Early Archaean spherule beds in the Barberton Mountain Land, South Africa: no evidence for impact origin

Abstract

Detailed petrographical and geochemical studies are reported for several well-documented occurrences of spherule-rich layers in the Barberton Mountain Land, South Africa, which were previously inferred by Lowe and co-workers to be the result of early Archaean meteorite impacts. In contrast, we made the following observations. The textures of most spherules are not necessarily the result of impact, but could instead be the product of radial or intersertal growth of crystals during secondary mineral formation. There is no difference in the content of siderophile elements between spherule layers and layers devoid of spherules. The siderophile element abundances are high where sulphide minerals (e.g., pyrite, gersdorffite, and chalcopyrite) and/or chromite are present, independent of the presence or absence of spherules. High contents of, e.g., Ir (up to 2700 ppb), Ni (0.96 wt%), and Cr (1.6 wt%) were found in various samples. Abundances of these elements in chondritic meteorites are approximately 600 ppb, 1.4 wt%, and 0.35 wt%, respectively, resulting in respective meteoritic components in these samples of 450%, 70%, and 460%. We do not accept these high concentrations as primary meteoritic signatures. Impact melt rocks have typically « 1% of a meteoritic component. Furthermore, spherules of any kind are very rare in known impact deposits (including the K-T boundary). If present at all, they are not usually associated with any significant Ir or PGE anomaly. The high abundances of the siderophile elements in some Barberton samples and their enrichment in secondary minerals indicate that these elements have been remobilized and reconcentrated. The PGE interelement ratios have changed during remobilization as well. Thus, the PGE abundance patterns and ratios are not primary and cannot be used as an argument in favour of an impact origin either. Nickel-rich Cr-spinels, which are found in some of the Barberton spherule samples, have low Fe 3+ Fe (total) ratios. These ratios are incompatible with an extraterrestrial or impact origin, because such impact-derived spinels (including those found at the K-T boundary) are highly oxidized and also have different chemical compositions. There is also no indication of any evidence of shock metamorphism associated with the Barberton spherule layers, which is the commonly accepted definitive criterion for recognition of an impact origin. This is unusual, because such evidence is preserved even in heavily altered samples from deeply eroded Archaean impact structures. We conclude that, while these spherule layers are clearly unusual and deserve further attention, there is no convincing evidence for an origin by impact. We suggest that they were formed by volcanic processes, followed by extensive hydrothermal alteration.