Reconstructing primary mineral chemistry from inclusion microstructures using EPMA

Dr. Dirk Spengler, Institute of Geosciences, University of Potsdam

Abstract

Crystallographically oriented mineral inclusions in high-grade rocks are often interpreted as products of solid-state exsolution from a once chemically homogeneous precursor phase. This textural unmixing occurs in various solid-solution systems, e.g. K-Na feldspar, garnet-pyroxene and Fe-Ti oxides, typically in response to cooling. Such microstructures preserve information that is otherwise lost to late-stage metamorphic re-equilibration, and chemical and deformational overprinting. They therefore form a key archive of the most extreme part of a rock's metamorphic history, and reconstructing the primary mineral chemistry is essential to unravelling it.

I will first show, using well-equilibrated ultramafic rocks from the Western Gneiss Region (WGR), Norway, containing rods of pyroxene in garnet, how chemical integration of the mineral microstructure by EPMA, using a defocused beam, compares with a 2D image-quantification approach. Both methods yield reproducible results, and for small regions of interest (ROIs) their means overlap within error, though EPMA carries larger uncertainties than the 2D image approach. However, small ROIs are insufficient to ensure accuracy, as they remain subject to sampling error. Achieving adequate accuracy therefore requires a sufficiently large ROI, a requirement that EPMA can rarely meet within a justifiable analytical time. The 2D image approach can cover much larger areas efficiently, making it the more practical method for well-equilibrated, unzoned samples. These results are combined with experimental data and thermodynamic datasets to address contrasting interpretations of the possible origin of the microstructure and its geodynamic implications.

In a second, more complex example, WGR mafic rocks contain rods of quartz that are either fresh, partially reacted to albite or intermingled with pargasite within clinopyroxene. In this case, the integrative EPMA approach offers a distinct advantage over the 2D image method. Because the accuracy of individual analyses is difficult to assess in isolation, mineral-chemical systematics between reconstructed and present-day pyroxene compositions are used instead to test whether the inclusion microstructure originated in situ or through an allochemical process, with implications for the metamorphic history.