Effect of pH and dissociation on the fate and exposure of ionizable chemicals

Ionizable organic chemicals comprise an important fraction of pharmaceuticals, pesticides as well as industrial chemicals. It has been estimated that 33% of the preregistered REACH substances is mostly ionized at pH 7. To extend the appliccability of existing exposure models, a Multimedia Activity Model for Ionics (MAMI) was recently developed and tested. In the present study, the impact of the parameters describing ionization was assessed by performing the sensitivity and the uncertainty analysis on MAMI for the acids 2,4-D, pentachlorophenol, bisphenol-A, perfluorooctanoic acid and the bases aniline, 4-chloroaniline and trimethoprim. A realistic emission scenario was simulated for each test chemical. First, a sensitivity analysis was performed to assess the influence of individual input parameters on the results. Then, probability density functions were derived for the most sensitive input parameters. The sensitivity analysis showed that the parameters describing ionization, pH and the dissociation constant (pKa), are among the most sensitive model parameters. The uncertainty analysis, however, indicated that these parameters are not the major source of uncertainty, which statistically justifies the use of species-specific models for ionics. The water content in air is a sensitive parameter for the PEC in air of molecules with negligible air-water partition coefficient, such as ions. The uncertainty of the QSARs for solid-water sorption significantly affects the PECs in soils and sediments. In most cases, the uncertainty of PECs and of persistance is largely explained by the uncertainty of (bio)degradation rates, which may be caused by model assumptions, experimental or estimation errors or by the environmental variability, including the effect of pH.