Implications of the carbon isotopes and mineral inclusion record for the formation of diamonds in the mantle underlying a mobile belt: Venetia, South Africa

Peter Deines*, Fanus Viljoen, J. W. Harris

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

36 Citations (Scopus)


A total of 199 diamonds from the Venetia kimberlite, South Africa, whose mineral inclusion chemistry had already been measured, were analyzed for their carbon isotopic composition. Silicate inclusions in these diamonds either belong to a peridotitic (P-Type), an eclogitic (E-Type) or transitional, websteritic (W-Type), paragenesis. The carbon isotopic composition of 161 P-Type diamonds ranges from δ13C= -2.23 to -18‰ vs. PDB. The large number of samples available and the wide range in δ13C permitted, for the first time, a detailed analysis of the relationships between P-Type inclusion chemistry and the carbon isotopic composition of the diamond host. The δ13C sampling frequency distribution is multi-modal. Examination of the inclusion chemistry (chromite, olivine, garnet) as a function of the carbon isotope composition mode to which the host belongs, as well as multivariate regression analyses, revealed no correlation between inclusion chemistry and 13C content. The inclusion compositions in diamonds of low13C content are not distinctive. For a given carbon isotopic composition the combination of Ni/Fe and Mg/(Mg + Fe) of olivine inclusions varies systematically along fractionation trends. The composition of the olivine inclusions and the 13C content of their hosts can be interpreted as reflecting similar petrogenetic processes occurring in several mantle environments into which carbon of variable isotopic composition was introduced. The iron/magnesium distribution between coexisting garnets and olivines permits an estimate of their pressure/temperature equilibration conditions. Diamonds whose inclusions were equilibrated at lower temperatures and pressures tend to have, on average, lower 13C contents. The compositions of coexisting olivines and chromites suggest oxygen fugacities between 2.9 and 5.8 log units below the quartz-fayalite-magnetite buffer at 50 kbar, and temperatures between 1280 and 1490°C prevailed during diamond formation. Inclusions from diamonds of lower 13C content do not indicate systematically lower fO2 values during their formation. The fO2/T conditions determined suggest minimal (0.0 to -0.5‰) isotope fractionation between a C-H-O vapor phase (carbon dioxide, carbon monoxide, methane, water, hydrogen) and diamond. The large 13C depletion of some Venetia P-Type diamonds appears to be unrelated to the composition of their inclusions, igneous fractionation trends, oxygen fugacity, and vapor isotope fractionation processes. This conclusion is consistent with evidence deduced from more limited data sets from other kimberlites. Eclogitic diamonds constitute less than 10% of the inclusion bearing Venetia diamonds. The δ13C values of 19 E-Type samples range from -4.39 to -15.6‰. The nature of the relationship between inclusion chemistry and carbon isotopic composition of the hosts parallels that observed for P-Type diamonds. W-Type diamonds occur least frequently; the δ13C range for six samples is -3.74 to -5.91‰. The carbon isotope distribution and inclusion chemistry of Venetia diamonds are akin to those of diamonds from most kimberlites located on the Kaapvaal craton. This indicates that diamond formation in the mantle underlying the Limpopo Mobile Belt followed processes and involved carbon sources that are very similar to those involved in diamond formation in the mantle beneath the Kaapvaal craton.

Original languageEnglish
Pages (from-to)813-838
Number of pages26
JournalGeochimica et Cosmochimica Acta
Issue number5
Publication statusPublished - 2001
Externally publishedYes


Dive into the research topics of 'Implications of the carbon isotopes and mineral inclusion record for the formation of diamonds in the mantle underlying a mobile belt: Venetia, South Africa'. Together they form a unique fingerprint.

Cite this