Browsing by Author "Erlank, Tony"

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Open Access
    A geochemical study of metasomatised peridotite and marid nodules from the Kimberley pipes, South Africa
    (1987) Waters, Frances Gillian; Erlank, Tony
    A comprehensive study has been made of a large collection of metasomatised peridotite, and MARIO (Mica-Amphibole-ButileIlmenite-Qiopside) xenoliths from the Kimberley pipes, with the aim of obtaining an improved understanding of enrichment processes operating in the sub-continental lithospheric mantle. The metasomatised peridotite suite is divided into garnet peridotites (GP), which contain no texturally equilibrated or primary phlogopite, garnet phlogopite peridotites (GPP), which constitute the most abundant peridotites, phlogopite peridotites (PP) and phlogopite K-richterite peridotites (PKP). Diopside can be present in all four groups. PKP rocks may also contain exotic incompatible element-rich Crtitanates such as lindsleyite and mathiasite, Nb-Cr rutile, ilmenite and armalcolite. Petrographic and chemical evidence presented here suggest that metasomatism increases progressively in the sequence GPGPP-PP-PKP, with the PKP group being richest in Fe, Ti and incompatible elements such as K, Na, Rb, Ba, Sr, Zr, Nb and the LR.EE. Mineral compositions change progressively from the garnet-bearing rocks to the PP and PKP rocks, showing decreases in Al 2o 3 and cr2o 3 contents, and increases in FeO and Tio2 contents. Data from this study show that most PKP and some PP rocks were derived from Al-deficient harzburgites. Other PP rocks probably had garnet-bearing precursors, as they display chemical similarities to GPP rocks, and typically contain aggregates of phlogopitediopside-spinel which are interpreted and modelled as being garnet replacement textures. Pressures and temperatures of equilibration of the peridotites in this study, combined with published experimental mineral stability data, suggest that metasomatism increases in intensity upwards in the subcontinental lithospheric mantle (SCLM) over a considerable depth interval from _170 km to Sr isotope data suggest that the metasomatism is young (1 Ga) have been recorded by Nd isotopes. New and published peridotite mineral and whole-rock Nd and Sr isotopic compositions range from moderately depleted to highly enriched and are interpreted in terms of mixing between variably enriched "ancient" SCLM and young metasomatic fluids with isotopic compositions close to Bulk Earth values. A rnineralogic expression of the ancient enrichments which might suggest that they resulted from older metasornatic events, cannot be clearly identified, but is best represented by enriched Nd-isotopic compositions of nonmetasornatic garnet. Rare phlogopites with low 1 43 Nd/144Nd ratios are interpreted as the product of complex mixing between the young metasomatic fluids and old enriched SCLM, rather than as older phlogopite. A group of unusually diopside-rich peridotites (±Phlogopite, ±ilmenite), are described and discussed. They are probably not directly related to the main group of metasomatised peridotites. Their chemical differences and greater pressures and temperatures of equilibration are consistent with formation by interaction between a diapir of asthenospheric melt and surrounding peridotitic mantle at the base of the SCLM. Comparison with chemical and isotopic data from the literature for Cr-rich "granny-smith" diopside rnegacrysts and glimmerites (phlogopite-diopside-ilmenite aggregates), suggests that they may have close genetic links with the diopside-rich suite. MARIO nodules are modally dominated by phlogopite, Krichterite and diopside, with lesser ilmenite and rutile, and accessory minerals such as calcite, barite and apatite. Olivine is absent, as is (with