Environmental Engineering Reference
In-Depth Information
soils and pure-phase metal oxides appears to be driven by non-specific electrostatic
attraction, with specific electrostatic attraction and van der Waals interactions being
secondary. Both the carboxylate and the heterocyclic nitrogen groups may partici-
pate in sorption of quinmerac, facilitated by specific and non-specific electrostatic
attraction with charged soil mineral surfaces, and surface complexation. The hetero-
cyclic nitrogen, amine, and carbonyl groups of norflurazon do not appear to interact
with soil minerals.
A phenomenon that is observed when studying sorption of lipophilic, polar and
ionic contaminants is related to the displacement of contaminants initially sorbed
to the solid phase. This is for instance reported for sorption of pharmaceuticals to
sewage sludge by Jones et al. ( 2006 ). This phenomenon may be ascribed to initially
fast binding of the contaminant to a large number of low affinity binding sites (like
weak binding to the surface layer), followed by displacement of the contaminants by
for instance more hydrophobic constituents of the DOC or by (newly formed) DOC
constituents capable of interacting more strongly with the sorption sites. Thereupon,
these constituents may out-compete micro contaminants as they may be formed in
relative high concentrations due to (microbial) degradation of the organic matter
present in either the solid phase or in the pore water.
Bintein and Devillers ( 1994 ) suggested an empirical regression using K ow and
pK a as molecule descriptors and pH and f OC to describe the sorbent properties for
the estimation of the K d of both ionised and non-ionised contaminants:
log K d =
0.93
·
log K OW +
1.09 log f OC +
0.32
·
CF a
0.55
·
CF b +
0.25 (16.4)
where the correction factors CF a and CF b quantify the fraction of dissociated acids
and bases in the system:
1
CF a =
log
10 pH pKa
1
+
(16.5)
1
CF b =
log
10 pKa ( pH 2)
1
+
According to these authors, the soil pH should be entered for acids into the equa-
tion for the correction factor CF a . The pH required for the correction factor CF b
is the pH at the surface of the soil colloids and is lowered 2 pH units below the
pH of the bulk soil solution. Even though the coefficient of determination of this
equation was very high ( R 2
0.93), a short-coming of the regression, regarding
charged organic contaminants, is their insufficient representation in the test set (nine
acids and three bases out of 87 contaminants). Besides, it may also surprise that
the fraction of bases, CF b , was negatively correlated to the sorption coefficient K d ,
even though it might be expected that electrical attraction of the positively charged
cations by the negative potential of soil colloids would increase sorption of bases.
In summary, although the predictability of newly developed empirical formulae
is large, there is no rigorously tested method available to predict the K d or the K oc
of the majority of contaminants. Separate expressions are needed for the neutral and
the ionic molecule fraction in order to account for dissociation of organic acids,
bases and amphoters.
=
 
Search WWH ::




Custom Search