Journal of Petrology Advance Access originally published online on June 2, 2006
Journal of Petrology 2006 47(9):1751-1783; doi:10.1093/petrology/egl026
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Magmatic Evolution and Ascent History of the Aries Micaceous Kimberlite, Central Kimberley Basin, Western Australia: Evidence from Zoned Phlogopite Phenocrysts, and UV Laser 40Ar/39Ar Analysis of Phlogopite–Biotite
1 WESTERN AUSTRALIAN MUSEUM, FRANCIS ST PERTH, WESTERN AUSTRALIA, 6000, AUSTRALIA
2 JOHN DE LAETER CENTRE OF MASS SPECTROMETRY, DEPARTMENT OF APPLIED GEOLOGY, CURTIN UNIVERSITY OF TECHNOLOGY HAYMAN ROAD, BENTLEY, WESTERN AUSTRALIA, 6102, AUSTRALIA
3 CENTRE FOR MICROSCOPY AND MICROANALYSIS, UNIVERSITY OF WESTERN AUSTRALIA CRAWLEY, WESTERN AUSTRALIA, 6009, AUSTRALIA
RECEIVED SEPTEMBER 27, 2004; ACCEPTED APRIL 3, 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHYRESULTSWHOLE-ROCK GEOCHEMISTRY40AR/39AR GEOCHRONOLOGYDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
The Neoproterozoic Aries kimberlite was emplaced in the central Kimberley Basin, Western Australia, as a N–NNE-trending series of three diatremes infilled by lithic-rich kimberlite breccias. The breccias are intruded by hypabyssal macrocrystic phlogopite kimberlite dykes that exhibit differentiation to a minor, high-Na–Si, olivine–phlogopite–richterite kimberlite, and late-stage macrocrystic serpentine–diopside ultramafic dykes. Mineralogical and geochemical evidence suggests that the high-Na–Si, olivine–phlogopite–richterite kimberlite was derived from the macrocrystic phlogopite kimberlite as a residual liquid following extended phlogopite crystallization and the assimilation of country rock sandstone, and that the macrocrystic serpentine–diopside ultramafic dykes formed as mafic cumulates from a macrocrystic phlogopite kimberlite. Chemical zonation of phlogopite–biotite phenocrysts indicates a complex magmatic history for the Aries kimberlite, with the early inheritance of a range of high-Ti phlogopite–biotite xenocrysts from metasomatized mantle lithologies, followed by the crystallization of a population of high-Cr phlogopite phenocrysts within the spinel facies lithospheric mantle. A further one to two phlogopite–biotite overgrowth rims of distinct composition formed on the phlogopite phenocrysts at higher levels during ascent to the surface. Ultra-violet laser 40Ar/39Ar dating of mica grain rims yielded a kimberlite eruption age of 815·4 ± 4·3 Ma (95% confidence). 40Ar/39Ar laser profiling of one high-Ti phlogopite-biotite macrocryst revealed a radiogenic 40Ar diffusive loss profile, from which a kimberlite magma ascent duration from the spinel facies lithospheric mantle was estimated (assuming an average kimberlite magma temperature of 1000°C), yielding a value of
0·23–2·32 days for the north extension lobe of the Aries kimberlite. KEY WORDS: 40Ar/39Ar; diamond; kimberlite; mantle metasomatism; phlogopite–biotite
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHYRESULTSWHOLE-ROCK GEOCHEMISTRY40AR/39AR GEOCHRONOLOGYDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
The Kimberley region of northern Western Australia preserves an extended record of alkaline mafic to ultramafic magmatism dating from Mesoproterozoic to Tertiary times (Jaques et al., 1986
). Kimberlites, lamproites, ultramafic lamprophyres and carbonatites intrude the Kimberley Craton, and the surrounding Proterozoic King Leopold and Halls Creek Orogens (Fig. 1). Apart from the eruption of the Argyle lamproite at 1178 ± 47 Ma (Rb–Sr; Pidgeon et al., 1989), alkaline magmatism in the region falls into two major episodes: (1) an 800 Ma event that produced the North Kimberley Province kimberlites, kimberlites and ultramafic lamprophyres in the East Kimberley Province, and the Aries kimberlite in the Central Kimberley (Edwards et al., 1992); (2) a 22–18 Ma phase that produced the extensive West Kimberley lamproite field (Jaques et al., 1986) and lamproites in the Bonaparte Gulf (Gorter & Glikson, 2002).
