Sorption of trivalent actinides (Cm, Am) and their rare earth analogues (Lu, Y, Eu, Nd, La) onto orthoclase: Batch experiments, Time-Resolved Laser Fluorescence Spectroscopy (TRLFS) and Surface Complexation Modeling (SCM)

Sorption is one of the main processes, which determine the retention of radionuclides (RN) in a repository for nuclear waste. In a multi-barrier system, the host rock poses the ultimate barrier retarding the release of RN into the environment. Feldspars (e.g. orthoclase) are one of the main constituents of crystalline rock (e.g. granite), which is considered one potential host rock type in many countries (e.g. Finland, Sweden, Germany). In this study, the sorption of trivalent actinides (Cm, Am) and their rare earth analogues (Lu, Y, Eu, Nd, La) onto orthoclase (K feldspar) is investigated. For reliable predictions concerning the migration of RN, a process understanding on the molecular level of such processes is necessary. To achieve this, batch sorption experiments are combined with TRLFS and SCM.
Batch experiments were performed covering a broad range of experimental conditions (pH 4-11, oxic and anoxic conditions, [M3+] = 10-6-10-4 M, 3-50 g/L orthoclase (grain size: < 21 and 63-200 µm; SSA: 4.2 and 0.2 m2g-1)). Weak retardation below pH 5, followed by a strong increase between pH 5 and 7 and complete removal from solution at pH ≥ 8 was observed for all investigated metals. Cm- and Eu-TRLFS-measurements suggested the formation of an outer-sphere surface complex at lower (pH<5) and two different inner-sphere surface complexes at higher pH values (pH > 5 and pH > 7.5, respectively). Surface precipitation was observed for higher metal concentrations (10-4 M). As the investigated metals revealed a similar behavior over a broad range of conditions, a generic approach was used for the SCM to describe the system as a whole. Experimental data of different series with different metals were simultaneously fitted by coupling PHREEQC with UCODE using the same underlying speciation model. Resulting generic stability constants for the involved surface complexes will be presented.
The identification of comparable processes and their unified description with one suitable model is important to map the complexity of natural systems onto simplified geochemical models. This step is crucial for large-scale reactive transport calculations needed for a reliable safety assessment of potential repository sites, as they require enormous computing efforts.