The Light Stable Isotope (Hydrogen, Boron, Carbon, Nitrogen, Oxygen, Silicon, Sulfur) Composition of Orogenic Gold Deposits

Abstract

AbstractOrogenic gold deposits formed in various terranes of most ages since the Paleoarchean and generally consist of quartz veins hosted in shear zones formed at the ductile brittle transition under greenschist to lower amphibolite metamorphic conditions. Vein mineralogy is dominated by quartz with various amounts of silicates, carbonates, phyllosilicates, borates, tungstates, sulfides, and oxides. The isotopic composition of these minerals and fluid inclusions has been investigated since the 1960s to constrain the characteristics of orogenic fluid systems involved in the formation of gold deposits worldwide. This review is based on 8580 stable isotope analyses, including δ18O, δD, δ13C, δ34S δ15N, δ11B, and δ30Si values, from 5478 samples from 558 orogenic gold deposits reported in the literature from 1960 to 2010. This contribution describes the variability of the light stable isotopic systems as function of the minerals, the age of the deposits, their regional setting, and their country rocks. The temperature of isotopic equilibrium of orogenic gold veins is estimated from mineral pairs for oxygen and sulfur isotopes. Based on these temperatures, and on fractionation between mineral and fluid components (H2O, CO2 and H2S), the isotopic composition of fluids is estimated to better constrain the main parameters shared by most of auriferous orogenic fluid systems. Orogenic gold deposits display similar isotopic features through time, suggesting that fluid conditions and sources leading to the formation of orogenic gold deposits did not change significantly from the Archean to the Cenozoic. No consistent secular variations of mineral isotope composition for oxygen (−8.1‰ ≤ δ18O ≤ 33‰, n = 4011), hydrogen (−187‰ ≤ δD ≤ −4‰, n = 246), carbon (−26.7‰ ≤ δ13C ≤ 12.3‰, n = 1179), boron (−21.6‰ ≤ δ11B ≤ 9‰, n = 119), and silicon (−0.5‰ ≤ δ30Si ≤ 0.8‰, n = 33) are documented. Only nitrogen (1.6‰ ≤ δ15N ≤ 23.7‰, n = 258) and sulfide sulfur from deposits hosted in sedimentary rocks (−27.2‰ ≤ δ34S ≤ 25‰, n = 717) display secular variations. For nitrogen, the change in composition is interpreted to record the variation of δ15N values of sediments devolatilized during metamorphism. For sulfur, secular variations reflect incorporation of local sedimentary sulfur of ultimate seawater origin. No significant variation of temperature of vein formation is documented for orogenic gold deposits of different ages. Quartz-silicate, quartz-carbonate and sulfide-sulfide mineral pairs display consistent temperatures of 360 ± 76 °C (1σ; n = 332), in agreement with the more common greenschist facies hostrocks and fluid inclusion microthermometry. Fluid sources for orogenic gold deposits are complex but the isotopic systems (hydrogen, boron, carbon, nitrogen, oxygen, sulfur) are most consistent with contributions from metamorphic fluids released by devolatilization of igneous, volcano-sedimentary and/or sedimentary rocks. The contribution of magmatic water exsolved from magma during crystallization is not a necessary component, even if permissible in specific cases. Isotopic data arrays can be interpreted as the result of fluid mixing between a high T (~550 °C)—high δ18O (~10‰)—low δD (~−60‰) deep-seated (metamorphic) fluid reservoir and a low T (~200 °C)—low δ18O (~2‰)—high δD (~0‰) upper crustal fluid reservoir in a number of orogenic gold deposits. The origin of the upper crustal fluid is most likely sea- or meteoric water filling the host rock porosity, with a long history of water–rock isotope exchange. Mixing of deep-seated and upper crustal fluids also explains the large variation of tourmaline δ11B values from orogenic gold veins. Regional spatial variations of oxygen and hydrogen isotope compositions of deep-seated fluid reservoirs are documented between orogenic gold districts. This is the case for the Val-d’Or (Abitibi), Coolgardie and Kalgoorlie (Yilgarn) where the oxygen isotope composition of the deep-seated fluid end-member is 4‰ lower compared to that from the Timmins, Larder Lake, and Kirkland Lake districts (Abitibi). However, both mixing trends converge towards a common, low δ18O upper crustal fluid end-member. Such variations cannot be related to fluid buffering at the site of deposition and suggest provinciality of the fluid source. The contribution of meteoric water is mainly recorded by fluid inclusions from Mesozoic and Cenozoic age deposits, but micas are not systematically in isotopic equilibrium with fluid inclusions trapped in quartz from the same vein. This suggests late involvement of meteoric water unrelated to deposit formation. Yet, a number of deposits with low δD mica may record infiltration of meteoric water in orogenic gold deposits. Isotope exchange between mineralizing fluid and country rocks is documented for oxygen, carbon, sulfur and silicon isotopes. Large variations (> 10‰) of sulfide δ34S values at the deposit scale are likely related to evolving redox conditions of the mineralizing fluid during reaction with country rocks. Deposits hosted in sedimentary rocks show a shift to higher δ18O values as a result of fluid/rock oxygen exchange with the regional sedimentary country rocks.