one exception) orthopyroxene, which serves to distinguish MARIO rocks from highly metasornatised peridotites. Relative modal proportions, textures and grain-size of MARIO rocks vary considerably, both within and between nodules. Textures suggest that they are igneous cumulates. The maximum stability depth of Krichterite restricts their depths of crystallisation to <120km. Data obtained in this study show that MARIO mineral compositions are Fe and Ti rich compared to most peridotites, and are much poorer in Cr, Al, Mg and Ni. Bulk compositions are alkali- and magnesian-rich (K20=4-9 weight%, Mg0=20-25 weight%), with moderate to high concentrations of i ncompatible trace elements. New and published MARIO mineral Nd and Sr isotopic measurements ranges from slightly depleted to highly enriched compositions. They are interpreted in terms of mixing of recent (phlogopite metasomatism is presented. The model proposes that there was recent input of metasomatic fluids at two distinct levels in the sub-continental mantle lithosphere. (1) Deep level fluids were generated at the base of the subcontinental lithosphere between 170-200 km depth - l l I ) possibly evolved from failed asthenospheric melts. The intensity of metasomatism progressively increased as these fluids percolated upwards to _100 km depth, and the GPP and some of the PP rocks were formed. (2) Shallow level metasomatic fluids were evolved during the crystallisation of MA.RID magmas ponded at 60-100 km depth. They were released into the surrounding SCLM, and formed the PKP and some PP rocks. These MA.RID-derived fluids appear to have overprinted the deep-derived metasomatism in places; some PP rocks have metasomatic signatures with characteristics of both events, which is interpreted as evidence for the passage of both types of fluid through them. Various lines of evidence including variations in mineral and whole-rock chemistry, suggest that both types of fluids contained Si, Al, Fetotal, Fe 3+, Ti, ±Ca, incompatible elements such as K, Na, Rb, Ba, Sr, Zr, Nb, V, LREE, s, and possibly F and Cl. The MARIO-derived fluids may have been generally richer in these elements, as the PKP rocks are more intensely metasomatised than even the most metasomatised GPP. u and related PP rocks. The behaviqr of Ba appears to have been different in the two sources/fluids - the deeper-derived fluids produced Ba-rich phlogopites, whereas the MARID magmas and MARID-derived fluids resulted in the formation of Ba-poor phlogopites in MA.RID and PKP rocks. Modelling of fluid compositions using published crystal/melt and crystal/fluid partition coefficients and PKP mineral trace element abundances is inconclusive but suggests that the MARIDderived fluids (and possibly the deep-derived fluids as well), were strongly LREE-enriched with REE patterns comparable to lamproites or kimberlites, and with moderately high Rb and Ba contents. However, inferred concentrations of Sr in the metasomatic fluids are of the order of 1% or more. The PKP rocks have attracted more attention in the literature than the GPP and PP groups, due to their spectacular J metasomatic assemblages. However, it is worth noting that they are relatively rare (~10% of peridotites from the Kimberley pipes). The ¥.LA.RID-derived metasomatism was probably intense, but strongly localised. In contrast the GPP and PP rocks are extremely abundant at Kimberley (_50% and _30% respectively) and were equilibrated over a large depth interval (170-100 km). The deep-derived metasomatism therefore appears to have been much more pervasive, and although less intense (most GPP rocks have 1-2% phlogopite) , it is considered to be a more significant phenomenon in the context of mantle metasomatic processes, especially as it appears to predate host kimberlite formation and emplacement.
