A Reconstitution Approach for Whole Rock Major and Trace Element Compositions of Granulites from the Kapuskasing Structural Zone

Abstract

Current estimates for the composition of the lower continental crust show significant variation for the concentrations of the highly incompatible elements, including large uncertainties for the heat-producing elements. This has consequences for models of the formation of lower crust. For example, is lower continental crust inherently poor in incompatible elements or has it become so after extraction of partial melts caused by thermal incubation? Answering these questions will require better agreement between estimates for the chemistry of the lower crust. One issue is that granulite samples may have been altered during ascent. Xenoliths often experience contamination from the entraining alkaline magma, potentially resulting in elevated concentrations of incompatible trace elements when analysed by conventional bulk rock techniques. To avoid this, we assessed an in situ approach for reconstructing whole rock compositions with granulites from the Kapuskasing Structural Zone, Superior Province, Canada. As terrain samples, they have not been affected by host magma contamination, and as subrecent glacial exposures, they show minimal modern weathering. We used scanning electron microscope electron dispersive spectroscopy (SEM-EDS) phase mapping to establish the modal mineralogy. Major and trace element concentrations of mineral phases were determined by electron microprobe and laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS), respectively. These concentrations were combined with the modal mineralogies to obtain reconstructed whole rock compositions, which were compared to conventional bulk rock analyses. The reconstructed data show good reproducibility relative to the conventional analyses for samples with massive textures. However, the conventional bulk rock chemistry systematically yields higher K concentrations, which are hosted in altered feldspars. Thus, even in terrain samples, minor alteration can lead to elevated incompatible element estimates that may not represent genuine lower continental crust.

Using the traditional microscope for mineral grain orientation determination: A prototype image analysis pipeline for optic‐axis mapping (POAM)

“…In the quiver map, plagioclase was manually inserted (stippled yellow contour) for comparison. Sample: 6KB‐67 75 …”

“…40% each plagioclase and hornblende, ca. 7% each of garnet, biotite and quartz with accessory titanite, apatite, pyrite, calcite and chlorite 75 . The sample is banded on the centimetre‐scale (Figure 10), alternating between well‐equilibrated mafic (hornblende, garnet, biotite) mineral bands and more felsic domains (plagioclase ± quartz).…”

“…Sample: 6KB-67. 75 The orientation map (bottom centre) shows a 6-mm garnet with internal orientation zonation from core to rim with two intermediate areas of lesser inclination (greener hue and highest saturation). There is a strong azimuth variation around the staurolite and biotite crystals.…”

“…7% each of garnet, biotite and quartz with accessory titanite, apatite, pyrite, calcite and chlorite. 75 The sample is banded on the centimetre-scale (Figure 10), alternating between wellequilibrated mafic (hornblende, garnet, biotite) mineral bands and more felsic domains (plagioclase ± quartz). In Figure 10, segmentation (yellow circle in WSI) was achieved for a single FoV, from optical multipol stack information alone.…”