A review of magnetite geochemistry of Chilean iron oxide-apatite (IOA) deposits and its implications for ore-forming processes

Highlights

• Magnetite can be used as a tracer of the geological environment of ore formation and as a fingerprint of the source reservoir of iron.

• The formation and evolution of IOAs deposits can be assessed using compatible trace elements in magnetite such as Al, Ti, V, Cr, Mn, Ni, Cu and Ga.

• Chilean IOA deposits formed at different conditions ranging from high-temperature purely magmatic, to magmatic-hydrothermal, to low-temperature hydrothermal conditions.

• Magnetite is susceptible to multiple microtextural and compositional re-equilibrium events during the formation and evolution of IOA deposits.

• The flotation model as a basal transfer mechanism of Fe allows variations and the definition of different subtypes of IOA mineralization, controlled by the depth of formation, structures/faults, composition of the host rock, the source of hydrothermal fluids and the fluid/rock ratio.

Abstract

Magnetite is the most important iron ore in iron oxide-apatite (IOA) deposits which represent the Cu-poor end-member of the iron oxide-copper–gold (IOCG) clan. Magnetite chemistry has been used as a petrogenetic indicator to identify the geological environment of ore formation and as a fingerprint of the source reservoir of iron. In this study, we present new textural and microanalytical EPMA and LA-ICP-MS data of magnetite from Carmen, Fresia, Mariela and El Romeral IOA deposits located in the Cretaceous Coastal Cordillera of northern Chile. We also provide a comprehensive summary and discussion of magnetite geochemistry from Andean IOAs including Los Colorados, Cerro Negro Norte, El Romeral (Chilean Iron Belt) and the Pliocene El Laco IOA deposit located in the Central Volcanic Zone of the Chilean Andes. Microtextures coupled with geochemical data were used to define and characterize the occurrence of different magnetite types. Magnetite exhibits a variety of textural features including oscillatory zoning, colloform banding, re-equilibration textures, exsolution lamellae and symplectites. The magmatic vs. hydrothermal origin of the different magnetite types and the evolution of IOA deposits can be assessed using diagrams based on compatible trace elements. However, magnetite is very susceptible to hydrothermal alteration and to both textural and compositional re-equilibration during magmatic and superimposed hydrothermal events. Based on the data presented here, we conclude that V and Ga are possibly the most reliable compatible elements in magnetite to trace ore-forming processes in the Andean IOA deposits. Magnetite chemistry reveals different conditions/events of formation for each IOA deposit ranging from high-temperature, low-oxygen fugacity (ƒO₂), purely magmatic (>600 °C) conditions; to lower temperature and higher ƒO₂ magmatic-hydrothermal (300–600 °C) to low-temperature hydrothermal (<200–300 °C) conditions. Specifically, a continuous transition from high-temperature, low- ƒO₂ conditions in the deepest portions of the deposits to low-temperature, relatively higher ƒO₂ conditions towards surface are described for magnetite from El Laco. The new and compiled magnetite data from IOA deposits from the Chilean Iron Belt and El Laco are consistent with a transition from magmatic to hydrothermal conditions. The flotation model plausibly explains such features, which result from the crystallization of magnetite microlites from a silicate melt, nucleation and coalescence of aqueous fluid bubbles on magnetite surfaces, followed by ascent of a fluid-magnetite suspension along reactivated transtensional faults or through fissures formed during the collapse of the volcanic structure (El Laco). The decompression of the coalesced fluid-magnetite aggregates during ascent promotes the continued growth of magnetite microlites from the Fe-rich magmatic-hydrothermal fluid. As with any general genetic model, the flotation model allows variation and the definition of different styles or subtypes of IOA mineralization. The deeper, intrusive-like Los Colorados deposit shows contrasting features when compared with the Cerro Negro Norte hydrothermal type, the pegmatitic apatite-rich deposits of Carmen, Fresia and Mariela, and the shallow, subaerial deposits of El Laco. These apparent differences depend fundamentally on the depth of formation, the presence of structures and faults that trigger decompression, the composition of the host rocks, and the source and flux rate of hydrothermal fluids.

Introduction

Magnetite (Fe₃O₄) is a common accessory mineral in igneous, metamorphic and sedimentary rocks and can form under a wide range of temperature and pressure conditions. It can also become concentrated in large quantities to form ore deposits such as banded iron formations (e.g., Angerer et al., 2012, Araújo and Lobato, 2019), magmatic Fe-Ti-V oxide (e.g., Zhou et al., 2005), iron skarns (e.g., Nadoll et al., 2014, Nadoll et al., 2015), iron oxide-copper–gold (IOCG) and Kiruna-type iron oxide-apatite (IOA) deposits (e.g., Williams et al., 2005).

