STRUCTURAL ENVIRONMENT AROUND TH4+ IN SILICATE-GLASSES - IMPLICATIONS FOR THE GEOCHEMISTRY OF INCOMPATIBLE ME4+ ELEMENTS

The structural environment around Th4+ in several silicate glasses containing 1-3 wt% Th4+ was investigated as a function of melt composition and polymerization using Extended X-ray Absorption Fine Structure (EXAFS) spectroscopy. One set of glasses had the bulk chemical composition of monticellite (MO), diopside (DI), anorthite (AN), albite (AL), olivine basalt (O-BAS), tholeiitic basalt (T-BAS), calc-alkaline dacite (DAC), and calc-alkaline rhyolite (RHY). A second set of Th-bearing glasses had the bulk AL composition but also contained either 5.5 wt% F or 2000 ppm Cl. In glasses containing 3 wt% Th, Th-coordination varies from 8 (d[(VIII)Th-O] almost-equal-to 2.41 +/- 0.02 angstrom in MO and DI) to a mixture of 8 and 6 coordinations (d[(VI+VIII)Th-O] almost-equal-to 2.35-2.36 +/- 0.02 angstrom in all other compositions). The presence of (VIII)Th is related to the thorianite (ThO2)-saturation in the glass. In AL glasses containing 1 wt% Th, 6-fold coordinated Th dominates (d[(VI)Th-O] almost-equal-to 2.32 +/- 0.02 angstrom). No clear evidence for F or Cl complexes around Th is observed in any of the halogen-containing glasses. Weak but significant interactions of Th with tetrahedral {Si, Al} network and/or alkali elements are observed, especially in highly polymerized compositions (RHY), in which Th-{Si, Al, Na, K} contributions are detected between 3.25 and 3.50 +/- 0.05 angstrom. These results suggest that, in this set of glasses, Th does not form "complexes" but rather acts as a weak network modifier. The presence of (VI)Th in silicate glasses, an unusual low coordination for Th, suggests that the ionic radius of Th in magmatic systems is lower than previously reported (r(Th) = 0.94 angstrom). The resulting high Th-O bond strength (almost-equal-to 0.7 vu), unknown in minerals, may explain the observed low crystal-melt partition coefficients of Th4+, i.e., its incompatible character during magmatic differentiation. Increasing polymerization is predicted to favor higher coordination around Th (VIII, IX) as a result of a shortage of NBO to which Th preferentially bonds. Hence, dramatic changes in the crystal chemistry of Th are predicted, i.e., the incorporation of Th in minerals. Comparison with U4+ suggests that (VI)Th is less abundant in similar glass/melt systems, compared with (VI)U. This may explain the U/Th fractionation often observed in zircon, as well as the preferential incorporation of Th in accessory minerals containing large sites in which Th4+ can be substituted (e.g., the Ca2+ site in titanite, apatite, or zirkelite).