Craig B. Smith

U-Pb zircon age for a tuff in the Campbell Group, Griqualand West Sequence, South Africa: Implications for Early Proterozoic rock accumulation rates

An ion-microprobe U-Pb age of 2552 ±11 Ma has been obtained on zircon separated from a regional banded-tuff horizon in the Nauga Formation (Beukes, 1980b) (upper Campbell Group, Griqualand West Sequence, South Africa). This age permits time constraints to be placed on lithologically correlated units within the adjacent Transvaal Sequence and correlations to be made with the further removed Hamersley Group in Australia. Calculated rock accumulation rates of 2 to 4 m/m.y. for these strata of mainly shale and banded iron formation suggest that sedimentation rates were significantly slower in the late Archean-Early Proterozoic than is generally assumed.

A feasibility study of U−Pb and Pb−Pb dating of kimberlites using groundmass mineral fractions and whole-rock samples

Abstract This paper describes partly successful attempts to determine emplacement ages of kimberlites by U−Pb and Pb−Pb methods, involving groundmass minerals with high U content (notably perovskite) and whole-rock kimberlite samples. U/Pb ratios in perovs-kite in the matrix of kimberlites can be two orders of magnitude larger than in the rest of the kimberlite material, and with simple mineral separation techniques moderate success was achieved in U−Pb dating of fresh samples of younger kimberlites (around 100 Ma). The differences in U/Pb ratios between kimberlite samples from different parts of the same pipe have also been found to be large enough, in some cases, to allow reasonably accurate U−Pb age determination. In older kimberlites the U−Pb ages obtained were mostly incompatible with geological constraints and results obtained by other methods. However, for such pipes use of Pb−Pb systematics yields realistic age limits in some cases.

Emplacement ages of kimberlite occurrences in the Prieska region, southwest border of the Kaapvaal Craton, South Africa

Abstract RbSr (mica) and UPb (perovskite) emplacement ages for kimberlites from the southwest part of the Kaapvaal Craton, near Prieska, South Africa, vary from as old as 170 Ma to as young as 74 Ma, similar to ages elsewhere in southern Africa. Isotopically defined Group-I and -II kimberlites are found on the Kaapvaal Craton northeast of the Doornberg and Brakbos lineaments (Domains I and II as defined by Skinner et al., 1992), with ages of 74, 108 and 118–125 Ma. Kimberlites in the craton margin area bounded by the two lineaments (Domain III) are exclusively Group II, with ages of 116–119 Ma. Kimberlites in Domain IV, to the south of the craton boundary, are of Group-I affinity emplaced in two episodes at 74 and 100–103 Ma. Southernmost Domain V is dominated by kimberlites petrographically, chemically and isotopically transitional between Group-I and -II types, but with the oldest ages at 140 and possibly 170 Ma. Phlogopite from a peridotite xenolith from the Sanddrift kimberlite has retained an age of 2250 Ma, presumably the enrichment age of the nodule.

STABLE AND RADIOGENIC ISOTOPE STUDY OF ECLOGITE XENOLITHS FROM THE ORAPA KIMBERLITE, BOTSWANA

Abstract Eclogite xenoliths from Orapa can be accurately classified as Group I or Group II on the basis of Na 2 O in garnet and K 2 O in clinopyroxene. Group I xenoliths are commonly diamondiferous while Group II xenoliths are diamond-free. Both xenolith varieties may contain graphite. Isotopic character is to some degree correlated with major- and trace-element chemistry. Group II samples with Ca-poor garnet have clinopyroxenes with radiogenic 87 Sr 86 Sr (0.705–0.709) and the least radiogenic 143 Nd 144 Nd (0.5122–0.5125). Group I eclogites with higher Ca and Fe in garnets have less radiogenic 87 Sr 86 Sr (0.702–0.7066) and bulk-Earth or higher 143 Nd 144 Nd ratios. Group I eclogites have more radiogenic 206 Pb 204 Pb (18.6–19) than Group II xenoliths (16.5–18.6). In contrast, Group II xenoliths have more variable and, in some cases, more radiogenic 206 Pb 204 Pb (36.6–39.3) than Group I xenoliths (38.3–38.4). The Sr, Sm, Nd and Pb concentrations of minerals in Orapa Group I eclogite xenoliths are much lower than in Group II samples. All the Group II xenoliths are inferred to be enriched in light rare-earth elements while Group I xenoliths are probably characterised in many cases by light rare-earth element depletion. Constituent garnet and clinopyroxene in both Group I and II eclogite xenoliths are essentially in isotopic equilibrium at the time of pipe emplacement. Mineral as well as calculated whole-rock 143 Nd 144 Nd compositions of most of the Group I eclogites are too close to bulk-Earth and depleted-mantle estimates in order to obtain useful model age information. Depleted-mantle model ages derived from the much lower 143 Nd 144 Nd compositions of the Group II eclogite xenoliths range from 661 to 1248 My, with an average clinopyroxene model age of 908 My and an average whole-rock model age of 1016 My. On the basis of an observed covariation of O and Sr isotopic compositions the entire Orapa Group I eclogite xenolith suite can be modelled as mixtures of oceanic basalt with or without a few percent of ocean floor sediment. The Group II xenoliths might have crystallised from a melt which derives from a protolith with time-averaged LREE depletion. Their radiogenic Sr isotope character could be due to interaction of the melt with metasomatised lithosphere, or might be a superimposed metasomatic signature.

