The geology and geochemistry of the El Laco iron oxide deposit (Central Andes, Chile) support a genesis related to the ascent, degassing, and subvolcanic emplacement of an unusual oxidized silica-poor but water-and iron-rich melt that took place during the growth of the host Pliocene-Holocene andesitic volcano. The model proposed in this paper for the evolution of the deposit involves the formation of a shallow telescoped magmatichydrothermal system with complex melt-fluid unmixing in a vertical column of less than 1 km. The dominant mineralization occurs as large stratabound apatite-poor magnetite bodies interfingered with an andesite host and rooted in vertical dikes of magnetite with minor apatite. The stratabound mineralization is crosscut by abundant coeval diatreme-like structures indicative of vigorous degassing. The andesite underlying the mineralization is pervasively replaced by a high-temperature alkali-calcic alteration assemblage (K feldspar-diopside-magnetite-scapolite) that includes coarse-grained diopside-magnetite-anhydrite veins and large subvertical bodies of magmatic-hydrothermal breccias. The host andesite also shows a large strata-bound steam-heated acid alteration that is devoid of any magnetite but has produced the replacement of a significant proportion of the early magnetite by hematite. The El Laco system is rich in anhydrite but poor in sulfides, suggesting that there were persistent oxidizing conditions that inhibited the formation of a sulfide-bearing mineralization.
Field evidence, oxygen isotope geothermometry, and thermodynamic constraints suggest that the magnetite mineralization formed close to the surface at temperatures above 800°C. The magnetite textures, similar to those of subaerial low-viscosity basalts, and the presence of melt inclusions in the host andesite recording the presence of immiscible Fe-Mg-Ca-(Si-Ti-P-S) and Si-K-Na-Al melts, suggest that the magnetite ore formed by direct crystallization from an iron-rich melt; its chemistry inhibited the formation of most other magmatic phases except minor apatite, anhydrite, and diopside. The crystallization of the iron-rich melt at shallow depths promoted the separation of large amounts of two immiscible aqueous fluids: a dominant low-density vapor phase and a small volume of hypersaline fluid. Diopside-magnetite-anhydrite veins are interpreted as the product of the crystallization of the residual melts, whereas the interaction of the brine with the host andesite formed the deep alkali-calcic hydrothermal assemblage. The condensation and mixing of the low-density magmatic vapor with meteoric water produced the steam-heated alteration.
Isotope data from the host andesite (87Sr/86Sr: 0.7066–0.7074; εNd: −5.5 to −4.1; δ18Owhole rock: 7.2–9.6‰; δ18Omagnetite: 5.1–6.2‰) and an underlying andesite porphyry (87Sr/86Sr: 0.7075–0.7082; εNd: −5.9 to −4.6) reflect the interaction of a primitive mantle melt with Andean crustal rocks. The isotope geochemistry of the magnetite ore (87Sr/86Sr: 0.7083; εNd: −5.4 to −5.1; δ18O 3.5–5.5‰) and the alkali-calcic alteration and related diopside-magnetite-anhydrite veins (87Sr/86Sr: 0.7080–0.7083; εNd: −5.1 to −4.6; δ18Odiopside: 7.2–8.2%c; δ18Omagnetite 4.4–6.3‰) show that the mineralization has a more crustal signature than the host andesite and all the volcanic rocks of the Central Andes. Therefore, ore-forming fluids/melts were not equilibrated with the host volcanic rocks and are interpreted as related to a deep yet undiscovered batch of highly contaminated igneous rocks. Crustal contamination is interpreted as due to major interaction of a juvenile melt with the underlying Late Mesozoic-Tertiary Salta Group, located 1 to 6 km beneath the volcano and which has high 87Sr/86Sr values (0.7140–0.7141).