Solid solutions of trace Eu(III) in calcite: Thermodynamic evaluation of experimental data over a wide range of pH and pCO 2

Abstract

The thermodynamics of dilute Eu-calcite solid solutions formed under widely different pH-pCO 2 conditions at T = 25°C and p = 1 bar were investigated using three sets of Eu(III) uptake experiments, two of which were taken from the literature: (a) recrystallization in synthetic cement pore water at pH ∼ 13 and pCO 2 ∼ 10 −13 bar (this work); (b) coprecipitation in 0.1 M NaClO 4 at pH ∼ 6 and pCO 2 ∼ 1 bar; (c) coprecipitation in synthetic seawater at pH ∼ 8 and pCO 2 ranging from 3 × 10 −4 to 0.3 bar. Solid solution formation was modeled using the Gibbs energy minimization (GEM) method. In a first step (“forward” modeling), we tested ideal binary solid solution models between calcite and the Eu end-members Eu 2(CO 3) 3 , EuNa(CO 3) 2, Eu(OH)CO 3 or Eu(OH) 3, for which solids with independently measured solubility products exist. None of these four binary solid solutions was capable of reproducing all three experimental datasets simultaneously. In a second step (“inverse” modeling), ideal binary solid solutions were constructed between calcite and the candidate Eu end-members EuO(OH), EuH(CO 3) 2 and EuO(CO 3) 0.5, for which no independent solubility products are available. For each single data point and each of these end-members, a free energy of formation with inherent activity coefficient term ( G α ⁎ = G α o + RT lnγ α) was estimated from “dual thermodynamic” GEM calculations. The statistical mean of G α ⁎ was then calculated for each of the three datasets. A specific end-member was considered to be acceptable if a standard deviation of ± 2 kJ mol −1 or less resulted for each single dataset, and if the mean G α ⁎ -values calculated for the three datasets coincided. No binary solid solution with any of the seven above mentioned end-members proved to satisfy these criteria. The third step in our analysis involved consideration of ternary solid solutions with CaCO 3 as the major end-member and any two of the seven considered Eu trace end-members. It was found that the three datasets can only be reproduced simultaneously with the ternary ideal solid solution EuH(CO 3) 2 – EuO(OH) – CaCO 3, setting G EuH(CO3)2 ⁎ = −1773 kJ mol −1 and G EuO(OH) ⁎ = −955 kJ mol −1, whereas all other end-member combinations failed. Our results are consistent with time-resolved laser fluorescence data for Cm(III) and Eu(III) indicating that two distinct species are incorporated in calcite: one partially hydrated, the other completely dehydrated. In conclusion, our study shows that substitution of trivalent for divalent cations in carbonate crystal structures is a more complex process than the classical isomorphic divalent-divalent substitution and may need consideration of multicomponent solid solution models.