Iron isotope fractionation during partial melting of metapelites and the generation of strongly peraluminous granites

The large variability in Fe isotope ratios of sedimentary rocks (particularly those from the Archean and Proterozoic) contrasts with that of igneous rocks, which display a much more limited range in values. Notably, among igneous rocks, those inferred to form via partial melting of siliciclastic sediments – strongly peraluminous granites (SPGs) – exhibit greater variability in their Fe isotope compositions, suggesting SPGs may capture isotopic variations in the sedimentary record. However, the extent and mechanisms of iron isotope fractionation between SPGs and their source remain poorly understood. Our study integrates iron isotope analyses with petrological modeling to investigate equilibrium isotopic fractionation during generation of SPG magmas. As a case study, we focus on the Neoarchean Ghost Lake Batholith and the adjacent metasedimentary rocks in Ontario, Canada. These units represent an internally differentiated SPG batholith and metamorphosed sedimentary rocks interpreted as the source of the batholith.

We measured δ⁵⁶Fe compositions of SPG samples, metasedimentary rocks, and a restitic rock. Sulfide grains were also measured in four metapelite samples and a granite sample. We find no correlation between the δ⁵⁶Fe composition of metasedimentary rocks and their metamorphic grade, indicating iron isotopes behave as a closed system during metamorphism. Modeling results show that iron isotopes in SPGs from the Ghost Lake batholith are consistent with equilibrium fractionation during biotite dehydration melting, with predicted δ⁵⁶Fe values for melts and restitic assemblages mainly controlled by the source composition. Our results predict negligible isotopic fractionation between the residue and the source, whereas ∼0.177–0.277 ‰ is expected between SPG melts and the residue, accounting for high δ⁵⁶Fe values in granite samples. Lower δ⁵⁶Fe values may indicate that some granites represent mixtures of melt and cumulus material or result from assimilation of restite or source/host rock. However, despite deviations from pure equilibrium fractionation, the variability in δ⁵⁶Fe values for SPGs is about one order of magnitude smaller than that seen in the sedimentary record for the Archean and Proterozoic (∼0.2 ‰ vs. >2 ‰). We posit that this narrower range of isotopic variation in SPGs results from metamorphism and partial melting, which can homogenize large isotopic variations in sedimentary protoliths. Thus, SPGs represent reliable archives for the bulk iron isotope evolution of siliciclastic sedimentary rocks through time.