Tin sorption to magnetite nanoparticles under anoxic conditions

The long-lived fission product 126Sn is of substantial interest in the context of nuclear waste disposal in deep underground repositories. However, the redox state (di- or tetravalent) under the expected anoxic conditions is still a matter of debate. We therefore investigated sorption and oxidation of Sn(II) in the presence of a typical corrosion product, magnetite (FeIIFeIII2O4), with a mean particle size of 9.4 nm. In order to simulate waste disposal conditions, the experiments were performed under strictly anoxic conditions in a glovebox at <2 ppm O2. Macroscopic parameters (pH, Eh, [Sn], [Fe]) were monitored along with redox state and local structure of Sn (X-ray absorption spectroscopy) and Fe (XPS) as a function of time, pH, and surface loading.
Magnetite rapidly sorbed Sn(II), reducing Sn concentration within 0.5 h from 10 to 0.0084 μμM. Tin was strongly sorbed by magnetite across a wide pH range from 3 to 9. Reduced sorption at pH <3 is in line with electrostatic repulsion between the positively charged surface of the magnetite nanoparticles (IEP ~6.7) and cationic Sn2+ or Sn4+ complexes. The reduced sorption at pH > 9 is in line with the transition from Sn(OH)20 to the anionic Sn(OH)3- which occurs at pH 9. Across the pH range 3-9 and reaction periods ≥1 h, EXAFS-derived sixfold oxygen coordination and XANES edge energy positions of ~29207 eV both indicate the presence of Sn(IV) at the magnetite surface. EXAFS shell fitting as well as Monte Carlo simulations showed formation of edge-sharing complexes of Sn(IV) with FeO6 octahedra (Sn-Fe distance of 3.15 Å), and formation of corner-sharing complexes with FeO4 tetrahedra (Sn-Fe distances of 3.60 Å). Even after the longest reaction periods of 1 month, we did not observe incorporation of Sn(IV) into the (compatible) magnetite structure. Also, precipitation of SnO2 was not observed in spite of an (initial) supersaturation.
In order to elucidate the reaction pathway, we also studied Fe in solution and at the surface (XPS). Starting with the PZC and increasing with [H+], the magnetite surface released Fe(II) into solution (0.11 g/L at pH 2). After addition of Sn(II), however, [Fe] in solution decreased as a function of Sn loading, in spite of the expected increase of structural Fe(II) due to the coupling to Sn(II) oxidation. This suggests a re-adsorption and possible re-precipitation of Fe(II) at the magnetite surface. Nevertheless, due to a protonation at low pH, Fe(II) again re-dissolved as a function of time. With XPS we were not able to detect an adequate increase of the Fe(II)/Fe(III) ratio at the surface, supporting an electron redistribution between bulk and surface Fe centers.
In conclusion, our study demonstrates that Sn is strongly retained by magnetite across a wide pH range, forming stable surfaces complexes and stabilising the magnetite surface against dissolution.