T9.7 Theoretical predition of isotopic fractionation processes: the case of Si, Fe and Li
Isotopic fractionation factors can be computed based on the vibrational properties of the relevant mineral phases, which in turn can be estimated form electronic structure calculations. These theoretical approaches have shown their capacity to semi-quantitatively predict these values, bnringing an indispensable help to the understanding of natural fractionation processes (Méheut et al. 2007). This is particluarly relevant for the "emerging" isotopic systems (e.g. Si, Fe, Mg, Ca, Li) .
Between these new systems, silicon presents a particular interest due to its ubiquity in terrestrial materials, but also for its role in alteration processes, or for the feeding of diatoms. In particular, magmatic processes fractionate silicon isotopes, giving a correlation between teh degreee of magmatic differentiation and the silicon isotopic signature of the rock (see e.g. Savage et al. 2011). In the absence of experimentally constrained values of the fractionation of silicon isotopes, this natural tendency seemed a good starting point for a detailled theoretical study. Following our preliminary work on the subject (Meheut et al 2009), we realized complementary calculations which permitted us to explain this tendency by general relationships between mineral structure, chemical composition and silicon fractionation properties (Méheut and Schauble, 2014).
The fractionation of iron isotopes also arouses considerable interest in Earth Sciences. As for silicon, these fractionation properties can be theoretically predicted based on electronic structure calculations (Blanchard et al. 2009), but also based on experimental estimates of the partial vibrational densities of state of iron from Mössbauer or Inelastic X-Ray Scattering spectroscopies (Polyakov and Soultanov 2011). The limits of this last technique, based on a comparison between the two methods, have been the subject of a critical comment from our group, in collaboration with researchers at IMPMC, Paris (Blanchard et al. 2012).
The PhD Thesis of Romain Dupuis, started in October 2011, is centered on the question of the theoretical prediction of isotopic fractionations implying liquid phases. We were particularly interested in the fractionation of silicon between mineral and solution (quartz-dissolved Si, kaolinite-dissolved Si), and between two dissolved species (H4SiO4/H3SiO4-, Dupuis et al. 2014, in rev. ). The development of new theoretical methods for the computation of fractionation properties, applied to the fractionation of lithium between mineral and solution, was also realized in collaboration with Prof. Mark Tuckerman at NYU (Marsalek et al. 2014, Dupuis et al., in prep. ).