, Hg sorbed to mineral components as
observed in S3 soils, is known to be effectively
202
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
desorbed with decreasing pH ŽAndersson, 1979.,
which might increase the mobility of Hg in these
soils.
4. Conclusions
This study has shown that HgŽ0., which is the
most important Hg species emitted from CAPs
Ž70%., is not preserved in soils surrounding the
plants. It is assumed that the major Hg compound
deposited near the plants is HgŽII., which is in
accordance with model calculation of Hg deposition near CAPs. HgŽ0. initially deposited to the
soils, is re-emitted within a short time or oxidised
to HgŽII. and sorbed to soil components suggesting that most HgŽ0. emitted from the plant is
directly or indirectly subject to long range
transport.
In soils which show a normal to high content of
organic matter, Hg was predominately bound to
organic matter. In sandy soils deficient in organic
matter, coupling of mercury to humic substances
occurs to a minor extent. In these soils, Hg was
less strongly bound to mineral soil components.
Leachable Hg in most soils, occurred as soluble
organic Hg complexes and was highest in soils
with a high content of organic matter and a low
content of clayey components. Leachability of Hg
was found to be inhibited in soils with a high
content of clayey soil components through the
formation of insoluble organo᎐mineral complexes. Soluble Hg compounds in organic rich
soils, mainly exist in complex bound non-reactive
form, whereas major occurrence of soluble reactive Hg or weak soluble Hg complexes was only
observed in sandy soils where the content of
organic matter was low. Mercury distribution in
soil profiles have shown that Hg contamination
reached down to approximately 20 cm in soils
where soluble Hg is organically bound following
the distribution of organic matter in the soil
column. In sandy soils lacking in organic matter,
contamination was restricted to the upper 5 cm
indicating that reactive Hg or weak Hg complexes
are effectively retained through sorption on mineral surfaces.
Acknowledgements
This study was initiated and funded by Eurochlor, Brussels. The help of the Eurochlor risk
assessment group and the helpful co-operation of
the site’s staff is acknowledged. Thanks to M.
Goodsite for his helpful comments on the
manuscript.
References
Alberts JJ, Schindler JE, Miller RW, Nutter DE. Elemental
mercury evolution mediated by humic acid. Science
1974;184:895᎐896.
Andersson A. Mercury in soils. In: Nriagu JO, editor. The
biogeochemistry of mercury in the environment.
ElsevierrNorth Holland Biomedical Press, 1979:79᎐112.
Avena MJ, Luuk K, Koopal LK. Kinetics of humic acid
adsorption at solid᎐water interfaces. Environ Sci Technol
1999;33Ž16.:2739᎐2744.
Biester H, Muller
G, Scholer
¨
¨ HF. Distribution and retention
of mercury in three different soils contaminated by mercury emissions from chlor-alkali plants. Sci Total Environ
2002;284:177᎐189.
Biester H, Nehrke G. Quantification of mercury in soils and
sediments ᎏ acid digestion vs. pyrolysis. Fresenius J Anal
Chem 1997;358:446᎐452.
Biester H, Scholz C. Determination of mercury phases in
contaminated soils ᎏ Hg-pyrolysis vs. sequential extractions. Environ Sci Technol 1997;31:233᎐239.
DiGiulio TR, Ryan EA. Mercury in soils sediments and clams
from a North Carolina peatland. Water Air Soil Pollut
1987;33:205᎐219.
Dmytriw R, Mucci A, Lucotte M, Pichet P. The partitioning of
mercury in the solid components of dry and flooded forest
soils and sediments from a hydroelectric reservoir, Quebec
ŽCanada.. Water Air Soil Pollut 1995;80r1᎐4:1099᎐1103.
Driscoll CT, Blette V, Yan C, Schofield CL, Munson R,
Holsapple J. The role of dissolved organic carbon in the
chemistry and bioavailability of mercury in remote adirondack lakes. Water Air Soil Pollut 1995;80r1᎐4:499᎐508.
EPA. Fate and transport of mercury in the environment.
Mercury study ᎏ report to congress, vol. III, EPA-452rR97-005, 1997.
