Diffusion anisotropy of Ti in zircon and implications for Ti-in-zircon thermometry

Abstract

Ti-in-zircon thermometry has become a widely used tool to determine zircon crystallization temperatures, in part due to reports of extremely sluggish Ti diffusion perpendicular to the crystallographic c-axis in this mineral. We have conducted Ti-in-zircon diffusion experiments, focusing on diffusion parallel to the c-axis, at 1 atm pressure between 1100 and 1540°C, with oxygen fugacities equivalent to air and the Ni-NiO buffer. There is no resolvable dependence of Ti diffusion in zircon upon silica or zirconia activity, or upon oxygen fugacity. The diffusion coefficient of Ti in zircon is found to be a weak function of its own concentration, spanning less than 0.5 log units across any profile induced below 1300°C. Ti diffusion in zircon, parallel to the c-axis at 1 atm pressure, is well described using:log10⁡DTi=[1.34(±1.44)−555425(±44820)Jmol−12.303RT(K)]m2s−1 where R is the gas constant in J/(mol⋅K). In conjunction with diffusion coefficients for Ti in zircon perpendicular to the c-axis reported by Cherniak and Watson (2007), strong diffusion anisotropy for Ti in zircon is observed. Diffusion parallel to the c-axis is ∼4-5 orders of magnitude faster than diffusion perpendicular to the c-axis within the experimentally constrained temperature range shared between these two studies (1540-1350°C). This difference increases if the data are extrapolated to lower temperatures and reaches ∼7.5-11 orders of magnitude between 950-600 °C, a typical range for zircon crystallization. Diffusion of Ti in natural zircons will predominantly occur parallel to the c-axis, and the Ti-in-zircon thermometer appears susceptible to diffusive modification under some crustal conditions. Temperatures calculated using this system should therefore be evaluated on a case-by-case basis, particularly when considering high-T, slowly cooled, reheated and/or small zircons.