Impact of accessory minerals on the distribution of trace elements in the continental crust: an integrated petrologic and phase equilibria modeling study of migmatites, The

Abstract

Partial melting of the lower continental crust and subsequent melt emplacement at higher structural levels is an important crustal differential process that is predicted to produce an enriched granitic upper crust and depleted, more mafic lower crust. However, the common occurrence of accessory mineral-rich migmatite residues and upper crustal granites that are depleted in incompatible elements (U, Th, and REE) suggests our understanding of the controls on element distribution during recycling is limited. The behavior of key accessory minerals during partial melting of metasedimentary rocks, is interpreted to impact the composition of melts generated and extracted from mid-crustal migmatites, and therefore has important implications for our understanding of the process of recycling in the continental crust. Two broad approaches were used in this research project to assess the behavior of accessory minerals in anatectic melts: a natural migmatite study, and modeling of theoretical melts generated through migmatization of metasedimentary rocks. Detailed in situ textural and mineral chemical characterization of migmatitic paragneiss from the Larsemann Hills, east Antarctica, provides insight into the behavior of migmatitic rocks during polycyclic metamorphism. Important trace element-bearing accessory minerals (zircon and monazite) record complex trace element zoning patterns that are linked with major mineral (garnet) growth and breakdown. The work provides insight into the utility of trace element analysis in the interpretation of complex age relationships and constrains the relative extent of melting during overprinting anatectic events. The extraction of melts generated during the first partial melting event, D1, left a predominantly residual composition, and the overprinting partial melting that occurred during D2 was localized to fertile domains. Results also suggest that the textural location and assemblage context affects accessory mineral growth and breakdown, and as a result affect the flux of trace and heat producing elements from the site of melting into the upper crust. Modeled S-type granitic melts calculated from thermodynamic phase equilibria show that pressure, temperature, and changes to melt composition (e.g. XH2O) affect the behavior of zircon in partial melts. Under equilibrium conditions zircon will be saturated in the melt at the site of melting. Slight decreases in XH2O in the melt and/or in temperature will facilitate zircon crystallization. This suggests that zircon may grow close to the site of melting in the presence of melt, and zircon in granulites does not necessarily reflect growth during cooling of the terrane below the solidus. The lower crust will be enriched in Zr and therefore zircon, which is supported by petrographic observations in some natural granulites. As a consequence of zircon enrichment, trace elements (REEs, HPEs) will also not be depleted in residual lower crustal rocks. Modeling also demonstrates that the major and accessory element compositions of melts that are extracted from the deep crust and emplaced at higher structural levels are highly dependent on melt segregation, extraction rate and interaction with wall rock.