REE fractionation and Nd-isotope disequilibrium during crustal anatexis: constraints from Himalayan leucogranites

Abstract

The processes controlling rare earth element (REE) behaviour during crustal anatexis have profound implications not only for REE abundances of crustally derived granites but also for their model Nd ages. For typical protoliths of anatectic granites, such as are exposed in metapelites from the High Himalayan Crystalline Series (HHCS), a high proportion (90%) of bulk-rock LREE and MREE are sited in the accessory phases apatite and monazite, and of HREE in garnet and apatite. Mass-balance constraints indicate that observed REE abundances of Himalayan leucogranites cannot be generated from REE-deficient reactant phases in their metasedimentary protoliths (muscovite, biotite, plagioclase and quartz) but must result from the dissolution of accessory phases. REE and Th abundances in anatectic melts are modelled by combining empirical expressions for monazite, apatite and zircon solubilities (as a function of temperature, melt composition and melt water content) with analysed accessory-phase REE abundances. This approach, which allows the contribution to the melt from each accessory phase to be identified, successfully predicts Himalayan leucogranite trace-element abundances. The LREE and Th concentrations of a metapelite-derived melt are buffered by monazite stability, although apatite dissolution contributes up to 10% of LREE to the melt together with ∼50% of the MREE and 90% of the HREE. Zircon contributes a negligible proportion of REE to the melt. During anatexis, dissolution of monazite and apatite in the melt results in: (1) increasing REE abundances and LREE/HREE ratios with increasing temperature of melting; (2) a negative Eu anomaly in the melt; (3) high Sm/Nd ratios in the melt relative to its metasedimentary protolith; and (4) Nd-isotope disequilibrium between the melt and its protolith. All of these features are observed in the REE geochemistry of Himalayan leucogranites. The combined effects of Sm/Nd fractionation and Nd-isotope disequilibrium during anatexis imply that model Nd ages calculated for anatectic granites derived from apatite and monazite bearing protoliths should be treated with considerable caution.