A TEM INVESTIGATION OF SHOCK METAMORPHISM IN QUARTZ FROM THE VREDEFORT DOME, SOUTH-AFRICA

The origin of the Vredefort structure in South Africa is still debated. Several causes have been discussed, namely asteroid impact, internal gas explosion or tectonic processes. Evidence of dynamic rock deformation is pervasive in the form of planar features in quartz grains, shatter cones, veins of pseudotachylite and occurrence of coesite and stishovite (high-pressure quartz polymorphs). A number of these characteristics is widely believed to support an impact origin. However, the planar features in quartz, which are generally considered as one of the strongest indicators of impact, are in the Vredefort case considered as anomalous when compared with those from accepted impact structures. We have investigated by optical and transmission electron microscopy (TEM) the defect microstructures in quartz grains from different lithologies sampled at various places at the Vredefort structure. Whatever the locality, only thin mechanical Brazil twin lamellae in the basal plane are observed by TEM. So far, such defects have only been found in quartz from impact sites, but always associated with sets of thin glass lamellae in rhombohedral planes {101nBAR} with n = 1, 2, 3, and 4. At the scale of the optical microscope, Brazil twins in (0001) are easily detected in Vredefort quartz grains because of the numerous tiny fluid inclusions which decorate them. Similar alignments of tiny fluid inclusions parallel to other planes are also detected optically, but at the TEM scale no specific shock defects are detected along their traces. If these inclusion alignments initially were shock features, they are now so severely weathered that they can no longer be recognized as unambiguous shock lamellae. Fine-grained coesite was detected in the vicinity of narrow pseudotachylite veinlets in a quartzite specimen, but stishovite was not found, even in areas where its occurrence was previously reported. Finally, definite evidence of high-temperature annealing was observed in all the samples. These observations lead us to the conclusion that our findings regarding microdeformation in quartz are consistent with an impact origin for the Vredefort structure. Most of the original shock defects are now overprinted by an intense post-shock annealing episode. Only the thin mechanical twin lamellae in the basal plane have survived.