Petrologists are generally faced with the problem of estimating the Fe3+/Fetot-ratios for mantle minerals. calculation of Fe3+-contents from microprobe analyses is always problematic, whereas Moessbauer spectroscopy seems to be a good method to evaluate ferric iron. This problem is especially significant in mantle clinopyroxenites and hornblendites which may contain substantial amounts of Fe3+. Here, we have investigated the effect of Fe3+-treatment on widely used geothermobarometers. Further, we use Moessbauer data on the Fe3+ content in amphibole to estimate the fH2-values in amphibole assemblages, following Popp et al. (1995). The investigated xenoliths come from the Bayuda Main Volcaninc Field/Sudan and consist of Gt-pyroxenites, Gt-websterites and Gt-hornblendites, occurring together with Ol-Sp-pyroxenites, amphibole-bearing pyroxenites and Sp-lherzolites (Zinngrebe, 1996). Subsolidus reactions such as exsolutions of garnet and ilmenite in Cpx or recrystallized Cpx are developed in some samples, and spinel-to-garnet reactions are observed. The garnet bearing pyroxenites contain Al-rich augite (Mg#: 0.80-0.84), Opx (Mg-# 0.80-0.78), garnet (Mg# 0.60-0.62), hercynite-rich spinel, ilmenite with 4-6 % MgO and some amphibole (titanina pargasite with very low Si content). Kelyphitic rims around the garnets and amphibole are conspicuous. They consist of Al-rich Opx, euhedral green spinel and plagioclase (An80-90) at the garnets and of euhedral Ol (Mg# 0.70-0.65), titanian Cpx, Ti-Mgt and plagioclase (An70-65) at the amphiboles. The rims of Cpx show enrichments of Si and Mg and a depletion of Al and Ti.
Gt-Cpx, Cpx-Opx and Gt-Opx Fe/Mg exchange reactions have been used to estimate the equilibrium P and T conditions for the garnet pyroxenites. However, high contents of Fe3+ in Cpx decrease calculated temperatures. If the Fe3+ content is neglected (Fe3+=0), the thermobarometers yield pressures of 13 to 15 kbar at temperatures of 1150 °C which are inconsistent with the presence of primary amphibole. Calculated Fe3+ contents from microprobe analyses (Fe3+/Fetot=0.27) yield pressures between 11 and 13 kbar at temperatures of 980 °C. To determine the exact P-T conditions for the equilibration of the xenoliths, the Fe3+ content in Cpx of the garnet pyroxenites has been analyzed by 57Fe Moessbauer spectroscopy. A surprisingly high Fe3+/Fetot ratio of 0.50 results for Cpx. This value leads to temperatures between 820 and 870 °C and low pressures of 8.5 kbar. Thus Fe3+ contents cannot be neglected in using thermobarometers for mantle assemblages. It should be noted that the temperature inferred from the erroneous Fe3+=0 assumption appears realistic at first sight. Thus, errors in xenolith thermometry associated with neglect of Fe3+ may often pass unnoticed. Lower temperatures derived from Moessbauer spectroscopy data are consistent with oberserved metamorphic textures.
The fH2 values are calculated from the Fe-oxy-component in amphiboles in the different xenolith types. However, the application of this fH2-barometer is critically dependent on correct Fe3+ analyses in amphibole. Preliminary Moessbauer data on amphibole from one sample (Fe3+/ Fetot =0.40) seem to confirm high Fe3+ contents from microprobe analyses (Fe3+/ Fetot =0.4). As Fe3+/ Fetot in amphiboles from different samples, calculated from microprobe analyses, varies from 0 to 0.6, there appear to exist large differences in fH2 (up to 3 log units) in this mantle cumulate suite. ar to exist large differences in fH2 (up to 3 log units) in this mantle cumulate suite.
Popp, R. K., et al., Am. Miner. 80, 534-548 (1995).
Zinngrebe, E., J. Conf. Abs. 1, (1996).