Adsorption of UVI and EuIII on illite: The important role of accessory minerals

Adsorption models able to predict the fate of (radio-)contaminants at water-mineral interfaces are important tools to assess the risk related to their migration in natural environments. In general, adsorption in a multimineralic environment is assumed to be governed by the major reactive mineral constituents. A common procedure to predict the adsorption by a soil/rock entity is to combine in an additive way the adsorption models for each single mineral composing the mineral assemblage, i.e. based on the relative reactive amounts of each mineral.
Ideally, robust and reliable adsorption models should be developed for each environmentally relevant mineral based on experimental data obtained under a wide range of conditions to guarantee their predictive capability. For this it is of utmost importance to study the retention of contaminants on well-characterized minerals, under well-defined and controlled conditions.
Diocthahedral alumina-silicates such as illite and montmorillonite are common mineralogical components of soils and sedimentary rocks which control the migration of (radio-)contaminants in natural systems through their exceptional bulk physical and chemical properties.
In the present study we will illustrate how the adsorption properties of a presumably pure clay mineral, illite du Puy (IdP), can be strongly altered by an accessory mineral (< 0.5 wt.%). The adsorption of UVI as well EuIII on purified IdP (<0.5 µm) under certain conditions (i.e., pH, concentration, presence/absence of dissolved carbonate) showed an unexpected enhancement of the retention of UVI and trivalent actinides and lanthanides (Cm/Am/Eu). The application of extended X-ray absorption fine structure spectroscopy and time resolved laser fluorescence spectroscopy on IdP samples loaded with UVI and EuIII, respectively, clearly suggest complexation by phosphate groups not expected in IdP. For U, the formed surface complex could be identified as an autunite-like phase. In the case of Eu, the formation of a ternary apatite-like Eu-Ca-PO4 solid phase was observed. Furthermore, leached phosphate from the clay mineral resulted in the precipitation of non-sorbed EuIII as rhabdophane (EuPO4×nH2O). The accessory mineral phase itself responsible for this anomalous behaviour is difficult to identify with common analytical methods such as TEM or XRD due to e.g. the low concentration and/or amorphous nature of the accessory phase or the low electron density of its constituent elements. Repeated acid-pre-treatment of IdP released P and Ca, pointing towards an apatite-like accessory phase as source of phosphate for the formation of the UVI and EuIII solid phases. In the pre-treated solid phosphate complexation was greatly reduced. In the case of UVI, an inner-sphere surface complex on IdP edge sites could be identified in the XAFS investigations. Our results demonstrate that adsorption models calibrated over a wide range of experimental conditions, complemented by a molecular-level control and a true mechanistic understanding, is of paramount importance for reliable modelling of adsorption processes at the solid-liquid interface.