EVIDENCE FOR O2-POOR ARCHEAN ATMOSPHERE FROM IMPACT SPINELS IN THE BARBERTON GREENSTONE BELT, SOUTH AFRICA

Ni-rich chrome spinel is one of the only primary minerals preserved in Archean impact spherules of the Barberton Greenstone Belt (BGB), South Africa. Analysis of the spinels can provide insight into the redox conditions of the impact plume and the abundance of oxygen in the early atmosphere. Elemental chemistry of individual Ni-rich chrome spinels in spherules from several sections of the 3.24 Ga S3 spherule bed was analyzed by microprobe. Fe²⁺ and Fe³⁺ were calculated from total Fe assuming a stoichiometric spinel composition. Detrital komatiitic spinels from S3 were also analyzed. The impact spinels are significantly more oxidized than the associated detrital spinels of komatiitic origin. The average Fe⁺³/Fe_T (atoms) value in the Ni-rich impact spinels is about 0.43 whereas the average Fe⁺³/Fe_T of detrital spinels is 0.17.

Fe³⁺/Fe_T ratios of the impact spinels, which range from 0.26 to 0.69 (atoms), suggest formation at oxygen fugacities below 10^-4 based on experimental data by Gayraud et al. (1996). In comparison to the K/T boundary layer spinels, with Fe³⁺/Fe_T values between 0.77 and 0.99 (Gayraud et al. 1996 and others) the BGB spinels reflect low-O₂ levels in the 3.24 Ga atmosphere. Oxidation of the impact-produced spinels likely occurred through their interaction with the atmosphere during formation and fallout, and the broad range in values may reflect the redox conditions in the plume and the atmosphere, particularly as some spinels show increased Fe³⁺/Fe_T toward the rims. This may suggest the atmosphere was more oxidizing than the plume. Cr₂O₃ and NiO are more abundant in the BGB spinels compared to K/T spinels, which tend to have more abundant Al₂O₃ and Fe₂O₃. This may be due to longer, higher temperatures in the Archean impact plume than in the K/T plume, though the BGB spherule layer clearly has a greater overall meteoritic component than the K/T spherules.

Reference: J. Gayraud, E. Robin, R. Rocchia, L. Froget, GSA Special Paper 307, 425-443 (1996).