Origin of cordierite-bearing granites by assimilation in the Central Iberian Massif (CIM), Spain

Abstract

The Central Iberian Massif (CIM) is characterized by the great abundance and variety of granitoids emplaced during Hercynian times. One of the most conspicuous is cordierite-bearing biotite granite (CG), which in the western part of the CIM is associated with other granites lacking cordierite [amphibole-bearing biotite granite (AG) and biotite granite (BG)] and with anatectic host-rocks [cordierite-rich nebulites (NB)]. Major-, trace- and rare-earth element data for AG and BG show that these granites have chemical affinities, and parallel or overlapping differentiation trends, suggesting two different granite magmas, with a common or similar deep-seated source. CG show two extreme facies: a marginal cordierite-rich facies (MCG) in direct contact with host-rocks (schists, migmatites), and an internal cordierite-poor facies (ICG) farther away from the host-rocks. Gradual transitions between MCG and ICG are commonly observed, as well as between ICG and BG. An anatectic hypothesis for CG is difficult to accept because: (a) CG have higher CaO contents relative to NB or other country rocks, a smaller Eu/Eu∗ ratio and similar Sr contents; and (b) the similar REE absolute contents and patterns in CG, BG and NB, and the absence of REE fractionation with respect to country rocks. Simple mixing equations are computed to test assimilation processes involving BG as a primary magma and NB as the contaminant in producing CG. CaO and MgO data have been used for this purpose since they are the most discriminating elements between BG and NB. The results of the computation indicate that: (a) assimilation is possible for nearly the whole CaO range in NB if CaO in the BG ranges between 1.91% and 2.86%, and MgO is between 0.93% and 1.18% (these extreme values are comprised within the observed range for CaO and MgO in BG); (b) CaO and MgO data for MCG are compatible with assimilation for nearly the whole range of data observed; and (c) the percentage of assimilation is < 30% if all data are considered, but 92% of the solutions can be explained by assimilation percentages of < 18% NB. These data make the assimilation model the one preferred by the authors.