Carbon isotope fractionation during diamond growth in depleted peridotite: counterintuitive insights from modelling water-maximum CHO fluids as multi-component systems

Because of the inability of depleted cratonic peridotites to effectively buffer oxygen fugacities when infiltrated by CHO or carbonatitic fluids, it has been proposed recently (Luth and Stachel, 2014) that diamond formation in peridotites typically does not occur by rock-buffered redox reactions as previously thought but by an oxygen-conserving reaction in which minor coexisting CH₄ and CO₂ components in a water-rich fluid react to form diamond (CO₂ + CH₄ = 2C + 2H₂O). In such fluid-buffered systems, carbon isotope fractionation during diamond precipitation occurs in the presence of two dominant fluid carbon species. Carbon isotope modelling of diamond precipitation from mixed CH₄- and CO₂-bearing fluids reveals unexpected fundamental differences relative to diamond crystallization from a single carbon fluid species: (1) irrespective of which carbon fluid species (CH₄ or CO₂) is dominant in the initial fluid, diamond formation is invariably associated with progressive minor (<1‰) enrichment of diamond in ¹³C as crystallization proceeds. This is in contrast to diamond precipitation by rock-buffered redox processes from a fluid containing only a single carbon species, which can result in either progressive ¹³C enrichment (CO₂ or carbonate fluids) or C13 depletion (CH₄ fluids) in the diamond. (2) Fluid speciation is the key factor controlling diamond δ¹³C values; as X_CO2 (X_CO2 = CO₂/[CO₂ + CH₄]) in the initial fluid increases from 0.1 to 0.9 (corresponding to an increase in fO₂ of 0.8 log units), the carbon isotope composition of the first-precipitated diamond decreases by 3.7‰. The tight mode in δ¹³C of −5 ±1‰ for diamonds worldwide places strict constraints on the dominant range of X_CO2 in water-rich fluids responsible for diamond formation. Specifically, precipitation of diamonds with δ¹³C values in the range −4 to −6‰ from mantle-derived fluids with an average δ¹³C value of −5‰ (derived from evidence not related to diamonds) requires that diamond-forming fluids were relatively reduced and had methane as the dominant carbon species (X_CO2 = 0.1–0.5). Application of our model to a recently published set of in-situ carbon isotope analyses for peridotitic diamonds from Marange, Zimbabwe (Smit et al., 2016), which contain CH₄ fluid inclusions, allows us to perfectly match the observed co-variations in δ¹³C, δ¹⁵N and N content and at the same time explain the previously counter-intuitive observation of progressive C13 enrichment in diamonds that appear to have grown from a fluid with methane as the dominant carbon species. Similarly, the almost complete absence in the published record of progressive C13 depletion trends within diamonds likely reflects ubiquitous precipitation from CH₄- and CO₂-bearing water-rich fluids, rather than diamond formation exclusively by carbonate-bearing and CH₄-free oxidized fluids or melts.