|
The Aries micaceous kimberlite was emplaced during the Neoproterozoic phase of alkaline magmatism and has been previously dated at 820 ± 20 Ma (Rb–Sr on a mica separate from diamond drill-hole PRBH 45 in the north extension lobe; Smith, 1989
). It has close mineralogical affinities to Group 2 kimberlites (orangeites) and olivine lamproites, but exhibits distinctive geochemical characteristics, notably the Nd–Sr isotopic signature of a Group 1 kimberlite (Edwards et al., 1992; Taylor et al., 1994). Initial studies of the Aries kimberlite included descriptions of its mineralogy, petrology and geochemistry (Edwards, 1990; Edwards et al., 1992), exploration history and evaluation (Towie et al., 1994) and indicator mineral chemistry (Ramsay et al., 1994). Since that time, an improved understanding of the geology of the Aries kimberlite has been made possible through further sampling of each of the kimberlite lobes by diamond drilling, which has produced over 7000 m of drill-core material that extends to a maximum depth of >300 m. There is, thus, a more extensive collection of material compared to the previous eight drill-cores that only penetrated to 
100 m (Edwards et al., 1992), which forms the basis of this study. Throughout this paper, the terms ‘macrocryst’, ‘phenocryst’ and ‘xenocryst’ are used, respectively, to refer to crystals of uncertain, cognate and foreign origin to their host.
One of the interesting features of the phlogopite-rich, hypabyssal phases of the Aries kimberlite is the presence of phlogopite phenocrysts with complex chemical zonation involving up to four stages of growth (Edwards et al., 1992). Such compositional zonation in phenocrysts has provided important insights into magma evolution for a range of different magma compositions, ranging from rhyodacite to basalt (e.g. Pearce et al., 1987; Singer et al., 1993, 1995; Nakagawa et al., 2002), phonolites and minettes (O'Brien et al., 1988), lamproites (Toscani et al., 1995) and micaceous kimberlites (Hunter et al., 1984; Neal & Taylor, 1989; Wiese et al., 1996). Studies of this type for micaceous kimberlite magmas are relatively rare, probably because kimberlites that preserve such a complex record of phlogopite growth are themselves rare. Thus, the availability of new samples from the hypabyssal phases of the Aries kimberlite allows a more detailed examination of this phenocryst zonation in relation to magma evolution.
Among the new phlogopite–biotite parageneses in the hypabyssal phases of the Aries kimberlite that are recognized here, we describe high-Ti phlogopite–biotite macrocrysts that are commonly overgrown by later generations of phlogopite and may form the cores of some phlogopite phenocrysts. These macrocrysts are similar in texture and composition to aluminous biotite macrocrysts from Group 2 kimberlites (orangeites), and other micaceous kimberlites from South Africa, the USA and Canada (‘type 1’ micas of Smith et al., 1978; Hall et al., 1989; Mitchell & Meyer, 1989; Mitchell, 1995; Wiese et al., 1996). The origin of these aluminous biotite macrocrysts is uncertain, but they may have been derived from mixing with other mantle-derived magmas or through inheritance from mantle or crustal lithologies (Smith et al., 1978; Mitchell, 1995). Mitchell (1995) suggested that the macrocrysts were distinctive of Group 2 kimberlite (orangeite) magmas, or their source regions in the Kaapvaal Craton. In our study, we test these hypotheses by examining a possible link between the Aries macrocrysts and high-Ti phlogopite–biotite in associated metasomatized ultramafic xenoliths.
