Geochemistry and geochronology of Midyan terrane granitoids, NW Saudi Arabia: Implications for growth of the Arabian-Nubian Shield

 The 900–550 Ma Arabian-Nubian Shield (ANS) represents the northern part of the East African Orogen (EAO) and was generated by Neoproterozoic juvenile crust addition and ultimately, collision of eastern and western Gondwana ca. 600 Ma. The ANS  encompasses the greatest volume of Neoproterozoic juvenile crust preserved on Earth and embodies over 610,000 km2 across NE Africa and the western part of the Arabian Peninsula. The ANS comprises stacks of thin-skinned nappes resulting from oblique convergence of bounding plates and resulting amalgamation of intra-oceanic arcs generated within the Mozambique Ocean. As such, the ANS is a classic example of an accretionary orogen. Final accretion and ANS consolidation were accompanied and followed by emplacement of within-plate alkaline plutons c. 640–550 Ma, which represent the largest volume of alkaline granitoids on the planet. Overall, ANS tectonic evolution encompasses an orogenic cycle beginning with the fragmentation of Rodinia (870–800 Ma) and ending with amalgamation of eastern and western Gondwana in the Cambrian. This study combines field observations, petrography, whole rock major-and-trace element chemistry, along with U-Pb zircon geochronology and Sr-Nd-Pb isotope data of four granitoid suites of the northwestern part of the Midyan terrane (Ifal, Muwylih, Midyan, and Lawaz), that provides an excellent opportunity for further understanding of the factors influencing the shift from A- to I-type granitoids. Samples are mostly granite-granodiorite-diorite with subordinate gabbro and gabbro diorite. Seven samples are A-type, while the remaining 26 are I-type. Petrographically, I-type rocks consist of K-feldspar, quartz, albite, mica, amphiboles, and sodic-pyroxene as major minerals with a variety of accessory minerals, including Fe-oxides and zircon. Major oxide abundances such as CaO, TiO2, and P2O5 manifest a clear decrease with increasing SiO2 content, except for K2O, which indicates fractionation of, for example, plagioclase, Fe-Ti oxides, and apatite. Whole rock data shows clearly distinct characters of the A-type being metaluminous, enrchied in ∑REE with an average of 355 µg/g. In contrast, the I-type is meta-luminous-peraluminous, which only has an average of ∑REE 197 µg/g. The chondrite-normalized patterns of the A-type show lower negative Eu anomalies (Eu/Eu*=0.61) in comparison to the I-types (Eu/Eu*=0.86), depicting the different degrees of plagioclase fractionation. The primitive-mantle normalized patterns show depletions in  Sr and Ti, indicating fractionation of feldspar and Fe-Ti phases in both types. The U-Pb zircon ages indicate two distinct pulses of granitoid magmatism.  The first pulse is attributed to A2 (post collision)-type ca. 634–640 Ma, followed by a second pulse of I-type in nature ca. 618–598 Ma. The A and I-type Sr-Nd isotope data attest to the juvenile nature of the crust from the ANS (εNd =4.88–5.15), while some I-types (εNd =3.91) are derived from partial melting of a pre-existing crust. We conclude that the magmatism in the Midyan terrane area is shown to have shifted from a within-plate setting associated with orogenic collapse and partial melting of the lower crust into volcanic arc affinity with a more depleted source.