Modeling metal ion-humic substances complexation in highly saline conditions

•Model VII and NICA-Donnan are tested at high ionic strength (I) using the SIT.•For I > 1 m, Model VII is applicable as a non-electrostatic model.•No modification is needed for NICA-Donnan.•Both models predict the effect of I on proton- and metal ion-humate (M-HA) binding.•SIT parameters for simpler M-HA binding models vary with pH and metal loading.

Because highly saline groundwaters are found at potential repository sites for nuclear waste, geochemical models should predict the speciation of relevant radionuclides in brines, including their complexation with substances such as humic acids (HA). In this study, available experimental radionuclide-HA complexation data in high 1:1 background electrolyte solutions (0.01 < mNaCl/NO3/ClO4 < 4 molal, m) are reviewed. Discrepancies in the amplitude of ionic strength effects on radionuclide-HA complexation are observed, which might depend on the nature of the interacting radionuclide or on the origin of HA. However, significant differences in the experimental conditions and calculations applied to determine conditional metal ion-HA complexation constants hamper direct comparison between these datasets. To clarify whether metal ion-HA binding in saline solutions can be described, two sophisticated humic-ion binding models (Model VII and NICA-Donnan) are presently used. This is the first time that Model VII and NICA-Donnan are applied to predict metal ion-HA binding at high ionic strength (I > 1 m). The advantage of these models, compared to more simple ones (e.g., the polyelectrolyte or the charge neutralization models), is that both electrostatic and chemical contributions to the overall metal ion-HA binding are explicitly taken into account. Model VII and NICA-Donnan are shown to produce very similar results. Trends in conditional metal ion-HA binding constants and in the maximum metal ion uptake by HA (e.g., the loading capacity) with I agree with experiments. The present data evaluation suggests that most of the apparent discrepancies between various experimental datasets arise from differences in the experimental conditions. Both Model VII and NICA-Donnan predict that the specific ion interaction theory (SIT) parameters for metal ion-HA systems, which are required for high ionic strength with more simple models, vary with pH and metal loading. Overall, Model VII and NICA-Donnan are able to account for various mechanisms involved in metal ion-HA complexation, including the metal loading effects and cation competition, and might be helpful predictive tools for performance safety assessment up to highly saline conditions.