Transition metal element concentrations in coexisting orthopyroxenes (opx) and clinopyroxenes (cpx) from various mantle xenoliths were measured using the electron- and
ion-microprobe. The grt-peridotite, sp-peridotite and grt-pyroxenite xenoliths from the East African Rift and Massif Central (France) studied equilibrated over a broad range of temperature and pressure, between 800° and 1400°C and
10-40 kbar, respectively. Major- and trace-elements show
no zonation from core to rim within the minerals and no
variations are observed between clinopyroxenes and between orthopyroxenes from the same xenolith. We present new data on the trace element partitioning between opx and cpx and evaluate their use for geothermobarometry.
It has been suggested that partitioning of trace elements in mafic- and ultramafic rocks are strongly temperature or pressure dependent. Previous studies (Hervig et al., 1986; Stosch, 1987) have shown, for example, that Sc-partitioning between olivine and cpx and between opx and cpx is temperature dependent, and that this partitioning can be used as a geothermometer. Aside from temperature, there are several other physical and chemical parameters such as valence state, ionic radius, size of crystal sites, crystal field stabilization energy (CFSE), bulk composition and coupled substitution, that can influence trace element partitioning. Strong compositional effects have been previously documented (Adam and Green, 1994; Lindstrom, 1976). Sc, for example, is preferentially incorporated into the M1 site of pyroxenes. The ability of M1- and M2-sites to accept different cations varies with major element composition and changing
PT-conditions. With increasing Al-component, for example, M1- and M2-sites decrease in size which subsequently controls the incorporation of trace elements (ionic radii relative to polyhedral volumes). Compositional variations namely in AlIV exhibit good correlations with Dopx/cpx for trivalent trace elements (M3+), suggesting coupled substitution as a major control. Na-content in opx and cpx varies strongly with bulk composition and has been shown to be a function of temperature and pressure (Brey and Köhler, 1990). Coupled substitution of M3+ cations on the M1 site and Na on M2 is likely too. Substitution of such a component (M3+NaSi2O6) would be much more restricted in opx in comparison to cpx.
To evaluate the usefulness of trace element partitioning in geothermobarometry it is necessary to test for possible controlling influences other than pressure and temperature. At constant temperature and pressure opx/cpx partition coefficients for transition metals increase from trivalent to divalent cations with DNi „ DCo > DMn > DCr „ DV > DTi „ DSc. With increasing temperature DM3+ approaches 1. Our data for Sc, V and Cr show very good positive correlation with temperature, while Ti shows a considerable spread. Mn, Co and Ni exhibit a broad negative correlation with temperature. DM3+ show no dependencies when plotted as a function of pressure. However, there seems to be a trend to higher D's for high pressure garnet-peridotites. It should be noted that natural samples have equilibrated along geothermal gradients which means high temperatures tend to correspond to high pressures and vice versa. In this case, it is difficult to isolate the effects of pressure and temperature. However, some samples which have equilibrated at similar pressure conditions show large differences in their D's, suggesting no pressure control.
The approximate co-variation of Al-content in opx and cpx as a function of pressure and/or temperature suggests that changing Al-content does not influence the partition coefficients significantly. The role of Na is not entirely resolved at this point.
Our new data on natural xenoliths suggest that within defined bulk compositions, e.g. spinel- or garnet- peridotites, DM3+opx/cpx are dominantly a function of temperature. DSc, DV and DCr correlate excellently with temperature and it is suggested that these transition metal elements are especially suitable for the use in geothermometry. However, there remains a need to test the empirical calibrations with high pressure experiments under controlled P-, T-, X- and fO2- conditions.
Adam, J. & Green, T.H., Chem. Geol. 117, 219-233 (1994).
Brey, G.P. & Köhler, T., Jour. Petrol. 31, 1353-1378 (1990).
Hervig, R.L., Smith, J.V. & Dawson, J.B., Trans. Roy. Soc. Edinburgh: Earth Sci. 77, 181-201 (1986).
Lindstrom, D.J., Ph.D. dissertation, University of Oregon, Eugene, OR (1976).
Stosch, H.-G., Fortsch. Miner. 65, 49-86 (1987).