The influence of natural organic matter on the speciation and solubility of Eu in Boom Clay porewater

Abstract

Eu solubility / Organic matter / Boom Clay water / Geochemical modeling Summary. The influence of natural organic matter (NOM) on the speciation and solubility of europium (Eu) was studied under geochemical conditions representative for the Boom Clay. Different organic matter types were used, and analysis was performed both after 0.45 µm microfiltration and after 30 000 MWCO ultrafiltration to distinguish between larger colloids (assumed to be immobile under in situ conditions) and small dissolved species.Equilibrium was approached from undersaturation starting from synthesised Eu(OH) 3 (s), which, during the experiment, transformed into EuOHCO 3 (s), in agreement with thermodynamic considerations. In the absence of NOM, the Eu solution concentrations after 0.45 µm filtration exceeded the thermodynamic solubility of EuOHCO 3 (s) by several orders of magnitude, indicating the presence of inorganic Eu colloids. In the presence of NOM, the Eu solubility increased with increasing NOM concentration as was expected, but, surprisingly, was dependent on the operational size cut-off: at an identical NOM concentration in the filtrate, the Eu solution concentration after 0.45 µm filtration was consistently higher compared to the Eu concentration after 30 000 MWCO filtration. This latter observation necessitates detailed knowledge concerning the pore size cut-off of Boom Clay under in situ conditions in order to use the correct Eu-NOM complexation constant and/or maximum solubility in transport calculations. At higher NOM concentrations (TOC > 30 mg/L) the Eu solubility after 0.45 µm filtration was seemingly independent of the NOM concentration.In contrast, after 30 000 MWCO ultrafiltration, the Eu solution increased linearly with increasing DOC, from the expected thermodynamic solubility (∼ 5 × 10 −7 mol L −1 ) at 0 mg L −1 DOC to ∼ 3 × 10 −5 mol L −1 at 80 mg L −1 DOC. All of the data sets were modelled using the Nagra/PSI database [1] for solubility, hydrolysis and inorganic aqueous complexation reactions, and fitted organic complexation reactions between Eu 3+ and NOM functional groups. Both a free ligand approach (with electrostatic correction) and the humic ion-binding model VI [2], which was for the first time successfully introduced into Phreeqc geochemical code, were tested and provided equally good fits to the data.

An improved description of the interactions between rare earth elements and humic acids by modeling: PHREEQC-Model VI coupling

“…By this method we found Q PHREEQC = 0.92 Â Q WHAM + 3 Â 10 À4 (R 2 = 1). Since SA decreases with P, our results are consistent with those of Liu et al (2008) who determined SA = 30000 m 2 g À1 at an ionic strength of 0.015 mol L À1 and P = À120.…”

“…The 80 types of sites considered by Model VI were defined as the SUR-FACE_MASTER_SPECIES, and their respective stability constants as the SURFACE_SPECIES in the "minteq.v4" database. This procedure is similar to that used previously by Liu et al (2008) to model the binding of Eu by HA in clay pore water using the same PHREEQC-Model VI combination, except that in this study, we used DLK 2 values -0. Therefore, the 80 binding sites defined by Tipping (1998) are taken into account versus only 32 sites in Liu et al (2008).…”

“…This procedure is similar to that used previously by Liu et al (2008) to model the binding of Eu by HA in clay pore water using the same PHREEQC-Model VI combination, except that in this study, we used DLK 2 values -0. Therefore, the 80 binding sites defined by Tipping (1998) are taken into account versus only 32 sites in Liu et al (2008). In order to model similar REE binding by each specific HA sites with WHAM 6/Model VI and PHREEQC/Model VI, REE-HA binding simulations were compared with the corresponding REE-simple ligands complexes in solution.…”

“…The stability constants of the REE-HA phenolic sites were thus independently estimated. For this purpose, Model VI equations were introduced into PHREEQC ( Parkhurst and Appelo, 1999), following Liu et al (2008). The PHRE-EQC/Model VI model was then used to fit the HA-REE binding datasets published by Sonke and Salters (2006), Pourret et al (2007b) and Marsac et al (2010).…”

Thiol groups controls on arsenite binding by organic matter: New experimental and modeling evidence

“…Previous studies, where Models V, VI or VII were coupled to PHREEQC, attempted to convert this empirical electrostatic humic ion-binding model into the diffuse layer model (DLM) formalism [40][41][42][43][44]. This type of conversion requires the calculation of a surface area (A) that depends on the ionic strength similar to the calculations made for polyelectrolytes such as polyacrylic acid [45].…”