Environmental Engineering Reference
In-Depth Information
forest soils. The soil columns were subjected to leaching
with different intensities and acidities. Schlüter observed
that at the lower end of the inorganic Hg and MeHg load-
ing range, Hg species were associated with SoM in the o
horizon, and Hg transport was due to the DoM transport.
At a higher loading, however, Hg species were less strongly
associated with the SoM and changes in the chemistry of
rain could potentially lead to their mobilization, prompt-
ing the author to suggest that in soil with a high level of
Hg contamination, ion exchange with H
1
may lead to Hg
mobilization. MeHg was consistently more mobile than
inorganic Hg in this study.
Biester et al. (2002b) studied Hg speciation in soils down-
wind from chlor-alkali plants using a mercury-thermo-
desorption (MTD) technique (Biester and Nehrke, 1997;
Biester and Scholz, 1997), as well as with chemical extrac-
tions. They defined the Hg extracted in NaBH
4
as the reac-
tive Hg fraction, distinguished from the organic-bound Hg
fraction. The Hg desorption spectra for organic-rich (9-15%
SoM) and sandy (average 0.5% SoM) soils resembled those
of Hg-humic standard and Hg-Fe(III) hydroxide standard,
respectively. In the organic-rich soils, water-soluble Hg was
mainly associated with the DoM, whereas in the sandy
soil, up to 50% of the water-soluble Hg was in the NaBH
4
-
extractable fraction (Biester et al., 2002b). Hg mobility was
highest in soils with a high level of soluble SoM and low
level of clayey material. Accordingly, they proposed that
Hg mobility is not dominated solely by the DoM and that
the clay content also plays a role. In the soil profiles, they
observed a significant vertical Hg transport only in high-
SoM samples. Hg in the low-SoM sample showed very little
vertical transport because of the low vertical transport of
DoM. Effect of pH on Hg mobility may be understood as
its effect on the mobility of organic matter, and not on the
binding of Hg to soil. They did not report the DoM con-
centrations in the leachate, however.
in the environment. The initial Hg concentration used in
Andersson's experiments was 40.2 ppm, inevitably resulting
in an extremely high Hg surface coverage and, especially
in the case of mineral oxides, perhaps surface precipitation
of Hg(oH)
2(s)
(Tiffereau et al., 1995). At such high Hg con-
centrations, organic matter's reduced sulfur groups would
largely be saturated, and as such, depending on its amount,
organic matter may not have the capacity for further sorp-
tion. Adsorption experiments conducted with high Hg con-
centrations may indeed result in a gross underestimation of
the binding strength (Lövgren and Sjöberg, 1989; Skyllberg
et al., 2000). Therefore, it is difficult to compare the behav-
ior of different sorbents in the environment at such high
Hg concentrations. Also, a pH increase results in enhanced
SoM dissolution and Hg mobilization, which is a different
mechanism than a loss in SoM binding strength as com-
pared with the mineral constituents.
In tropical soils, Fe and Al hydroxide minerals play a very
important role in binding Hg (Semu et al., 1987; roulet et
al., 1998; Lacerda et al., 2004). Weathered oxisols (later-
ites) contain high concentrations of Fe and Al hydroxides.
Fadini and Jardim (2001) observed a strong positive corre-
lation between Hg and the soil Fe and Al concentrations
down to a depth of 1 m in the Negro river basin (Amazon),
with soil Fe exhibiting a higher Hg adsorption efficiency
(i.e., higher molar Hg:Fe ratios) than soil Al. A higher soil
Al content, however, resulted in presumably equivalent or
higher Hg adsorption to Al than to Fe minerals. Fadini and
Jardim found a poor correlation between SoM and Hg,
as also reported for other tropical soils (Aula et al., 1994;
roulet and Lucotte, 1995; roulet et al., 1998). roulet et al.
(1998) observed a strong correlation between Hg and the
reductively extractable Al in acidic equatorial soils. Since Al
hydroxide is not known to strongly adsorb Hg (Andersson,
1979), roulet et al. (1998) concluded that Hg adsorption may
have been enhanced by Al substitution of the Fe hydroxide,
resulting in an increase in the mineral-specific surface area.
Semu et al. (1987) observed a strong correlation between
Hg adsorption capacity and the cation exchange capacity
for a Tanzanian oxisol, suggesting the importance of ion
exchange in Hg adsorption in this soil. They also observed
a negative correlation between kaolinite and inorganic
Hg, confirming the observations by Andersson (1979) that
kaolinite is indeed a weak sorbent for inorganic Hg.
In the absence of Cl
2
, inorganic Hg(II) was shown to
form an inner-sphere surface complex via two deproton-
ated surface oxygen atoms coordinated to two edge-sharing
Fe atoms on goethite (a-FeooH) (Collins et al., 1999).
Using surface complexation theory, Tiffereau et al. (1995)
modeled the adsorption of inorganic Hg onto a-Sio
2
and
amorphous Fe(oH)
3(s)
. The experimental data were from
the work of MacNaughton (1973) and Avotins (1975). The
modeled Hg binding constants were greater for Fe(oH)
3(s)
than a-Sio
2
, as expected from the reactivity of these sur-
faces. For the a-Sio
2
system, the initial Hg:solid ratio was at
920 ng g
2
1
, where up to ~70% of the added Hg was removed
ASSociATion of MeRcuRy Specie S
wiTh Soil MineR AlS
Andersson (1979) reviewed the Hg adsorption literature,
especially with respect to pure aluminosilicates and other
minerals. Based on a set of experiments with organic soils
and some pure minerals, Andersson classified SoM as the
most efficient component in soils for Hg adsorption at
acidic pH values, but mineral fractions as the most efficient
component at circumneutral pH. Among soil minerals,
illite and Fe
2
o
3
were the most effective sorbents, followed
by montmorillonite, kaolinite, and Al(oH)
3
. Later reviews
agreed with the conclusions of Andersson's work as to the
effectiveness of soil minerals over organic matter in adsorb-
ing Hg at circumneutral and higher pH (Schuster, 1991;
Jackson, 1998). Because of the analytical challenges of low-
level Hg measurements at the time of Andersson's research,
the Hg concentrations used in these experiments were
several orders of magnitude higher than those observed
