the chemical are defined by the capital letters: E is the excess molar refraction,
S represents polar interactions, A the H-bond donor properties, B the H-bond
acceptor properties, and V is the molar volume of the chemical. J
+ and J
À characterize the interactions that are connected to the positive or negative charge of the ion.
The equation was originally developed for solvents, but has been demonstrated to
give reasonable predictions for structural proteins as well (Henneberger et al.
2016b). Unfortunately, for ions this modeling approach appears rather empirical
with no clear mechanistic background. Furthermore, the required chemical descriptors are not available for many ions and the influence of pH, competing ions, or ion
pairing cannot be considered. Various other attempts have been made to predict
sorption of ionic species in soils. These models appear to cover some aspects quite
well such as the relative influence of competing ions, the cation exchange capacity,
or some molecular structural entities within a class of compounds (Droge and Goss
2013a; Figueroa et al. 2004; Higgins and Luthy 2007; Jones et al. 2005; MacKay and
Vasudevan 2012; Sassman and Lee 2005). However, none of these models is able to
predict soil or sediments sorption coefficients for some standardized conditions (pH,
ionic strength, and composition) from molecular structure across a large number of
chemical classes. Sorption of ions to albumin was shown to depend on the 3D
geometry of the ions (Henneberger et al. 2016a). A 3D QSAR model was able to
cover and explain these effects (Linden et al. 2017), but this model is neither user
friendly nor able to cover a wider chemical domain. Much more successful was the
attempt to predict the sorption of both anions and cations to phospholipid membranes (Bittermann et al. 2014, 2016; Timmer and Droge 2017). This can be done
with a commercial software, COSMOmic, that is based on quantum chemical
calculations and has a wide application domain due to its fundamental nature. For
235 molecules (among them 24 cations and 51 anions) the predictive error was 0.65
log units (Bittermann et al. 2014) and for 19 cationic amine surfactants in a separate
work the predictive error was even smaller (Timmer and Droge 2017).
6 Toxicity and Toxicokinetics of IOCs
6.1 Membrane Sorption and Narcosis
As stated above, the membrane sorption of ionic species from water is usually
smaller than that of their corresponding neutral species, but it is by no means
negligible. In general, anions show a stronger distribution to membranes than cations
because of the positive dipole potential in the interior of the membrane (Flewelling
and Hubbell 1986a, b). This is nicely illustrated by tetraphenylborate and
tetraphenylphosphonium which are almost identical with practically the same 3D
geometry and the same charge density but with opposite signs. However, their
membrane–water partition constants (K mem/water ) differ by several orders of magnitude with log K mem/water ¼ 5.2 [L water /L mem ] for the tetraphenylborate anion and log
K mem/water ¼ 1.2 for the tetraphenylphosphonium cation (Flewelling and Hubbell
1986a).
Environmental Sorption Behavior of Ionic and Ionizable Organic Chemicals
53
S represents polar interactions, A the H-bond donor properties, B the H-bond
acceptor properties, and V is the molar volume of the chemical. J
+ and J
À characterize the interactions that are connected to the positive or negative charge of the ion.
The equation was originally developed for solvents, but has been demonstrated to
give reasonable predictions for structural proteins as well (Henneberger et al.
2016b). Unfortunately, for ions this modeling approach appears rather empirical
with no clear mechanistic background. Furthermore, the required chemical descriptors are not available for many ions and the influence of pH, competing ions, or ion
pairing cannot be considered. Various other attempts have been made to predict
sorption of ionic species in soils. These models appear to cover some aspects quite
well such as the relative influence of competing ions, the cation exchange capacity,
or some molecular structural entities within a class of compounds (Droge and Goss
2013a; Figueroa et al. 2004; Higgins and Luthy 2007; Jones et al. 2005; MacKay and
Vasudevan 2012; Sassman and Lee 2005). However, none of these models is able to
predict soil or sediments sorption coefficients for some standardized conditions (pH,
ionic strength, and composition) from molecular structure across a large number of
chemical classes. Sorption of ions to albumin was shown to depend on the 3D
geometry of the ions (Henneberger et al. 2016a). A 3D QSAR model was able to
cover and explain these effects (Linden et al. 2017), but this model is neither user
friendly nor able to cover a wider chemical domain. Much more successful was the
attempt to predict the sorption of both anions and cations to phospholipid membranes (Bittermann et al. 2014, 2016; Timmer and Droge 2017). This can be done
with a commercial software, COSMOmic, that is based on quantum chemical
calculations and has a wide application domain due to its fundamental nature. For
235 molecules (among them 24 cations and 51 anions) the predictive error was 0.65
log units (Bittermann et al. 2014) and for 19 cationic amine surfactants in a separate
work the predictive error was even smaller (Timmer and Droge 2017).
6 Toxicity and Toxicokinetics of IOCs
6.1 Membrane Sorption and Narcosis
As stated above, the membrane sorption of ionic species from water is usually
smaller than that of their corresponding neutral species, but it is by no means
negligible. In general, anions show a stronger distribution to membranes than cations
because of the positive dipole potential in the interior of the membrane (Flewelling
and Hubbell 1986a, b). This is nicely illustrated by tetraphenylborate and
tetraphenylphosphonium which are almost identical with practically the same 3D
geometry and the same charge density but with opposite signs. However, their
membrane–water partition constants (K mem/water ) differ by several orders of magnitude with log K mem/water ¼ 5.2 [L water /L mem ] for the tetraphenylborate anion and log
K mem/water ¼ 1.2 for the tetraphenylphosphonium cation (Flewelling and Hubbell
1986a).
Environmental Sorption Behavior of Ionic and Ionizable Organic Chemicals
53