Chemistry Reference
In-Depth Information
or in different laboratories. These physicochemical properties are labeled A-F in
Table 5.20
.
The covalent index
X
r
m
( )
, labeled A in Table 5.20 was used in 4 differ-
ent QSARs by Newman's group (McCloskey et al. 1996; Newman and McCloskey
1996; Tatara et al. 1997; Tatara et al. 1998; Newman et al. 1998] and the QSAR with
the highest
r
2
value was developed by McCloskey et al. (1996) for 20 cations. The
covalent index
X
r
m
( )
was also used by Enache et al.
(2003), Van Kolck et al. (2008),
and Mendes et al. (2010) to develop QSARs (Table 5.20). The QSARs developed
by Enache et al.
(2003) for 11 cations and Mendes et al. (2010) for 18 cations had
the highest
r
2
values (Table 5.20). The log of the first hydrolysis constant (|log
K
OH
|)
labeled B in Table 5.20 was used in 4 different QSARs by Newman's group and the
QSAR with the highest
r
2
value was developed by McCloskey et al. (1996) for 9
cations. The log of the first hydrolysis constant (|log
K
OH
|) was also used by Enache
et al.
(2003), van Kolck et al. (2008), and Mendes et al. (2010) to develop QSARs
(Table 5.20). The QSARs developed by Enache et al.
(2003) for 12 cations and
Mendes et al. (2010) for 18 cations had the highest
r
2
values (Table 5.20). Ionization
potential (IP), labeled C in Table 5.20, was used by 3 different teams of investiga-
tors (Sauvant et al. 1997; Enache et al.
2003; and Sacan et al. 2009) to develop
QSARs. The IP QSAR developed by Sauvant et al. (1997) to predict the toxicity of
8 cations on oxygen consumption by nitrifying bacteria had the highest
r
2
values
(Table 5.20). QSARs developed with
X
r
m
2
and |log
K
OH
| are labeled D in Table 5.20.
These QSARs were used only by Newman's group (McCloskey et al. 1996; Tatara
et al. 1998) and Mendes et al. (2010). All 3 QSARs included a large number of
cations (17-20) and had high
r
2
values, but the Mendes et al. (2010) QSAR had
the lowest p value. Van Kolck et al. (2008), Sacan et al.
(2009), and Mendes et al.
(2010) developed QSARs with the ionic index (Z
2
/r), labeled E in Table 5.20. The
r
2
values of the van Kolck et al. (2008) and Mendes et al. (2010) QSARs were
lower than the
r
2
values for the Sacan et al. (2009) QSARs (Table 5.20). The only
other QSARs that used less numerous physicochemical properties used to predict
cation toxicity were those developed by Newman's group (McCloskey et al. 1996;
Newman and McCloskey 1996; Tatara et al. 1997; Tatara et al. 1998) for
X
r
m
2
and
Z
2
/r (Table 5.20). These QSARs are labeled F in Table 5.20 and had higher
r
2
values
for QSARs for predicting toxicity for a high (17-20) number of cations (Table 5.20).
5.4 NONPHYSICOCHEMICAL PROPERTIES USED
TO PREDICT CATION TOXICITY
Six non-physico-chemical properties have been used to predict cation toxicity
(
Table 5.21
). These include serum concentrations, calmodulin activity, freshwater con-
centrations, abundance in the earth's crust, soil concentrations, and elemental compo-
sition of soils and plants.
Joseph and Meltzer (1909) established that the toxicity of the alkaline earth met-
als (magnesium, calcium, potassium, and sodium) to dogs was in inverse proportion
to the amounts in the dog's serum. Cox and Harrison (1983) compared Williams
et al.
(1982) mouse LD
50
cation toxicity values with the cation's ability to mimic Ca
2+
in stimulating the intracellular Ca
2+
receptor protein, calmodulin. Williams et al.
