Chemistry Reference
In-Depth Information
similar in action. The absolute value of the difference in the probit model slopes
for each metal (Equations [1.6] and [1.7]) reflected the degree of deviation from
the similar action assumption. As anticipated, a clear trend emerged if this abso-
lute difference were plotted for ten binary mixtures against the differences in the
softness indices of the mixed metals (
Figure 1.8
bottom panel). The more similar
the paired metals' softness indices, the more they tended toward similar action.
It is interesting to note that this approach does not require that a complex series
of binary metal mixture experiments be conducted. It requires only probit model
slopes from single-metal tests. However, mixture experiments are required to
explore this approach from the vantage of independent action. Based on metal
independent action modeling, the magnitude of deviation of the interaction coef-
ficient (ρ in Equation [1.4]) from 1 suggests the degree of deviation from perfect
independent action (Figure 1.8 top panel). The more dissimilar the softness indices
of the paired metals, the more the estimated ρ deviated from 1. Independence
increased the more dissimilar the coordination chemistry of the paired metals.
Clearly, joint action of metals was influenced by coordination chemistry of the
combined metals.
1.4 CONCLUSION
[S]everal resolvable issues require attention before the QICAR approach has the same
general usefulness as the QSAR approach. These issues include exploration of more
explanatory variables, careful evaluation of ionic qualities used to calculate explana-
tory variables, examination of models capable of predicting effects for widely differing
metals (e.g., metals of different valence states), effective inclusion of chemical specia-
tion, examination of more effects, and assessment of the applicability of QICARs to
phases such as sediment, soils, and foods.
Newman et al. (1998, p. 1424)
It is now generally accepted that QSAR-like models can be generated for metal
ions based on fundamental coordination chemistry trends. Examples ranging from
adsorption to biological surfaces, to metal interactions in mixtures, to accumulation
and effects in metazoans were used in this short chapter to demonstrate this point.
Class (a) metal toxicity was easily related to electrostatic interactions with biologi-
cal ligands. Trends in effects of class (b) and intermediate metals were related to
qualities more closely linked to covalent bonding. Models involving a wider range of
class (a), intermediate, and class (b) metals might require more than one explanatory
variable based on different binding tendencies.
What is currently needed is a sustained exploration of the approach and further
refinement of metrics and methodologies. Progress toward filling the information
gaps highlighted in the above quote is evident in the literature, e.g., Can and
Jianlong (2007), Kinraide and Yermiyahu (2007), Newman and Clements (2008),
Ownby and Newman (2003), Walker et al. (2003), Wolterbeek and Verburg (2001),
and Zamil et al. (2009) and Zhou et al. (2011). These models will very likely emerge
in the next two decades to the level enjoyed now by QSAR models for organic
compounds.