Iron oxide-apatite deposits represent the Cu-poor end-member of the IOCG clan (Sillitoe, 2003) and are an important source of Fe. Iron oxide-apatite mineralization is dominated by low-Ti magnetite, which can be accompanied by variable amounts (1–50% modal) of apatite, actinolite, pyroxene, epidote and sulfides (Williams et al., 2005). On the other hand, IOCG deposits are mined for their Cu content, but in some cases Au, U, REE, P, Co, Bi and Nb are relevant by-products (Hitzman et al., 1992, Sillitoe, 2003, Williams et al., 2005, Groves et al., 2010, Barton, 2014). IOCG mineralization is characterized by Cu-Fe sulfides (chalcopyrite and minor bornite), Cu oxides and abundant iron oxides (magnetite and/or specular hematite). Hydrothermal alteration is represented by sodic-calcic (albite, actinolite, epidote) and potassic (biotite, orthoclase) assemblages with minor chlorite, sericite and late calcite. Regardless of the high concentration of iron oxides in IOCGs, they are rarely mined for Fe.

Both IOCG and IOA deposits occur globally with reported ages ranging from Archean (e.g., Carajás Province, Brazil; de Melo et al., 2017), early Proterozoic (e.g., Kiruna district, Sweden and Olympic Dam district, Australia; Westhues et al., 2017a, Westhues et al., 2017b, Apukhtina et al., 2017), middle Proterozoic (e.g., Pea Ridge and Pilot Knob deposits, USA, and Cloncurry district, Australia; Rusk et al., 2010, Day et al., 2016, Childress et al., 2016), Cretaceous (e.g., Los Colorados, El Romeral, Candelaria and Mantoverde, Chile; Benavides et al., 2007, Rieger et al., 2010, Rieger et al., 2012, Knipping et al., 2015a, Knipping et al., 2015b, Rojas et al., 2018a) and Pliocene (e.g., El Laco, Chile; Maksaev et al., 1988, Nystrӧm and Henríquez, 1994, Naranjo et al., 2010).

In Chile, most of the IOCG and IOA deposits occur spatially and temporally associated with one another in the Jurassic − Early Cretaceous Chilean Iron Belt (CIB) within the Coastal Cordillera of northern Chile between latitudes 25° and 31°S (Fig. 1). These deposits occur closely associated with coeval Mesozoic intrusions and are structurally controlled by the arc-parallel Atacama Fault System (AFS) (Sillitoe, 2003, Williams et al., 2005). As the youngest-known IOCG − IOA province in the world, mostly unaffected by metamorphism and deformation (Sillitoe, 2003), the CIB is an ideal area to study the processes involved in the formation of IOA and IOCG deposits and to refine the current genetic model. Noteworthy, the Pliocene El Laco Volcanic Complex (ELVC), located further north (23°48′S) in the high Andes, hosts the youngest, least altered, and best preserved IOA deposit discovered to date (Naranjo et al., 2010; Fig. 1). Within an area of ~ 35 km², six magnetite ore bodies have been described at El Laco (i.e., Laco Norte, Laco Sur, San Vicente Alto, Rodados Negros, Cristales Grandes and San Vicente Bajo). These ore bodies display remarkable volcanic and subvolcanic features that have fueled the controversy regarding a magmatic (immiscibility model) or metasomatic origin for IOA deposits (Frutos and Oyarzún, 1975, Nystrӧm and Henríquez, 1994, Sillitoe and Burrows, 2002, Naranjo et al., 2010, Tornos et al., 2016, Velasco et al., 2016, Ovalle et al., 2018).

It is well known that IOCG deposits are mostly formed by hydrothermal processes (Sillitoe, 2003, Williams et al., 2005, Mumin et al., 2007, Groves et al., 2010, Barton, 2014). However, the origin of IOA deposits remains controversial and, historically, two contrasting “classic” hypotheses have been proposed. The first hypothesis comprises a purely magmatic origin involving the separation of an immiscible Fe-P-rich melt from a silicate melt with the subsequent intrusion and crystallization of an Fe-rich ore body at upper crustal levels (e.g., Nystrӧm and Henríquez, 1994, Naslund et al., 2002, Chen et al., 2010, Velasco et al., 2016). The second hypothesis involves a purely hydrothermal origin where the magnetite ore is formed by metasomatic replacement of host rocks by Fe-rich fluids sourced from magmatic or non-magmatic sources (e.g., Rhodes and Oreskes, 1995, Rhodes and Oreskes, 1999, Sillitoe and Burrows, 2002, Barton and Johnson, 1996, Barton and Johnson, 2004, Pollard, 2006, Dare et al., 2015, Westhues et al., 2016, Westhues et al., 2017b, Westhues et al., 2017a). Most of the current debate regarding the origin of IOA deposits has been centered on El Laco, where some researchers have argued that the magnetite ore bodies preserve purely igneous/volcanic features, such as pahoehoe- and aa-like lava flows, different degrees of vesicularity, contorted flow layering, spherulitic and dentritic magnetite, and vertical structures interpreted as gas escape tubes (Nystrӧm and Henríquez, 1994, Naslund et al., 2002). However, others have interpreted these textures as relicts of the andesite lava flows that have been replaced by hydrothermal iron oxides, based on magnetite veins and breccias, and hydrothermal alteration assemblages observed around the magnetite bodies (Rhodes et al., 1999, Sillitoe and Burrows, 2002).

