Insights into sources and ore-forming processes of Cu-sulphide mineralization of the Phalaborwa Igneous Complex, from coupled Cu and Fe isotope and trace element systematics

Abstract

The Loolekop Pipe of the Phalaborwa Igneous Complex, Limpopo, South Africa, is the only known occurrence of a carbonatite –phoscorite sequence-hosted Cu-sulphide deposit of economic relevance. Despite significant research effort, some aspects of carbonatite-and phoscorite-hosted Cu-sulphide metallogenesis have remained elusive. In an attempt to address some outstanding research questions, this study focused on several specific points: (i) To characterize and classify sulphide textures and types, (ii) to constrain relative timing of ore-forming process(es), and (iii) to identify the metal source(s)involved in Cu mineralization. To this end, microfocus X-ray tomography and Mineral Liberation Analysis, combined with trace elements, Platinum-Group Elements (PGE) and Cu –Fe isotope systematicshave been used to establish a comprehensive model of the metallogenesis in the Loolekop Pipe. The Loolekop Pipe is the result of the successive emplacement of carbonatized melt pulses. The first unit to be emplaced was composed of a phoscorite –carbonatite sequence, where phoscorite is concentrated at the margins of the intrusion and carbonatite towards its centre. Due to the presence of mineral layering, this rock type is described as banded carbonatite. A structural event caused renewal of igneous activity and the intrusion of a second carbonatite pulse within a stockwork, known as transgressive carbonatite. Several Cu-sulphide generations were identified across the Loolekop Pipe. The first major form of Cu mineralization is characterized by chalcopyrite veins, which, considering the sharp contact with the host-rock, precipitated along fractures within the Loolekop Pipe. The second form is composed of sub-vertical sulphide layers hosted in both the banded carbonatite and the phoscorite. These layers are parallel to banding in the phoscorite and banded carbonatite. The magmatic assemblages are composed of bornite and chalcopyrite in various proportions. Three main textures can be observed among Cu-sulphide layers. The first texture consists of small (<1 mm) bornite and chalcopyrite grains. These grains are dispersed within the layer and are not associated with each other. Bornite can be found as isolated grains dispersed in the gangue without the presence of structural features, combined with chalcopyrite replacement or exsolution patches or laths. Bornite grains may locally host chalcocite veinlets and combined chalcocite and covellite replacement along its external boundaries. Isolated chalcopyrite can also be found, as well as trace amounts of cubanite. The second texture is characterized by stretched mm-sized bornite-chalcopyrite assemblages, which, considering the irregular surface, most likely derived from the merging of bornite and chalcopyrite grains. These assemblages are stretched in the same direction, suggesting the intervention of a magmatic flow process. The third and last texture shows mm(s)-wide chalcopyrite planar cumulates. Therefore, the mineralogical makeup cannot be considered as representative. The last major form of Cu-sulphide is constituted by valleriite [(Fe2+,Cu)4(Mg,Al)3S4(OH,O)6 ], which resulted from both in-situ replacement of magmatic and hydrothermal sulphides, as well as dissolution and re-precipitation of Cu along small fractures in close proximity to magmatic and hydrothermal sulphides. Similar PGE contents in the Primitive Mantle (5 ppb for Ru; 3.6 ppb for Pd; 6.6 ppb for Pt, respectively) and the Loolekop Pipe (1 to 4 ppb Ru; 2 to 8 ppb Pd; 1 to 9 ppb Pt) suggest a mantle origin of the sulphide material. Furthermore, low PGE contents coupled with heavy Fe isotope signatures (δ56Fe of 0.07 to 0.29‰) of banded carbonatite and phoscorite indicate a limited degree of partial melting and sulphide liquid extraction (~10%). Considering the preferential fractionation of Co and Zn into sulphide phases, the enrichment of both Co and Zn contents in whole-rocks (5 to 100 ppm and 15 to 100 ppm, respectively) relative to sulphides (0.1 to 40 ppm and 1 to 10 ppm, respectively) suggest some assimilation of external material and early interruption of chemical exchange between sulphide liquid(s) and melt(s).Platinum-Group Element concentrations in the Cu-sulphide assemblages reveal an enrichment of Pd-group PGE (PPGE; Pt and Pd) over Ir-group PGE (IPGE; Ir, Os and Ru). Palladium concentrations in sulphides range from 35 to 310 ppb, while Os and Ir range from <4 to 49 ppb and <1 to 4 ppb, respectively. Platinum contents range from <2 to 31 ppb, which can be explained