PERITECTIC ASSEMBLAGE ENTRAINMENT: THE MAKING OF S- AND I-TYPE GRANITES
Partial crustal melt can escape its solid matrix carrying small crystals of peritectic phases formed in the melting reaction, in ratios decreed by the stoichiometry of the reactions. For elements with low solubilities in the melts, PAE, with co-entrainment of accessories, explains most of the primary variation in the magmas. Concentrations of elements with high solubilities in the melts reflect the protolith compositions in a simpler way.
For S-types, derived through melting of metasedimentary rocks, the entrainable peritectic assemblage (formed during breakdown of aluminous biotite) involves Pl, Grt and Ilm, with Opx in the less aluminous sources (e.g. greywackes). Isotope and geochemical evidence shows that I-types are not generally produced through fractionation of mantle-derived mafic magmas. Despite examples of spectacular magma mingling, the geochemistry of I-types is also incompatible with magma-mixing. Crustal igneous and volcano-sedimentary rocks, themselves not long separated from the mantle, are the likely protoliths. Melting reactions here involve simultaneous breakdown of biotite and hornblende, and the entrainable peritectic assemblages involve Pl, Cpx, Opx, Ilm and possibly Grt.
Pure melts of both S- and I-type protoliths are peraluminous. S-type magmas remain peraluminous or increase in ASI with PAE. I-type magmas normally only become metaluminous with increased entrainment of peritectic Cpx. Thus, S- and I-type rocks commonly display opposing trends on plots of ASI vs Mg+Fe.
As Bruce Chappell and Allan White impressed upon us, the protolith is the primary control on granite magma chemistry. However, granites do not image their sources in the simple way that they proposed. Rather, the bulk composition and mineralogy of the protolith dictates what will remain behind as restite, what will be available to dissolve in the melt and what will be formed as the entrainable peritectic assemblage.