CARBONADO: MORE CLUES TO A COMMON IMPACT ORIGIN FOR SAMPLES FROM BRAZIL AND THE CENTRAL AFRICAN REPUBLIC

Brazilian and Central African carbonados can be matched via their carbon and nitrogen isotopic compositions to suggest a common origin. Features seen at the margins of the stones investigated may be interpreted as indicating very high temperature processing consistent with an impact origin. Carbonado is a form of polycrystalline diamond mined as a placer mineral in Brazil and known from other areas, particularly in Venezuela and the Ubangui region of the Central African Republic. A number of observations distinguish carbonado from kimberlite diamonds (1-4). The contradictory nature of these observations makes it difficult to assign an origin to carbonado. Nevertheless several hypotheses have been put forward including impact (5) transformation of carbon rich material as a result of radiation (6). In respect of the impact hypothesis (7), we recently suggested that a giant impact, proposed (8) to explain the magnetic anomaly spread over 700,000 krn2 of Central Africa, might have been responsible for the production of Ubangui and Brazilian carbonados as a single population which were subsequently separated by plate tectonics (7). To evaluate this possibility, we have investigated nitrogen abundance and its isotopes, with carbon isotopic composition for twelve samples of carbonados from Ubangui and Brazil, along with nine specimens of framesite (polycrystalline diamonds from the Jwaneng and Orapa (Botswana) kimberlite pipes); three shock produced diamonds from the Popagai crater and a sample of Yakutite (a distinct form of carbonado from Siberia). Our results are sumrnarised in the Table 1. There is essentially no difference between our 613C for Brazilian carbonados and the results obtained by Galimov (reviewed in 9) in a number of studies during the last 20 years. Based on our data, carbonados from Ubangui are indistinguishable from those collected in Brazil in respect of 6l3C, N abundance and 61%. However, one sample from Ubangui was so different from all the others (613C = -6%0 (repeated measurements) 615N = -22%0) as to be totally unrelated to the main dataset; the cause of this huge discrepancy is presently unknown. The similarity of the majority of the Brazilian and Ubangui carbonados, leads us to believe that our hypothesis favouring a Precambrian age impact is worthy of much more detailed consideration; a more statistical sample set is needed for analysis. We note that our Brazilian and Ubangui samples are easily discerned from Jwaneng and Orapa framesites, the other microcrystalline form of diamond common from African sources. Popagai crater diamonds and yakutites have not been studied yet in depth but initially appear distinct, certainly in respect of carbon isotopes. O Lunar and Planetary Institute Provided by the NASA Astrophysics Data System 1282 LPS XXVI Carbonado More Clues: Shelkov et al. One of the unusual features common to Brazilian and Ubangui carbonados, but not seen in our other specimens, is the presence of a surface rind with the texture and appearance of a fusion crust. We have now made an SEM study of a number of specimens broken and polished orthogonal to the surface layer. Fig. 1 shows the region immediately adjacent to the edge of the sample. Whereas the interior of the carbonado is characterised by a highly vesicular texture, the outer 200pm exhibits clear evidence that material was plastic to the point where voids and cavities were largely annealed out. Given the known properties of diamond, the environment in which the melting process occurred (if that is what caused the surface layer) must have been extremely high temperature and oxygen deficient. The former is obviously achievable under impact conditions. The Precambrian atmosphere could have been much less oxidising or conditions created during supersonic flight whereby low oxygen fugacity was obtained. The 613C and 615N values of all the carbonados studied here, save the single anomalous Ubangui specimen, fall in the range typical of organics. However none of the samples investigated has a N abundance of >I200 ppm; most are 100 ppm or less, hardly in keeping with a putative sedimentary deposit being the source for the carbon and nitrogen. Biologically derived material has a N content of several per cent, although a violent process, such as imuact. could cause severe N loss.