by the extraction of Pt from sulphide phases through crystallization at depth of sperrylite (PtAs2) and its sequestration elsewhere. The PPGE enrichment over IPGE indicates the involvement of a monosulphide solid solution (mss) –intermediate solid solution(iss)system, where a sulphide liquid formed an IPGE-rich mss remaining at depth, and a Cu-and PPGE-rich sulphide liquid, which, by further ascent in the pipe and cooling down of the system, evolved into magmatic bornite and chalcopyrite. Iron and Cu isotope compositions of Cu-sulphides confirm a hypogene origin of these phases. Employing in-situ micro-drilling, values of δ65Cu range between -1.3 and 0.8‰and δ56Fe values between -0.6 and 0.3‰. Likewise, sulphide separates show similar variations in δ65Cu values between -0.7 and 0.4‰and δ56Fe values between -0.7 and 0.4‰. The alternation of sub-vertical banded carbonatite and phoscorite bandings, and a progressive decrease of phoscorite abundance towards the core of the pipe suggest the involvement of several magma pulses with a decreasing Fe-oxide and apatite component, from which phoscorite and banded carbonatite formed by immiscibility due to pressure drop. Relative proportions of bornite and chalcopyrite change from one sulphide layer to another, suggesting the presence of sulphide liquids of different compositions in the melt pulses. Despite this aspect, the similar Fe and Cu isotopic signatures indicate both a common material source and similar ore-forming processes. The similar distribution patterns, but lower concentrations of PGE in hydrothermal chalcopyrite relative to magmatic assemblages suggest that hydrothermal chalcopyritealong fractures within the core of the Loolekop Pipe may have resulted from the dissolution of magmatic sulphides and an iss phase by high-temperature hydrothermal fluids at depth, and subsequent re-precipitation of the dissolved material into a stockwork. However, light Fe isotope compositions (δ56Fe values between -0.7 and -0.2‰) of hydrothermal chalcopyrite indicate leaching and breakdown of magnetite at depth, which is the most likely assumption. Late-stage valleriite is consistently associated with both magmatic and hydrothermal sulphides. It shows that valleriite resulted from the alteration of these sulphide assemblages. The presence of valleriite along fractures within primary sulphides and around the boundaries of the grains indicate in-situ alteration of magmatic and hydrothermal sulphide assemblages and their replacement by valleriite. The presence of valleriite along fractures in proximity to primary sulphide phases indicates limited dissolution –precipitation. The preservation of primary sulphide assemblages in fracture-free samples illustrates the importance of late-stage fluid circulation for valleriite formation. The presence of gangue-filled veins in close spatial association to valleriite and other primary sulphide assemblages confirms the limited influence of dissolution –precipitation. Therefore, the presence of large valleriite veins in previous studies can be explained by the alteration and replacement of primary and hydrothermal assemblages. The higher concentrations of As and Pb in valleriite relative to magmatic sulphides (1 ppm vs 0.1 ppm for As; 100 to 840 ppm vs 13 to 380 ppm Pb, respectively) suggest a crustal origin of late-stage fluids. However, dissolution of magmatic and hydrothermal sulphides and precipitation of valleriite may also have contributed to this anomaly. The involvement of several magma pulses seems to be a classical feature for the formation of a banded carbonatite –phoscorite sequence, as observed for instance, in the Kovdor complex, Russia. Likewise, formation of phoscorite and carbonatite either by immiscibility or cumulate formation appears to be similar to other phoscorite –carbonatite sequences, such as in Kovdor or Salitre in Brazil. The main difference between the Loolekop Pipe and other phoscorite –carbonatite occurrences is the presence of significant amounts of Cu-sulphides. The occurrence of a significant amount of hydrothermal chalcopyrite can be partly explained by massive leaching of phoscorite-hosted magnetite by high-temperature fluids, although the presence in the Loolekop Pipe of sub-vertical magmatic Cu-sulphide layers remains unexplained, and the metal source unconstrained. Another phoscorite –carbonatite sequence hosted by the Salmagorskii Igneous Complex also contains Cu-sulphides, which appears to be post-magmatic, and which cannot be the case for the Loolekop-hosted sulphide layers. A possible explanation may be contribution from a Cu-and S-rich, and compositionally heterogeneous mantle underlying the Phalaborwa Igneous Complex at the time of magmatism