DIFFERENTIATION IN THE LAYERED MAFIC INTRUSIONS: INSIGHTS FROM THE SONJU LAKE-FINLAND GRANOPHYRE SYSTEM (MN)

Layered mafic intrusions provide the textbook example of magma differentiation following mineral melt thermodynamic equilibrium. For instance phase appearance closely follows examples given by laboratory experiments. Whether differentiation reflects fractional crystallization involving crystal settling is less clear. The Sonju Lake Intrusion (SLI), a 1200 m thick layered mafic intrusion, and the Finland Granophyre(FG), an equal volume silicic intrusion that directly overlies it, provide a puzzle to standard models of igneous differentiation. Mineral modes, compositions and even incompatible trace element ratios show smooth gradation between the two bodies as if the two intrusions were genetically related. Yet, differences in Sr and Pb isotope ratios and simple volume relationships argue against any fractional crystallization-like process linking the intrusions. Lundstrom et al. (2011) provided an alternative model showing that SLI modes and compositional trends could be reproduced by a top down sill injection process of a uniform composition basalt with differentiation occurring by temperature gradient based diffusion-reaction process above each emplaced sill. Lundstrom and Gajos (2013) provide experiments and a model showing that the observed PGE-rich horizon in the SLI can form by the same top down process. Testable predictions of this model include non-traditional stable isotope analysis and age dating. I will present d56Fe for 40+ samples showing that heavier d56Fe occurs as the top of the SLI is approached and lower d56Fe occurs as the bottom of the SLI is approached—both are consistent with the prediction of the thermal gradients occurring during the top down sill emplacement process. Furthermore, Lu-Hf in 2 apatite separates and a WR analysis were measured to produce an isochron for the apatite gabbro in the upper SLI previously dated by Paces and Miller (1992) at 1096 Ga. The apatite-WR produces a low MSWD isochron age of 1088 +/- 1 Myrs. The age offset is unlikely to reflect analytical issues with either analysis. Instead, it is consistent with slow top down formation of the SLI with the 1088 age representing the shutting off of ascending reactive fluids which reset the apatite age.