V13A-3095
An improved geospeedometry using chemical-isotopic profiles in olivines
Abstract:
In-situ isotopic analyses can be used to identify diffusion-controlled zoning in minerals. Recent studies [1,2] show that in olivine, diffusion-driven chemical zoning is associated with negatively coupled Mg-Fe isotopic profiles. Crystallization in a slowly cooled magmatic body, on the other hand, produces no isotopic fractionation. A wealth of information can be obtained from chemical-isotopic profiles measured in olivines if the ß-exponent, as in D1/D2 = (m2/m1)^ß (where D is diffusivity and m is mass), can be experimentally determined. For this purpose, we have performed diffusion experiments at 1,400 °C using crystallographically oriented Fo90-Fo100 crystals.We will present the experiments as well as modeling results to show how at a given ß-exponent, isotopic fractionations are affected by crystal growth rates and boundary conditions (e.g. changes in melt composition driven by changes in temperature) used in the model. When the Péclet number (νR/D, where ν is the crystal growth rate, R is radius and D is diffusivity) is less than 100, crystal growth does not affect isotopic fractionations significantly. As for the effect of boundary conditions, isotopic fractionation increases with increasing compositional contrast between olivine and melt. This means that isotopic fractionations may be used to constrain boundary conditions, which in turn affect calculated cooling timescales.
We will demonstrate the virtue of modeling both chemical and isotopic profiles in a zoned olivine using a natural sample. While a large number of randomly generated time-temperature histories can produce good fits to the measured Mg-Fe chemical profile, only a few can produce acceptable fits to the measured Mg-Fe isotopic profiles in the zoned olivine. Thus, by modeling both chemical and isotopic profiles in zoned minerals, thermal histories of magmatic bodies may be determined with a new level of confidence.
[1] Sio et al. (2013) GCA 123, 302-321 [2] Oeser et al. (2015) GCA 154, 130-150.