Over the past few years, and in light of a renewed interest in magnetite geochemistry studies, a novel genetic model was proposed by Knipping et al., 2015a, Knipping et al., 2015b to explain the formation of IOA deposits. By using a combination of field observations, trace element geochemistry and Fe and O stable isotopes composition in magnetite grains from Los Colorados, those authors propose that IOA deposits formed by a combination of igneous and magmatic-hydrothermal processes. In their flotation model, microlites of magnetite crystallized from a silicate melt triggering bubble nucleation, forming a magnetite-fluid suspension that transported magnetite to shallow depths through pre-existing fault structures that are reactivated by crustal extension. This new model not only reconciles the two opposing genetic hypotheses (liquid immiscibility vs. metasomatic replacement), but also provides a plausible explanation for the formation of spatially associated IOCG deposits as part of the same evolving system (Reich et al., 2016, Barra et al., 2017, Simon et al., 2018). A genetic connection between IOA and IOCG deposits of the Coastal Cordillera of northern Chile has been previously proposed, where IOA (sulfide-poor) deposits may represent the roots or deep portion of IOCG (sulfide-rich) systems (Espinoza et al., 1996, Sillitoe, 2003).

The current discussion about the genesis of IOA deposits is based on, among other proxies, the major and trace element geochemistry of magnetite, which is the modally dominant mineral in all IOA deposits. Both electron microprobe analysis (EMPA) and laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) have been used to measure the chemical composition of major and trace elements in magnetite, respectively. While EMPA allows for a low-micron size beam (<5 μm) to analyze a limited number of elements, e.g., Mg, Al, Ca, Ti, V, Cr and Mn at detection limits of tens to hundreds of ppm (e.g., Dupuis and Beaudoin, 2011), LA-ICP-MS allows in-situ measurements on 20–50 µm spots of a large number of trace elements (>40) at ppm to sub-ppm levels. These analytical advances have resulted in a growing interest in the study of magnetite as a petrogenetic indicator and as tool to identify past geological environments of ore formation or as a fingerprint of mineral deposit types (Dupuis and Beaudoin, 2011, Dare et al., 2012, Dare et al., 2014, Nadoll et al., 2012, Nadoll et al., 2014, Huang et al., 2018, Huang et al., 2019, Huang and Beaudoin, 2019). In particular, Knipping et al., 2015a, Knipping et al., 2015b documented distinct chemical zoning in magnetite grains from Los Colorados IOA deposit, where cores are enriched in Ti, Al, Mn and Mg indicating crystallization from a silicate melt. In contrast, magnetite rims show a pronounced depletion in these elements, consistent with magnetite grown from a Fe-rich magmatic-hydrothermal aqueous fluid (Knipping et al., 2015b).

After Knipping et al., 2015a, Knipping et al., 2015b, several geochemical and textural studies have been carried out on magnetite and other mineral phases (pyrite, actinolite, apatite) in the Andean IOCG and IOA deposits including: El Romeral (Rojas et al., 2018a, Rojas et al., 2018b); Los Colorados (Bilenker et al., 2016, Reich et al., 2016, Deditius et al., 2018, La Cruz et al., 2019, Knipping et al., 2019a); El Laco (Ovalle et al., 2018, La Cruz et al., 2019, La Cruz et al., 2020); Cerro Negro Norte (Salazar et al., 2020); Carmen, Fresia and Mariela (Palma et al., 2019); Candelaria (Rodriguez-Mustafa et al., 2020); Mantoverde (Childress et al., 2020). The chemical composition of magnetite grains from Cerro Negro Norte, El Romeral and El Laco deposits reflect their formation conditions, revealing systems that evolved from high-temperature purely igneous/magmatic conditions at depth to lower temperature magmatic-hydrothermal conditions at shallow depths (Ovalle et al., 2018, Rojas et al., 2018b, Salazar et al., 2020, La Cruz et al., 2020). The data presented in such studies have been described as consistent with the flotation model to explain the origin of Andean IOA deposits.

In this contribution we present a review of the extensive magnetite chemical data gathered during these last few years on Andean IOA and IOCG deposits. In addition, we present and discuss new micro-textural, EPMA and LA-ICP-MS data for magnetite from the Carmen, Fresia, Mariela and El Romeral IOA deposits. The trace element composition of magnetite, geochemical discrimination diagrams (e.g., Dupuis and Beaudoin, 2011, Nadoll et al., 2014) and Principal Component Analysis (PCA) are used here to fingerprint the texturally different magnetite types observed in each deposit and provide new constraints on the formation of these deposits. We also demonstrate how magnetite compositional data should be closely inspected and complemented with micro-textural observations to avoid misinterpretations. Finally, we propose a mechanism of formation for the apatite-rich (~20–50% modal), pegmatitic IOA subtype of Carmen, Fresia and Mariela.