The formation of Cu-Au-Mo porphyry deposits in Iran is linked to continental collision, yet there are no clear temporal relations between geochemical signatures of Miocene (collisional) and Eocene (pre-collisional) intrusive arc rocks, and the presence/lack of ore mineralization. We compare geochemical scenarios from five different segments along the Urumieh-Dokhtar arc and propose a geodynamic model to explain this discrepancy. For instance, in the Natanz arc segment in central Iran, contrasting geochemical signatures of copper ore hosting Eocene and some barren undeformed Miocene diorites to granites temporally overlap with the Alpine-Himalayan collision. These changes provide key implications on the existence and lack of Cu mineralization during collisional magmatism. High Sr and low Y (and Yb) contents of Eocene arc rocks in the Natanz arc segment reflect thickened, Andean-type orogenic arccrust (~45 km), whereas barren Miocene Natanz arc rocks (21-19 Ma) indicate thin arc crust similar to collisional volcanism in Anatolia. Geochemical modeling indicates a change in the mineralogy of the melt residual, from precollisional Eocene basaltic garnet-bearing (5-30%) amphibolite to syn- or postcollisional Miocene metasomatized mantle peridotite, which can be explained by collision-induced delamination of the arc lithospheric root. Subsequent recharge of hot asthenosphere and melting of metasomatized mantle peridotite and lack of interaction with a garnet-bearing arc crustal keel explain the low Sr and high Y (and Yb) contents, the relatively enriched initial Sr isotope ratios of postcollisional Miocene Natanz rocks, and the lack of copper mineralization in postcollisional Miocene Natanz arc rocks. Arc-root delamination removes the copper- and sulfurenriched metasomatized lithospheric arc root and hydrous cumulate reservoir required to form copper ore deposits. Lack of the dense melt residues also provides an alternative explanation for the elevated, thin crustal Irania