Hogstrom
¨ ¨ U, Enger L, Svedung I. A study of atmospheric
mercury dispersion. Atmos Environ 1979;13:465᎐476.
Jernelov
¨ A, Wallin T. Airborne mercury fall-out on snow
around five Swedish chlor-alkali plants. Atmos Environ
1973;7:209᎐214.
Kerndorf H, Schnitzer M. Sorption of metals on humic acid.
Geochim Cosmochim Acta 1980;44:1701᎐1708.
Lindqvist O, Johansson K, Aastrup M, Andersson A,
˚ Meili
Brinkmark L, Hovsenius G, Hakanson
L, Iverfeldt A,
˚
H. Biester et al. r The Science of the Total En¨ ironment 284 (2002) 191᎐203
M, Timm B. Mercury in the Swedish environment ᎏ
recent research on causes, consequences and corrective
methods. Water Air Soil Pollut 1991;55:261.
Lodenius M. Dry and wet deposition of mercury near a
chlor-alkali plant. Science of Tot Env 1998;213:53᎐56.
Lodenius M, Tulisalo E. Environmental mercury contamination around a chlor-alkali plant. Bull Environ Contam
Toxicol 1983;32:439᎐444.
Maserti BE, Ferrara R. Mercury in soils, plants and atmosphere near a chlor-alkali complex. Water Air Soil Pollut
1991;56:15᎐20.
˚ Hakanson
Meili M, Iverfeld A,
L. Mercury in the surface
˚
water of Swedish forest lakes ᎏ concentrations, speciation
and controlling factors. Water Air Soil Pollut 1991;
56:439᎐453.
Meili M. The coupling of mercury and organic matter in the
biogeochemical cycle ᎏ towards a mechanistic model for
the Boreal forest zone. Water Air Soil Pollut 1991;
56:333᎐348.
Mierle G, Ingram R. The role of humic substances in the
mobilization of mercury from watersheds. Water Air Soil
Pollut 1991;56:349᎐357.
Muller
G, Gastner M. The Karbonatebombe ᎏ a simple
¨
device for the determination of carbonate content in sediments, soil and other material. N Jb Miner Mh 1971;
10:466᎐469.
Nriagu JO, Pacyna JM. Quantitative assessment of worldwide
contamination of air, water and soils by trace metals.
Nature 1988;333:134᎐139.
Rogers RD, McFarlane JC. Factors influencing the volatiliza-
203
tion of mercury from soil. J Environ Qual 1979;
8Ž2.:255᎐260.
Stein ED, Cohen Y, Winer AM. Environmental distribution
and transformation of mercury compounds. Crit Rev Environ Sci Technol 1996;26:1᎐43.
Szakacs O, Lasztity A, Orvath Z. Breakdown of organic Hg
compounds by hydrochloric acid᎐permanganate or bromine
monochloride solution for the determination of Hg by cold
vapor atomic absorption spectrometry. Analyt Cimica Acta
1980;121:219᎐224.
Wallin T. Deposition of airborne mercury from six Swedish
chlor-alkali plants surveyed by moss analysis. Environ Pollut 1976;10:101᎐114.
Weber JH. Binding and transport of metals by humic materials. In: Frimmel FH, Christman RF, editors. Humic substances and their role in the environment. John Wiley and
Sons, 1988:165᎐178.
Weber JH. Review of possible paths for abiotic methylation of
mercuryŽII. in the aquatic environment. Chemosphere
1993;26Ž11.:2063᎐2077.
Xia K, Skyllberg UL, Bleam WF, Bloom PR, Nater EA,
Helmke PA. X-ray absorption spectroscopic evidence for
the complexation of HgŽII. by reduced sulfur in soil humic
substances. Environ Sci Technol 1999;33Ž2.:257᎐261.
Xu H, Allard B. Effects of a fulvic acid on the speciation and
mobility of mercury in aqueous solutions. Water Air Soil
Poll 1991;56:709᎐717.
Yin Y, Alen HE, Huang CP. Kinetics of mercuryŽII. adsorption and desorption on soil. Environ Sci Technol
1997;4b31:496᎐503.