Middle to Late Miocene K-rich magmatism in Central Iran: Geochemical characterization of the post-collision mantle beneath the Urumieh–Dokhtar magmatic arc

Abstract

High-K magmatism of the central Urumieh–Dokhtar magmatic arc (UDMA) documents the final stages of Neo-Tethyan Ocean subduction. The temporal and spatial development of collision-related magmatism in this region remains ambiguous because of the paucity of systematic age and radiogenic isotopic data. Here, we revisit pre-final collisional (∼24 Ma; low-K calc-alkaline) and post-collisional (∼15–10 Ma; high-K) magmatic rocks and dikes from the central UDMA; the latter constitute the youngest exposed magmatic episode in the area. Post-collisional whole rock analyses display shallow negative Eu anomalies and higher Rb/Sr ratios than are observed in the pre-collisional samples. High-K mafic lavas have slightly higher Pb isotope ratios (206Pb/204Pb(t) = 18.28–18.42, 207Pb/204Pb(t) = 15.55–15.61, 208Pb/204Pb(t) = 38.22–38.52), and lower εNd(t) values (−0.6 to +1.7), while low-K samples have lower Pb isotope ratios (206Pb/204Pb(t) = 18.26–18.28, 207Pb/204Pb(t) = 15.52–15.55, 208Pb/204Pb(t) = 38.11–38.29), and higher εNd(t(+2.4 to +2.9). The genesis of low-K rocks is closely replicated by 8% partial melting (i.e., F = 8%) of a batch melt defined by 96% mantle components (isotopically comparable to the least radiogenic E-DMM-like and OIB-like melts) metasomatized by 3.5% of sediment melt and 0.5% sediment fluid. The best-fit pattern mimics that produced by ∼97% batch melt later contaminated by ∼3% Cadomian upper crustal components. High-K suites are closely replicated by a melt comprising ∼96% products of F = 5% batch melt (i.e., 94% phlogopite-bearing lherzolite+5% sediment melt+1% sediment fluid, subsequently contaminated by ∼4% Cadomian lower crustal components. Continental arc evolution beneath the central UDMA occurred with slab steepening and roll-back leading to low-K magmatism until Early Miocene (pre-final collision), and break-off and detachment of the Neo-Tethyan oceanic slab to form local high-K magmatism from Middle Miocene onward (post-collision). We envisage different scenarios for the post-collisional magmatic suites of Iran since our proposed model of the central UDMA (e.g., E-DMM-like and OIB-like melts) does not conform with those recently suggested for post-collisional magmatic suites of the NW- and SE Iran, e.g., MORB-like asthenosphere, an enriched lithospheric mantle contaminated by continental material, and anatexis of continental lower crust. We acknowledge that distinguishing geochemical features preserved in UDMA volcanic and plutonic rocks from across the duration of convergence and suturing need not be diagnostic of arc maturity. Rather, they likely reflect changes in the geodynamic setting of the arc itself, e.g., slab angle, crustal and lithospheric thickness, over time.