THE COMBINED EFFECTS OF FO2 AND MELT COMPOSITION ON SNO2 SOLUBILITY AND TIN DIFFUSIVITY IN HAPLOGRANITIC MELTS

Abstract

The diffusion profile method has been employed to measure tin diffusion coefficients and SnO2 solubility in water-saturated, peralkaline to peraluminous haplogranitic melts at 850°C, 2 kbar, and log fO2 conditions ranging from FMQ - 0.57 to FMQ + 3.49. At reduced conditions cassiterite is highly soluble and tin is present dominantly as a Sn2+ species, whereas at oxidized conditions SnO2 is much less soluble, and tin is present dominantly as a Sn4+ species. There is a strong melt composition control on SnO2 solubility; solubilities are at a minimum at the subaluminous composition, increase strongly with alkali content in peralkaline compositions and weakly with Al content in peraluminous compositions. In the case of the latter, this increase can only be distinguished at reduced conditions, e.g., at a logfO2 of FMQ - 0.57 cassiterite solubility increases from 2.78 to 4.11 wt% SnO2 for melt with Al/(Na + K) compositions (A.S.I.) of 1.0 and 1.2, respectively. At oxidized conditions SnO2 solubility is ~500 ppm for both the A.S.I. 1.0 and 1.2 compositions. By comparison SnO2 solubilities in the most peralkaline composition investigated range from 3.94 wt% to ~10 wt% SnO2, for the most oxidized to the most reduced conditions, respectively. Thermodynamic modelling of the data indicates that the Sn4+/#Sn ratio in the melt is also at a minimum at the subaluminous composition, ranging from ~0.4 at logfO2 of FMQ + 3.49 to ~0.01 at FMQ - 0.57. Over the same log fO2 range the Sn4+/#Sn ratio for the A.S.I. 0.6 composition ranges from ~0.98 to ~0.4 and for the A.S.I. 1.25 composition, from ~0.8 to ~0.02.Tin diffusivity is dependent on both fO2 and melt composition. The effective binary diffusion coefficient of tin at reduced conditions is approximately 10-7.5 cm2/sec for the peraluminous compositions and 10-8.2 cm2/sec for the peralkaline compositions. At oxidized conditions these values decrease to approximately 10-8.2 and 10-9.0 cm2/sec, respectively. These are interpreted to reflect relatively fast diffusion where Sn2+ is the dominant valence and tin in this case behaves similar to a network modifier and relatively slow diffusion where Sn4+ is dominant and tin likely has a lower coordination number. Alternatively, the coordination of Sn2+ and Sn4+ is the same, but the bond strengths are different. At fixed fO2 the faster diffusivity in the peraluminous compositions reflects the lower Sn4+/Sn2+ ratio. The fact the Sn4+/Sn2+ ratio in melts varies greatly with fO2 at redox conditions near FMQ suggests that the partitioning behaviour of tin possibly changes during the evolution of an igneous suite in general and of a peraluminous granite suite in particular.