Granite petrogenesis and the δ44Ca of continental crust

Abstract

Stable Ca isotopes are an increasingly useful tool for understanding the sources and processes leading to the formation of magmatic rocks, yet Ca isotope fractionation during genesis of silicic continental crust is still poorly understood. Here, we present Ca, Sr, and Nd isotope, as well as major- and trace-element whole-rock geochemical data for A-, I-, and S-type granites (n = 30) from Australia/Tasmania, Canada, and France (δ44CaBSE of -0.6‰ to +0.2‰) and compare them to phase-equilibrium models for partial-melting (pelite, greywacke, MORB, enriched Archean tholeiite) and crystallization (hydrous arc basalt, A-type granite) that incorporate novel ab-initio predictions for Ca isotope fractionation in epidote and K-feldspar. The ab-initio calculations predict that epidote has similar δ44Ca to anorthite and that K-feldspar is the isotopically lightest known silicate mineral at equilibrium (Δ44Cakspar-melt of -0.4‰ at 1000 K). Our phase-equilibrium model results suggest that δ44Ca variations in all three granite types can be fully explained through magmatic processes, without necessarily requiring addition of isotopically exotic components (e.g., carbonate sediments). Heavy Ca isotope enrichments in A-type granites from the Lachlan Fold Belt, however, require isotopic disequilibrium between plagioclase and melt, which we use to constrain average plagioclase growth rates in these systems. This also serves to illustrate that whole-rock Ca isotope measurements can be used to estimate crystal growth rates, even in the absence of analyzable phenocrysts. In general, low Ca diffusivities and strong isotopic diffusivity ratios (D44/D40) in low-H2O granitic magmas should lead to resolvable isotopic disequilibrium effects in plagioclase, even at relatively slow growth rates (e.g., > 0.03 cm/yr). Combining our data with those from previous studies, we demonstrate that average granitoids and upper continental crust (with newly estimated δ44CaBSE of -0.25 ± 0.02‰, 2SE) have resolvable low δ44Ca compared to basalts and oceanic crust. Given that pressure has a major influence on Ca isotope fractionation across all of our models, this implies that melts feeding upper crustal granitoids dominantly evolve in the lower crust (10-14 kbar, through either partial-melting or fractional crystallization). This observation also suggests that heavier Ca isotopes are preferentially recycled back into the mantle through subduction and/or lower-crustal delamination events, but this is unlikely to have had a significant influence on the δ44Ca evolution of the upper mantle through geologic time.