Trace element systematics and ore-forming processes in mafic VMS deposits: Evidence from the Troodos ophiolite, Cyprus

Abstract

The volcanogenic massive sulfide (VMS) deposits in the Troodos ophiolite (Cyprus) are ancient analogues for modern day seafloor massive sulfide mineralisation formed in a supra-subduction zone environment. In this study we present the first detailed in situ study of trace elements in sulfides from twenty VMS deposits hosted in the Troodos ophiolite to better understand factors that influence the distribution, enrichment and incorporation of trace elements in different sulfide minerals.On a mineral scale, trace elements exhibit systematic variations between pyrite, chalcopyrite and sphalerite. Pyrite preferentially incorporates As, Sb, Au and Te, whilst chalcopyrite is enriched in Co and Se. Sphalerite is trace element poor with the exception of Ag and Cd. Selenium averages 278 ppm (n = 150) in chalcopyrite but only 42 ppm (n = 1322) in pyrite. Bismuth and Te in pyrite show a weak positive correlation (R2 = 0.35) in some VMS deposits possibly linked with the occurrence of Bi-telluride inclusions. Trace element concentrations also vary between colloform and euhedral pyrite, with an enrichment of Au, As, Sb, Cu and Zn in colloform compared to euhedral pyrite.Time resolved laser ablation profiles reveal that the trace element distribution on a mineral scale is not uniform and varies with crystallographic effects, fluctuating physicochemical fluid conditions such as temperature, pH, fS2, fO2 and ligand availability during sulfide precipitation. Incorporation mechanisms in sulfides differ between elements in pyrite, Ag, As, Se and Pb are hosted in solid solution or as nanoscale inclusions, whilst Au, Sb and Te may form micro-scale inclusions.On a regional scale (20 km) the distribution of trace elements exhibits systematic variations between three major structural domains; namely the Solea, Mitsero and Larnaca grabens. The VMS deposits of the magmatic-tectonic Solea graben are enriched in Se, Co, Te, Au and Cu relative to Mitsero, which is a purely extensional feature. Therefore, we hypothesise that a variable magmatic volatile influx related to a) ‘magma’ volume, b) migration of the magmatic-hydrothermal crack front and associated brine liberation or c) a variation in protolith metal concentration are responsible for regional scale variations in VMS geochemistry. This is suggested to be intrinsically linked to the spreading architecture of Troodos.