On the cratonization of the Arabian-Nubian Shield: Constraints from gneissic granitoids in south Eastern Desert, Egypt

International audience; The Shaitian granite complex (SGC) spans more than 80 Ma of crustal growth in the Arabian–Nubian Shield in southeast Egypt. It is a voluminous composite intrusion (60 km2) comprising a host tonalite massif intruded by subordinate dyke-like masses of trondhjemite, granodiorite and monzogranite. The host tonalite, in turn, encloses several, fine-grained amphibolite enclaves. U-Pb zircon dating indicates a wide range of crystallization ages within the SGC (800 ± 18 Ma for tonalites; 754 ± 3.9 Ma for trondhjemite; 738 ± 3.8 Ma for granodirite; and 717±3.2 Ma for monzogranite), suggesting crystallization of independent magma pulses. The high positive εNdi (+6–+8) indicate that the melting sources were dominated by juvenile material without any significant input from older crust. Application of zircon saturation geothermometry indicates increasing temperatures during the generation of melts from 745±31 °C for tonalite to 810±25 °C for trondhjemite; 840±10 °C for granodiorite; and 868±10 °C for monzogranite. The pressure of partial melting is loosely constrained to be below the stability of residual garnet (3 kbar for the stability of residual amphibole as inferred from the significantly lower NbN and TaN compared with LREEN and the sub-chondrite Nb/Ta ratios exhibited by the granitic phases. The inverse relation between the generation temperatures and the ages estimates of the granitoid lithologies argue against a significant role of fractional crystallization. The major and trace element contents indicate the emplacement of SGC within a subduction zone setting. It lacks distinctive features for melt derived from a subducted slab (e.g. high Sr/Y and high (La/Yb)N ratios), and the relatively low MgO and Ni contents in all granite phases within the SGC suggest melting within the lower crust of an island arc overlying a mantlewedge. Comparisonwith melts produced during melting experiments indicates an amphibolite of basaltic composition is the best candidate as source for the tonalite, trondhjemite and granodiorite magmas whereas the monzogranite magma is most consistent with fusion of a tonalite protolith. Given the overlapping Nd-Sr isotope ratios as well as several trace element ratios betweenmonzogranite and tonalite samples, it is reasonable to suggest that the renewed basaltic underplating may have caused partial melting of tonalite and the emplacement of monzogranite melt within the SGC. The emplacement of potassic granite (monzogranite) melts subsequent to the emplacement of Na-rich granites (tonalite-trondhjemite-granodiorite) most likely suggests major crustal thickening following arc collision andamalgamation into the over thickened proto-crust of the Arabian-Nubian microcontinent. Eventually, after complete consolidation, the whole SGC was subjected to regional deformation, most probably during accretion to the Saharan Metacraton (arc–continent collisions) in the late Cryogenian -Ediacaran times (650–542 Ma).