2006 Philadelphia Annual Meeting (22–25 October 2006)

Paper No. 3
Presentation Time: 2:05 PM

NEW THERMODYNAMIC ANALYSIS AND CALIBRATION OF THE TI-IN-ZIRCON AND ZR-IN-RUTILE GEOTHERMOMETERS


FERRY, John M., Department of Earth and Planetary Sciences, Johns Hopkins Univ, Baltimore, MD 21218 and WATSON, E. Bruce, Earth and Environmental Sciences, Rensselaer Polytechnic Institute, 110 8th St., JSC 1W19, Troy, NY 12180, jferry@jhu.edu

This new analysis recognizes that ZrSiO4, ZrTiO4, and TiSiO4, but not ZrO2 or TiO2, are independently variable phase components in zircon (Zrn). Accordingly, the equilibrium controlling the Zr content of rutile (Rt) coexisting with Zrn is (1) ZrSiO4 = ZrO2 (Rt) + SiO2; the equilibrium controlling the Ti content of Zrn is either (2) ZrSiO4 + TiO2 = ZrTiO4 + SiO2 or (3) TiO2 + SiO2 = TiSiO4, depending on whether Ti substitutes for Si or Zr. The Zr content of Rt thus depends on the activity of SiO2 (aSiO2) as well as T; the Ti content of Zrn depends on aSiO2 and aTiO2 as well as T. The prediction from Eq. (1) that the Zr content of Rt should increase with decreasing aSiO2 is confirmed by the lower Zr content of Rt saturated with quartz (Qtz) in Run QTi13 of [1], 1697±56 ppm, than that of Rt in Qtz-undersaturated Run 87a at the same P and T, 2411±176 ppm. A new experiment with Zrn + Rt + baddeleyite (Bd) at 1.0 GPa, 1050°C, and log(aSiO2) = -0.538 crystallized Zrn with 368±28 ppm Ti compared to 108±23 ppm in Zrn in Qtz-saturated Run 66 of [1] at the same P and T. Results quantitatively confirm both the dependence of Ti in Zrn on aSiO2 and the substitution of Ti for Si in Zrn specified by Eq. (2). Following [1] in neglecting the effect of P and assuming aZrTiO4 and aZrO2 are proportional to ppm Ti in Zrn and ppm Zr in Rt, log(ppm Zr in Rt) + log(aSiO2) = A1 + B1/T(K) and log(ppm Ti in Zrn) + log(aSiO2) - log(aTiO2) = A2 + B2/T, where A and B are constants. The constants were derived from those experiments from [1] and 5 new experiments with aSiO2 buffered by either Qtz or Zrn + Bd and from data for natural samples in [1] except the Labait harzburgite. Results are A1 = 7.283; B1 = -4405; A2 = 5.564; B2 = -4659 with activity referenced to α-Qtz and Rt. Data for almost all runs with unbuffered aSiO2 in [1] are consistent with these best-fit values for A and B. T may now be calculated for rocks without Qtz and/or Rt provided that aSiO2 and aTiO2 are estimated. Thermometry of out-of-context Zrns must consider reduced aSiO2 as well as reduced aTiO2. For example, for Zrn with Ti content corresponding to 750°C in Qtz- and Rt-saturated rocks, T = 687°C if aSiO2 = 0.5 and T = 822°C if aTiO2 = 0.5 (the latter a plausible minimum for Zrn-saturated granitoid melts [1]). Because the effects of reduced aSiO2 and aTiO2 compensate, T = 750°C if aSiO2 = aTiO2 = 0.5.

[1] Watson et al. (2006) Contrib. Mineral. Petrol., v. 151, p. 413-433.