An estimate of ascent duration for the Aries kimberlite magma has been obtained by high spatial resolution ultra-violet (UV) laser 40Ar/39Ar analysis of a radiogenic 40Ar (40Ar*) diffusive loss profile in one of these high-Ti phlogopite–biotite macrocrysts. By integrating the paragenesis of the high-Ti phlogopite–biotite macrocrysts into the crystallization history of the magma, as recorded in the various phlogopite growth zones, combined with the estimate of the duration of ascent, it is possible to track the ascent path of the Aries kimberlite magma. Determining the ascent history of a kimberlite magma from its depth of origin to the surface is a task that is rarely achieved, but may have implications for the preservation of diamonds in the kimberlite (cf. Franz et al., 1996; Pearson et al., 1997; Canil & Fedortchouk, 1999; Wartho & Kelley, 2003).
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHYRESULTSWHOLE-ROCK GEOCHEMISTRY40AR/39AR GEOCHRONOLOGYDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
The Kimberley Craton is covered by an extensive Palaeoproterozoic sedimentary sequence that comprises the Kimberley and Speewah Basins (Fig. 1). The basement to the Kimberley and Speewah Basins generally is not exposed at the surface; however, geophysical and geochronological data suggest that the shallow basement comprises a late Archaean to early Proterozoic granitic terrain that is composed of NE-trending magmatic arcs and continental fragments (Gunn & Meixner, 1998
; Tyler et al., 1999). This terrain may have formed during Palaeoproterozoic NW-dipping subduction beneath the Kimberley Craton, prior to the eventual collision and amalgamation of the Kimberley Craton with the remainder of the North Australian Craton at 1820 Ma (Myers et al., 1996; Tyler et al., 1999; Griffin et al., 2000). The early Proterozoic age of emplacement of some of the granites within the basement (1867 Ma; Gunn & Meixner, 1998) correlates with the extensive 1865–1850 Ma Whitewater Volcanics and granites of the Paperbark Supersuite that occur in the Hooper and Lamboo Complexes within the King Leopold and Halls Creek Orogens bounding the Kimberley Basin to the southwest and east respectively (Fig. 1). Granite emplacement may have been related to Palaeoproterozoic subduction beneath the region (Gunn & Meixner, 1998); however, Griffin et al. (2000) and Sheppard et al. (2001) placed the Whitewater Volcanics and the Paperbark Supersuite granites within a post-collisional tectonic setting that pre-dated the amalgamation of the Kimberley Craton with the North Australian Craton. They suggested that this felsic magmatism resulted from the collision of one or more continental fragments with the eastern margin of the Kimberley Craton, producing the temporary termination of NW-directed subduction, followed by slab collapse, lithospheric delamination, and granite production in an extensional regime (Griffin et al., 2000; Sheppard et al., 2001).
Overlying the granitic basement, the Kimberley and Speewah Basins contain gently folded to flat-lying Palaeoproterozoic sequences (Griffin et al., 1993). The Aries kimberlite intrudes the Kimberley Group within the Kimberley Basin succession on the northern edge of the Phillips Range. Aries forms a series of three diatremes, trending N–NNE, that are infilled by massive, clast-supported, lithic-rich, volcaniclastic kimberlite breccias, and intruded by associated hypabyssal kimberlite dykes and plugs (Fig. 2). The current level of exposure of the diatremes is right on the boundary between the King Leopold Sandstone and the overlying tholeiitic basaltic lavas that comprise the Carson Volcanics (Roberts & Perry, 1969). Country rock debris from the King Leopold Sandstone and Carson Volcanics is a major component of the lithic-rich volcaniclastic breccias within the diatremes.
|
|
ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHYRESULTSWHOLE-ROCK GEOCHEMISTRY40AR/39AR GEOCHRONOLOGYDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Electron beam techniques
Mineral analyses were carried out on a JEOL 6400 Scanning Electron Microscope (SEM) with an Oxford instruments ISIS Energy Dispersive Spectroscopy (EDS) X-ray analytical attachment, and a JEOL 6400R electron microprobe, with an integrated Wavelength Dispersive Spectroscopy (WDS)–EDS X–ray analytical system, at the Centre for Microscopy and Microanalysis at the University of Western Australia. The operating conditions for EDS analysis include an accelerating potential of 15 kV, a beam current of 5·0 nA, a spot size of 1–2 µm and a count-time of 50 s. Peaks were calibrated to a copper standard. The operating conditions for WDS analysis include an accelerating potential of 15 kV, a beam current of 20 nA and a spot size of 1–2 µm. Element peaks were calibrated to standards that included well characterized minerals, pure metals and synthetic compounds. Analyses involved 40 s count-times on peaks and 10 s count-times on upper and lower backgrounds.
