O isotopes in the Azores : mantle melting versus AFC

The Azores archipelago in the central North Atlantic has been studied wideley over the last few decades, one reason being its formation history that is related to a slow-upwelling mantle plume. The plume sources identified in these ocean islands are predominantly inferred from Sr-Nd-Pb-(Hf) radiogenic isotope data along with major and trace element geochemistry. Contrastingly, few studies have dealt with the variability in stable isotopic composition. Here, we present a detailed isotopic study of oxygen isotope ratios (¹⁸O/¹⁶O) in olivines, clinopyroxenes and plagioclases that allow to infer on the quantiative and qualitativey involvement of altered oceanic crust (AOC) during ascent of the magmas. This is particularly important when aiming to place constraints on the primary O isotopic composition of the mantle plume source. Furthermore, such estimates can be combined with other newly availabe isotope data of boron (B) and Lithium (Li) of the Azores lavas. We provide a comprehensive data set of phenocryst O isotope data obtained by laser fluorination along with their individual mineral chemistry measured by electron microprobe on the same grain. The phenocrysts originate from lavas from the entire archipelago, namely Faial and Pico (central Azores) and Sao Miguel (east Azores), and, for the first time, include data for the two islands west of the mid-Atlantic ridge (MAR), Flores and Corvo. The samples were selected from geochemically well characterised host lavas such that the newly obtained O data are placed in a tight geochemical and petrological background. Our data indicate that the eastern Azores lavas are characterized by O ratios (i.e. δ¹⁸O₍ₒ₁₎ 4.7-5.2‰) slightly lighter than the primitive mantle range (δ¹⁸O₍ₒ₁₎ approximately 5.2±0.2‰), while the western lavas fall within the latter range. However, we find that decreasing forsterite contents in olivines are tightly correlated with decreasing δ¹⁸O₍ₒ₁₎. Such systematic is also observed in Hawaii', which is explained by assimilation of hydrothermally altered material during the crystallization-differentiation process. We test whether assimilation fractional crystallization (AFC) models sufficiently describe the δ¹⁸O₍ₒ₁₎ systematic of the Azores lavas, and if the highest forsterite olivines can be used to distinguish the δ¹⁸O₍ₒ₁₎ signal of the plume source. Our data are also compared with those of the MAR to better evaluate the nature and distribution of enriched components in the north Atlantic mantle.1 page(s