Whole-rock Mg isotopes distinguishing high- and low-temperature S-type granites

Abstract

Determining the temperature of crustal anatexis associated with granitic magmatism is vital for understanding the petrogenesis of granites and the processes underlying crustal anatexis. During the partial melting of metasedimentary rocks, the breakdown of biotite at elevated temperatures can lead to the formation of peritectic garnet, which is typically rare or absent at lower temperatures when melting is predominantly driven by muscovite breakdown. This distinction makes magnesium (Mg) isotopes a valuable tool for tracing relative variations in crustal melting temperatures, as garnet exhibits the lowest δ26Mg values compared to other Mg-bearing phases under equilibrium fractionation. However, the extent to which temperature variations influence melt Mg isotope compositions remains inadequately understood. In this study, we present Mg and oxygen isotope data for high- and low-temperature leucogranites from the Himalayan orogen, alongside whole-rock major and trace element data, and zircon Ti content from previous studies. Low-temperature leucogranites, with maximum Ti-in-zircon temperatures ranging from 739 to 801 °C, display negative δ26Mg values between −0.70 and −0.14 ‰, aligning with most global S-type granites. Conversely, high-temperature leucogranites, exhibiting maximum Ti-in-zircon temperatures of 800 to 855 °C, possess positive δ26Mg values ranging from 0.46 to 0.53 ‰, significantly exceeding those of Himalayan metasedimentary rocks and most global S-type granites. The high-temperature leucogranites also demonstrate relatively elevated Nb/Ta and Eu/Eu* ratios, while their whole-rock δ18O and CIA values are comparable to those of low-temperature leucogranites. These findings suggest that fluid alteration, fractional crystallization, and crustal assimilation have a minimal impact on the Mg isotope discrepancies between the two granite groups. We propose that the positive δ26Mg values of the high-temperature leucogranites result from a greater fraction of peritectic garnet in the melting residue, attributable to higher anatectic temperatures. Phase equilibrium modeling indicates that the peritectic reaction involving muscovite breakdown is swiftly replaced by biotite breakdown as temperatures rise, leading to a steady increase in the corresponding fraction of garnet. Equilibrium fractionation calculations further corroborate that melt δ26Mg values rise with an increasing fraction of peritectic garnet, with values at high temperatures significantly surpassing those at lower temperatures. Consequently, this study elucidates a causal relationship between Mg isotope composition and anatectic temperature, highlighting that whole-rock Mg isotopes can serve as a reliable indicator for distinguishing between high- and low-temperature S-type granites.