Whole-rock geochemical techniques
Whole-rock major and trace-element geochemistry was carried out by Australian Laboratory Services, Perth, Western Australia. Suitable samples of fresh hypabyssal kimberlite were selected from drill-core, crushed and pulverized. The concentrations of major and trace elements in each sample were measured using Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES) and Inductively Coupled Plasma Mass Spectrometry (ICP-MS) techniques, respectively. Fluorine content was determined by electrochemical techniques using a fluoride selective-ion electrode.
Laser ablation–inductively coupled plasma mass spectrometry (LA–ICP-MS)
Trace-element microanalysis of phlogopite–biotite by LA–ICP-MS was carried out on individual crystals in 
120 µm thick sections at the Research School of Earth Sciences, Australian National University. A Hewlett Packard Agilent 7500 ICP-MS instrument coupled to an ArF Excimer UV laser ablation microprobe, operating at a wavelength of 193 nm, was used. A detailed description of the techniques employed is outlined in Eggins et al. (1998). Each analysis involved the measurement of backgrounds for 20 s followed by the ablation of the laser spot for 40 s, for a total period of 60 s. A recovery time was allowed to enable washout of the sample chamber between each analysis. Analyses were monitored for possible compositional heterogeneties encountered during laser ablation by the collection of data in a time-resolved graphics mode. Standards and samples were ablated using pulse rates of 5 Hz and beam energies of 100 mJ per pulse. Spot diameters of 23 µm were used. A NIST 612 glass standard was used to obtain background and calibration measurements after every 10 analyses. Each analysis was normalized using TiO2, which was independently determined by SEM-EDS analysis, as an internal standard.
40Ar/39Ar laser dating
The 40Ar/39Ar analyses were undertaken at the Western Australian Argon Isotope Facility, operated by a consortium consisting of Curtin University and the University of Western Australia.
Polished thick sections of 100–200 µm, mounted in Canada Balsam resin, were removed from the resin and slides by soaking in acetone. The samples underwent ultrasonic treatment in methanol and subsequently deionized water. Approximately 10 x 10 mm sections containing mica grains of interest were broken off from the polished thick sections. Samples were individually wrapped in aluminium foil, and all the samples were loaded into an aluminium package, with biotite age standard Tinto B (Rex & Guise, 1995), with a K–Ar age of 409·24 ± 0·71 Ma, being loaded at 5 mm intervals along the package to monitor the neutron flux gradient. The package was Cd-shielded and irradiated in the 5C position of the McMaster University Nuclear Reactor, Hamilton, Canada, for 90 h. Upon return, the samples were loaded into an ultra-high vacuum laser chamber with a Suprasil 2 viewport and baked to 120°C overnight to remove adsorbed atmospheric argon from the samples and chamber walls.
