New thermodynamic models and revised calibrations for the Ti-in-zircon and Zr-in-rutile thermometers

The models recognize that ZrSiO4, ZrTiO4, and TiSiO4, but not ZrO2 or TiO2, are independently variable phase components in zircon. Accordingly, the equilibrium controlling the Zr content of rutile coexisting with zircon is ZrSiO4 = ZrO2 ( in rutile) + SiO2. The equilibrium controlling the Ti content of zircon is either ZrSiO4 + TiO2 = ZrTiO4 + SiO2 or TiO2 + SiO2 = TiSiO4, depending whether Ti substitutes for Si or Zr. The Zr content of rutile thus depends on the activity of SiO2 (a(SiO2)) as well as T, and the Ti content of zircon depends on a(SiO2) and a(TiO2) as well as T. New and published experimental data confirm the predicted increase in the Zr content of rutile with decreasing a(SiO2); and unequivocally demonstrate that the Ti content of zircon increases with decreasing a(SiO2). The substitution of Ti in zircon therefore is primarily for Si. Assuming a constant effect of P, unit a(ZrSiO4); and that a(ZrO2) and a(ZrTiO4) are proportional to ppm Zr in rutile and ppm Ti in zircon, [ log( ppm Zr-in-rutile) + loga(SiO2)] = A(1) + B-1/T( K) and [ log( ppm Ti-in-zircon) + loga(SiO2)-loga(TiO2)] = A(2) + B-2/T, where the A and B are constants. The constants were derived from published and new data from experiments with aSiO2 buffered by either quartz or zircon + zirconia, from experiments with aSiO2 defined by the Zr content of rutile, and from well-characterized natural samples. Results are A1 = 7.420 +/- 0.105; B-1 = - 4,530 +/- 111; A(2) = 5.711 +/- 0.072; B-2 = - 4,800 +/- 86 with activity referenced to alpha-quartz and rutile at P and T of interest. The zircon thermometer may now be applied to rocks without quartz and/or rutile, and the rutile thermometer applied to rocks without quartz, provided that a(SiO2) and a(TiO2) are estimated. Maximum uncertainties introduced to zircon and rutile thermometry by unconstrained aSiO2 and aTiO2 can be quantitatively assessed and are approximate to 60 to 70 degrees C at 750 degrees C. A preliminary assessment of the dependence of the two thermometers on P predicts that an uncertainty of +/- 1 GPa introduces an additional uncertainty at 750 degrees C of approximate to 50 degrees C for the Ti-in-zircon thermometer and of approximate to 70 to 80 degrees C for the Zr-in-rutile thermometer.