Formation of the Campbell-Red Lake gold deposit by H2O-poor, CO2-dominated fluids

Abstract

The Campbell-Red Lake gold deposit in the Red Lake greenstone belt, with a total of approximately 840 t of gold (past production + reserves) and an average grade of 21 g/t Au, is one of the largest and richest Archean gold deposits in Canada. Gold mineralization is mainly associated with silicification and arsenopyrite that replace carbonate veins, breccias and wallrock selvages. The carbonate veins and breccias, which are composed of ankerite ± quartz and characterized by crustiform–cockade textures, were formed before and/or in the early stage of penetrative ductile deformation, whereas silicification, arsenopyrite replacement and gold mineralization were coeval with deformation. Microthermometry and laser Raman spectroscopy indicate that fluid inclusions in ankerite and associated quartz (Q1) and main ore-stage quartz (Q2) are predominantly carbonic, composed mainly of CO2, with minor CH4 and N2. Aqueous and aqueous–carbonic inclusions are extremely rare in both ankerite and quartz. H2O was not detected by laser Raman spectroscopic analyses of individual carbonic inclusions and by gas chromatographic analyses of bulk samples of ankerite and main ore-stage quartz (Q2). Fluid inclusions in post-mineralization quartz (Q3) are also mainly carbonic, but proportions of aqueous and aqueous–carbonic inclusions are present. Trace amounts of H2S were detected by laser Raman spectroscopy in some carbonic inclusions in Q2 and Q3, and by gas chromatographic analyses of bulk samples of ankerite and Q2. 3He/4He ratios of bulk fluid inclusions range from 0.008 to 0.016 Ra in samples of arsenopyrite and gold. Homogenization temperatures (T h–CO2) of carbonic inclusions are highly variable (from −4.1 to +30.4°C; mostly to liquid, some to vapor), but the spreads within individual fluid inclusion assemblages (FIAs) are relatively small (within 0.5 to 10.3°C). Carbonic inclusions occur both in FIAs with narrow T h–CO2 ranges and in those with relatively large T h–CO2 variations. The predominance of carbonic fluid inclusions has been previously reported in a few other gold deposits, and its significance for gold metallogeny has been debated. Some authors have proposed that formation of the carbonic fluid inclusions and their predominance is due to post-trapping leakage of water from aqueous–carbonic inclusions (H2O leakage model), whereas others have proposed that they reflect preferential trapping of the CO2-dominated vapor in an immiscible aqueous–carbonic mixture (fluid unmixing model), or represent an unusually H2O-poor, CO2-dominated fluid (single carbonic fluid model). Based on the FIA analysis reported in this study, we argue that although post-trapping modifications and host mineral deformation may have altered the fluid inclusions in varying degrees, these processes were not solely responsible for the formation of the carbonic inclusions. The single carbonic fluid model best explains the extreme rarity of aqueous inclusions but lacks the support of experimental data that might indicate the viability of significant transport of silica and gold in a carbonic fluid. In contrast, the weakness of the unmixing model is that it lacks unequivocal petrographic evidence of phase separation. If the unmixing model were to be applied, the fluid prior to unmixing would have to be much more enriched in carbonic species and poorer in water than in most orogenic gold deposits in order to explain the predominance of carbonic inclusions. The H2O-poor, CO2-dominated fluid may have been the product of high-grade metamorphism or early degassing of magmatic intrusions, or could have resulted from the accumulation of vapor produced by phase separation external to the site of mineralization.