Note on the UPb perovskite method for dating kimberlites: Examples from the Wesselton and De Beers mines, South Africa, and Somerset Island, Canada

Abstract The previously reported UPb perovskite method of dating kimberlite has been substantially improved by obtaining high-purity perovskite concentrates using a combination of chemical, magnetic and hand-picking techniques. Reliable results are obtained from subhedral to euhedral perovskite, but not from anhedral perovskite commonly displaying atoll-like grain shapes. The latter may be weathered or may have been partially digested in the chemical treatment used in the initial concentration step. An age of 89 Ma was obtained from the Wesselton kimberlite and one of two De Beers Mine samples, while a second De Beers sample gave a younger age of ∼ 83 Ma. These ages are in agreement with previous age determinations by various other methods. The ThPb age for perovskite from De Beers Mines is 81 Ma. An age of ∼ 105 Ma for a Somerset Island kimberlite is in agreement with a RbSr errorchron age of ∼ 100 Ma for micas from a second Somerset Island occurrence, but a perovskite age of 40 Ma for a third locality is not realistic. Fresh kimberlitic perovskite has U and Pb concentrations of ∼ 150 and 15 ppm, respectively. Th concentrations fresh perovskite are of the order of 1000 ppm, though this is based on only one sample from the De Beers Mine. 87 Sr/ 86 Sr ratios and presumably 143 Nd/ 144 Nd ratios are easily obtained from groundmass perovskites, and reflect the initial Sr and Nd isotopic compositions of kimberlite at the time of emplacement.

Improved precision of RbSr dating of kimberlitic micas: An assessment of a leaching technique

Abstract RbSr isotopic analysis of phlogopite micas has commonly been used to determine the emplacement ages of kimberlite and other alkaline intrusives. Leaching of fresh as well as partially altered phlogopite samples in dilute (∼ 2 N) HCl, prior to isotopic analysis, increases the precision of the calculated ages significantly without disturbing the RbSr isotope systematics of the micas. The leaching step removes trace amounts of carbonate, which typically has a high Sr concentration (2000–9000 ppm), from the mica separates and thus increases the 87 Rb/ 86 Sr and 87 Sr/ 86 Sr ratios of the analysed samples. Leaching for a period of 10 min. had no affect on the Rb concentrations of the analysed phlogopite samples. However, leaching for a period of 12 hr. lowered the Rb contents of the phlogopites by a few per cent. All but one of the phlogopite samples leached for 12 hr. plot on an isochron. The slight loss in Rb content of these samples was thus complemented by loss of radiogenic Sr, which occurred along with the Rb in the large interlayer cation sites of the mica lattice. The emplacement age of the Makganyene kimberlite pipe is 121.2 ± 1.2 Ma, given by a RbSr isochron, including 19 phlogopite samples and one whole-rock sample. The initial 87 Sr/ 86 Sr ratio of 0.7100 ± 0.0002, in conjunction with the phlogopite mineralogy, indicates that the kimberlite is a Group-II kimberlite.

Ultramafic rocks in the centre of the Vredefort structure (South Africa): Possible exposure of the upper mantle?

Abstract The rocks in the central region of the Vredefort structure are interpreted as lower-crust that has undergone uplift of > 36 km during the Vredefort cryptoexplosion event 2.0 Ga ago. The outcrops in this central region are mainly high-grade leucogranulites interbanded with mafic granulites and supracrustals. Rocks from a borehole located near the centre of the structure indicate that the core region is underlain by ultramafic rocks which probably represent the upper mantle beneath this region of the Kaapvaal craton. The main lithologies sampled by the borehole are serpentinized hornblende-bearing harzburgite with minor bands of pyroxenite and glimmerite. The major- and trace-element chemistry of these ultramafic rocks is similar to that of fertile oceanic lithosphere. Mineral chemistry and petrographic data suggest that the Vredefort ultramafics are metamorphic rocks which have undergone a complex tectonic and metamorphic history. SmNd and RbSr isotopic data (together with the field relationships) provide some evidence that these rocks are Archaean but also that they have been chemically and isotopically altered. Similar hydrated and chemically altered ultramafic rocks occur in the Jamestown ophiolite complex in the Barberton greenstone belt. Accordingly, we speculate that the crust-mantle transition zone in the Kaapvaal craton consists of a zone of hydrated Archaean oceanic lithosphere.

Fluoride precipitates in silicate wet-chemistry: implications on REE fractionation

Abstract Precipitates of hieratite, ralstonite and NaUZr2F12 form during HFHNO3HClO4 dissolution of dolerite samples. These minerals are responsible for fractionation of Sm and Nd from the solution, giving rise to spurious analytical results. Fractionated Sm and Nd can be recovered by dissolution of hieratite in hot (373 K) water or HCl. The presence of ralstonite and NaUZr2F12 remains a problem, as these phases do not dissolve in acids or in hot water. However, their formation could be prevented by the addition of a complex-forming agent such as boric acid during the dissolution procedure. The use of a m-benzene disulphonic acid and hydrofluoric acid mixture is recommended during the dissolution of silicate rocks and minerals for purposes of routine analysis of trace elements.

Mass spectrometry in geochemistry: An established tool in the earth sciences

Abstract Thermal ionization mass spectrometry is an established and important tool in the earth sciences. Precise measurements of radiogenic isotope abundances have allowed the development of useful geochronometers and petrogenetic tracers. Further advances in isotope geochemistry will, in part, depend on the capability of analyzing smaller and smaller samples. Techniques highly suitable for small sample/high resolution analysis such as SIMS or AMS will therefore become increasingly more important in the earth sciences. Nevertheless, because of continued instrumentation improvements in conventional mass spectrometry, and the probable high costs and lower accessibility of SIMS/AMS techniques, conventional techniques will remain of major importance in the foreseeable -future.