Cations in olivine, Part 2: Diffusion in olivine xenocrysts, with applications to petrology and mineral physics

Diffusivities for calcium, iron, magnesium, manganese and aluminum have been measured for St. John's olivine undergoing cation exchange with synthetic basaltic melts. The variety of temperature, pressure and fO₂ conditions under which the diffusivities were measured complement the equilibrium-partitioning study of calcium in olivine-bearing basalts by Jurewicz and Watson, 1988. Olivine was found to be anisotropic with respect to the diffusion of calcium, iron, magnesium and manganese. This anisotropy is a weak function of temperature, but strongly dependent upon oxygen fugacity. Because diffusion is independent of olivine composition over the small range of compositions used in this study, it could be shown that the absolute values of the diffusion coefficients were also functions of temperature and fO₂. At near-atmospheric total pressure and an oxygen fugacity of 10^-8atm, D_Fe>D_Mn>D_Ca and D_Mg≥D_Mn for a range of geologically reasonable temperatures. These relative diffusivities were shown to change with oxygen fugacity. The power-law dependence of diffusion on oxygen partial-pressure was determined for each cation and the results are consistent with the range of values given by Stocker (1978) and by other workers. For Ca and Fe, the effect of hydrostatic pressure on diffusion appears to be weak, at least for transport parallel to the c crystallographic direction. Unfortunately, no true activation volumes (or other pressure-related parameters) could be computed because the oxygen fugacity was not held constant over changes in pressure, and because accurate post-experiment reconstruction of sample orientation was not possible. Al was found to enter high-pressure olivines at concentrations of up to 0.14 weight percent, thus allowing aluminum diffusion to be characterized. The diffusivity of aluminum is, within error, the same as iron at 20 kb at 1430° C at the ambient fO₂ of our piston-cylinder cells. This correspondence suggests that diffusion of Al may depend on transport of either Fe or of Fe ⁺³ defects. While the results of these experiments are generally consistent with results published elsewhere, there are important inconsistencies. Tracer diffusion and interdiffusion in pure, ordered, olivine endmembers (e.g., tephroite and forsterite) showed significantly higher activation energies. This discrepancy could reflect the role of Fe⁺³ defects in diffusion; however, it may also suggest that order-disorder phenomena may be significant factors influencing diffusion in analog systems. The results of this study are applied to four petrologic problems: (1) calculation of rates of equilibration for olivine xenocrysts; (2) calculation of closure temperatures for the CaO/MgO olivine/basalt geothermometer (Jurewicz and Watson 1988); (3) delineation of an intrinsic-/O₂ geobarometer; and (4) investigation of the dependence of olivine dissolution upon crystallographic orientation. In addition, it is demonstrated that diffusion-exchange experiments are useful for studying the dominant point-defect mechanisms for cation diffusion.