2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 6
Presentation Time: 9:35 AM

P-T PSEUDOSECTION MODELING OF MINERAL ASSEMBLAGE STABILITY AND COMPOSITIONS FOR A MINERALOGICALLY SIMPLE PELITE: THE IMPORTANCE OF FINDING THE EFFECTIVE BULK COMPOSITION


TINKHAM, Douglas K., Geology and Geophysics, Univ of Calgary, 2500 University Dr NW, Calgary, AB T2N 1N4 and GHENT, Edward D., Univ Calgary, 2500 University Dr NW, Calgary, AB T2N 1N4, tinkham@ucalgary.ca

The use of P-T pseudosections for modeling the theoretical equilibrium assemblage stability for a bulk composition to derive P-T conditions of metamorphism has increased recently. Although the comparison of predicted assemblage stability with observed assemblages is common, detailed comparisons of predicted phase compositions are not common. A pelitic gneiss from Mica Creek, British Columbia, is ideally suited to test the ability of pseudosections to model assemblage stability and phase compositions. The sample contains muscovite – biotite – kyanite – plagioclase – quartz – apatite – pyrite – rutile - graphite, all phases are compositionally homogeneous, and compositional layering is absent.

Pseudosections constructed in the chemical system MnO-Na2O-CaO-K2O-FeO-MgO-Al2O3-SiO2-H2O-TiO2 (MnNCKFMASHT) with rock compositions determined by X-ray fluorescence analysis are used to compare predicted assemblage stability and phase compositions with those of the sample. Garnet stability is predicted, but not observed, at the conditions of metamorphism. Removal of MnO from the system does not resolve the discrepancy. However, removal of iron associated with pyrite yields the observed assemblage at the estimated P-T conditions of metamorphism (7-8 kilobars, 650 – 670 C).

The agreement between predicted phase compositions and measured phase compositions ranges from good to poor. Agreement between plagioclase compositions is good ( Ab83 vs. Ab82 observed). Predicted muscovite compositions contain too much paragonite (Pg) content ( Pg20 vs. Pg12 observed). Biotite compositions are fair (0.36 octahedral aluminum vs. 0.44 observed; 0.010 Mn atoms vs. 0.008 atoms observed; Fe/(Fe+Mg)=0.49 vs. 0.48 observed).

These results highlight the importance of using an effective bulk composition to model an individual sample. An average bulk composition from a suite of pelites from the Mica Creek region would not predict the proper assemblage for many pelites, and should not be used for deriving detailed P-T paths. However, adjusting rock compositions for the presence of phases not modeled and accounting for the effect of chemically zoned phases on rock composition allows accurate modeling in the MnNCKFMASHT system, although there are some problems with predicted white mica compositions.