Workshop Report 29

Challenges and potential limitations: physicochemical properties

Todd Gouin

SEAC, Unilever, UK

Chemical activity (a), as it has been developed for neutral organic chemicals, is commonly estimated as the fraction of water solubility (liquid or sub-cooled liquid, if the chemical is a solid at room temperature): i.e.


where CW is the concentration of the chemical in the aqueous phase (e.g., mg/L) and SW is the water solubility (liquid or sub-cooled liquid for organic chemicals that are solids at the temperature of the test system). Equation 1 thus results in a dimensionless metric of between 0 and 1, which effectively provides a ratio of the energetic level in the aqueous phase observed in the environment/test system relative to the energetic level at the limit of solubility in water. Although estimating chemical activity using equation 1 is a relatively straight-forward exercise, it is important to acknowledge that uncertainty exists within both the numerator and the denominator. For instance, depending on how CW is derived it can include analytical error, associated with empirical methods, propagation of an error if estimated using an environmental fate model, or uncertainty related to the accuracy of a nominal concentration used in a toxicological test. Similarly SW will also have inherent uncertainty associated with values derived either empirically or estimated using a quantitative structure property relationship (QSPR).

The emphasis of this presentation was to highlight the influence that uncertainty in SW can have on estimates of chemical activity, which is meant to help stimulate discussion within the work group and focus activities towards approaches that could be used to help quantify the relative magnitude of error in SW.

In an attempt to provide preliminary insight regarding the variance that might exist in water solubility measurements, 233 chemicals reported by Mackay et al. (2006) were assessed with respect to their availability of solubility data. A graphical plot of the results is shown in Figure 3.4.1, which summarises 2440 solubility measurements for the 233 chemicals. The dataset reported by Mackay et al. (2006) are believed to provide a relatively good indication of the variance that might exist in empirically derived solubility data, with the majority of chemicals having >10 separate solubility measurements. A general observation from Figure 3.4.1 is that as solubility decreases the relative magnitude of the uncertainty increases, thus implying caution when relying on a limited number of solubility measurements for relatively insoluble organic chemicals (i.e. <0.01 mg/L).

Figure 3.4.1. Variability in reported water solubility measurements for 233 neutral organic chemicals represented as box and whisker plots for each individual chemical (primary axis)compiled from Mackay et al. (2006).


Green boxes represent the upper 25th percentile of data that are above the median value, whereas the orange boxes represent the lower 25th percentile of data that are below the median value. The maximum and minimum values are represented as the end of the whiskers above and below the median value, respectively. The number of data points included for each chemical plotted in the box and whisker plots are plotted as bars in relation to the secondary axis.

QSPRs used to estimate water solubility can also introduce different levels of uncertainty depending on the relative performance of the estimation method in relation to the chemical under investigation. Specifically, where test chemicals have properties that differ from those used in deriving the QSPR, it can be demonstrated that a higher level of uncertainty will undoubtedly be associated with the estimated value. Consequently, the use of estimated values of SW will require greater insight with respect to the performance of an individual QSPR in relation to the chemical being assessed. For organic chemicals that are solids at environmentally relevant temperatures, additional caution is warranted, particularly with respect to the manner by which melting point data are obtained and assumptions associated with Walden’s Rule in relation to the assumption that ΔSM = 56.5 J/mol K, when estimating fugacity ratios and deriving the sub-cooled liquid solubility of the chemical.

Lastly, the chemical activity concept has been demonstrated to work well for neutral organic chemicals, however, challenges currently exist with respect to applying equation 1 to miscible organic chemicals which do not have a quantifiable SW, and to ionisable organic chemicals, for which deriving SW may be problematic depending on the extent of ionisation that may exist, and which is a function of pH, ionic strength and on the influence of counter ions that may be present.