  • Thumbnail Image
    Item
    Open Access
    Geochemistry of the Sabie River Basalt Formation in the central Lebombo, Karoo Igneous Province
    (1988) Sweeney, Russell James; Sweeney, Russell James; Erlank, Tony; Duncan, Andy; Erlank, Tony; Duncan, Andy
    The Sabie River Basalt Formation is a group of tholeiitic basaltic rocks erupted ca 190 Ma ago in the eastern zone of the Karoo Igneous Province of southern Africa. It is traceable over a distance of 700 km from Zululand, northwards along the Lebombo monocline into the Transvaal and south-east Zimbabwe. An abrupt compositional change in this formation occurs about halfway down its length in the vicinity of the Sabie and Komati Rivers: basalts to the north are known to be enriched in certain incompatible elements relative to basalts in the south, which are comparable in geochemistry to most basaltic rocks in the southern part of the Karoo Igneous Province. New data obtained in this work include 134 major and trace element whole-rock analyses, some 400 analyses of constituent minerals, 38 ⁸⁷Sr/⁸⁶Sr ratio determinations, 19 ¹⁴³Nd/¹⁴⁴Nd ratio determinations, 16 common Pb determinations and 12 oxygen isotope analyses. The "normal" (N) and "enriched" basaltic rocks are distinguished by differences in the concentrations of Ti, P, Zr, Nb, Y, La, Ce and Nd (high field strength elements). Broadly these differences are substantiated by K, Rb, Ba and Sr, but with much more overlap. The "enriched" group of basaltic rocks has been further subdivided into a low-Fe "enriched" (LFE) group and a high-Fe "enriched" group (HFE). The LFE-group basalts, which predominate at the base of the stratigraphic sections, are considered to be equivalent to basalts occurring in the N. Lebombo. In the central Lebombo N-group basalts predominate in the mid- and upper portions of the sections and HFE-group basalt occurs near the top of each section. Interbedding of all basalt groups occurs in the Sabie River section at the northern end of the study area, while the N- and HFE-group basalts are interbedded in the Crocodile and Komati River sections further to the south. The decrease in LFE-group basalt abundance southwards is accompanied by an increase in N-group basalt abundance. HFE-group basalts appear to be unique to the central Lebombo area of the Karoo Igneous Province and are volumetrically less significant than N- or LFE-group basalts. Petrogenetic models involving closed-system fractional crystallization; coupled assimilation (of granitic crust) fractional crystallization; replenished, tapped and fractionated magma chambers and partial melting are examined. Granitic crustal contamination appears to have been significant only in some samples of the N group where assimilation of granitic material has proceeded in a bulk fashion described by an AFC model. RTF models are dynamically more realistic than closed-system fractional crystallization models and explain increases in incompatible elements with decreasing MgO in the LFE and HFE groups. Variations in the N group, however, require varying degrees of partial melting of a N-type source to be explained fully. RTF models may explain the absence of any stratigraphic correlations of element abundances in the three groups. The HFE group may be related to an uncontaminated N-type parent composition by a combination of continued fractional crystallization from an N-group parent composition and varying degrees of partial melting of an N-type source. The only petrogenetic process by which the N and LFE groups may be related is different degrees of partial melting. However, this demands a source composition which has no resemblance on trace element and isotopic grounds, to observed mantle xenolith compositions. The preferred model is one in which the LFE group is derived from old sub-cratonic mantle similar to garnet-bearing "cold" peridotite xenoliths and the N group from a source similar in composition to estimates of primitive mantle. The existence of two types of mantle derived continental flood basalt magmas occurs in other Mesozoic basalt provinces in "southern" Gondwanaland (e.g. Kirwanveggan of Antarctica, Etendeka of Namibia and the Parana Basin of South America). It is suggested that there is a geographical association of LFE-type basalts with Archaean crust (or Archaean crust re-worked in low temperature - high pressure events) and N-type basalts with post-Archaean crust (or Archaean crust re-worked in high temperature - low pressure events). This model suggests the derivation of the LFE group, from old sub-cratonic lithospheric mantle relatively enriched in incompatible elements and the N group being derived from more recently accreted and less enriched lithospheric mantle underlying younger crustal terraines.
  • Thumbnail Image
    Item
    Open Access
    The geology and geochemistry of the Etendeka Formation quartz latites, Namibia
    (1988) Milner, Simon Charles; Milner, Simon Charles; Erlank, Tony; Duncan, Andy; Erlank, Tony; Duncan, Andy
    The Etendeka Formation volcanics of north-western Namibia form part of the Karoo Igneous Province of southern Africa and consist of a series of basalts interbedded with quartz latites and minor volumes of latite. This thesis examines various aspects of the geology and geochemistry of the quartz latites, in particular the volcanological and petrogenetic origin of these rocks . This study has involved the geological mapping of ca. 5000 km² of the southern Etendeka region and the documentation of the field and petrographic characteristics of the quartz latites. 183 whole rock quartz latite samples have been analysed for 32 elements by x-ray fluorescence spectrometry and 17 of these samples were selected for detailed mineral analyses by electron microprobe.