A New Wave Research LUV 213X 4 mJ pulsed quintupled Nd–YAG (yttrium–aluminium garnet) 213 nm UV laser, with a variable spot size of 10–350 µm, and a repetition rate of 10 Hz, was used to ablate the mineral grains. The laser was fired through a Merchantek computer-controlled X–Y–Z sample chamber stage and microscope system, fitted with a high-resolution charge coupled device camera, 6x computer controlled zoom, high magnification objective lens and two light sources for sample illumination. The gases released by laser ablation were cleaned using three SAES AP10 getter pumps to remove all active gases (e.g. CO2, H2O, H2, N2, O2, CH4). Remaining noble gases were equilibrated into a high sensitivity mass spectrometer (MAP 215–50), operated at a resolution of 600 and fitted with a Balzers SEV 217 multiplier. The automated extraction and data acquisition system was computer controlled, using a LabVIEW program. The mean 5 min extraction system blank Ar isotope measurements obtained during the experiments were 3·6 x 10–12, 1·1 x 10–14, 6·1 x 10–15, 1·0 x 10–13 and 3·1 x 10–14 cm3 STP (standard temperature and pressure) for 40Ar, 39Ar, 38Ar, 37Ar and 36Ar, respectively. Samples were corrected for mass spectrometer discrimination (277·4) and nuclear interference reactions (39Ar/37ArCa – 0·00065, 36Ar/37ArCa – 0·000255 and 40Ar/39ArK – 0·0015). Errors quoted on the individual UV laser spot ages are 2
and 40Ar/39Ar ages were calculated using the decay constant quoted by Steiger & Jäger (1977). J values (±0·5%) were 0·019011 for AN15-99·55m, 0·018999 for AN65-36·82m, 0·018988 for AR61, 0·018976 for AR51 and 0·018965 for AN69-59·65m.
The 40Ar/39Ar ages of one high-Ti phlogopite–biotite 40Ar* loss profile were converted to normalized 40Ar*/39Ar ratios using the average 951–952 Ma grain core age (i.e. a value of 1) and the 815·4 Ma grain rim/eruption age (i.e. a value of 0) and plotted against the normalized distances from the rim of the grain. A diffusion curve was fitted through the 40Ar* loss profile datapoints, assuming a cylindrical diffusion geometry, which is appropriate for micas (Hames & Bowring, 1994), and using the Ar diffusion parameters of phlogopite (Giletti, 1974) and biotite (Grove & Harrison, 1996), using the method described by Wartho & Kelley (2003). In general, all the mica grains 40Ar/39Ar analysed in these samples were small (e.g. 188–1185 µm)—much smaller than mica grains analysed in previous studies in southern Africa, the Solomon Islands and Russia (Pearson et al., 1997; Kelley & Wartho, 2000; Kempton et al., 2001; Wartho & Kelley, 2003).
|
PETROGRAPHY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHYRESULTSWHOLE-ROCK GEOCHEMISTRY40AR/39AR GEOCHRONOLOGYDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
The mineralogy and petrology of the Aries kimberlite was originally described in detail by Edwards et al. (1992)
using material obtained from eight diamond drill-holes that had sampled the kimberlite to depths of 100 m. Their petrographic and mineralogical descriptions were based on samples from the central (PRBH 5, PRBH 17), north (PRBH 1, PRBH 2) and north extension (PRBH 45) lobes. Samples from the South Lobe diatreme were not described because the hypabyssal lithologies were heavily altered. Edwards et al. (1992) recognized four lithologies: (1) uniformly textured medium-grained macrocrystal rock (Lithology ‘M’—‘Medium’) that comprises 30–40% olivine macrocrysts and microphenocrysts in a uniformly textured, phlogopite-rich groundmass that includes apatite, titanite, spinels and deuteric minerals; (2) uniformly textured aphanitic macrocrystal rock (Lithology ‘A’—‘Aphanitic’), consisting of 5% olivine macrocrysts and microphenocrysts in a fine-grained, phlogopite-rich groundmass that includes interstitial calcite and serpentine; (3) macrocrystal segregated rock (Lithology ‘S’—‘Segregated’), a phlogopite-rich lithology that includes irregularly shaped groundmass segregations of calcite and serpentine; (4) breccias rich in country-rock xenoliths. On the basis of our new petrological studies, we have combined the first three lithologies into one hypabyssal lithofacies, termed a ‘macrocrystic phlogopite kimberlite’ (MPK); however, we also recognize a sub-facies—a globular segregationary-textured macrocrystic kimberlite (SK), in which a globular segregationary-textured groundmass is well developed. This texture can be easily recognized by macroscopic examination of the drill-core.
Edwards et al. (1992) also described the chemical zonation of phlogopite phenocrysts from the hypabyssal lithologies. Rare phlogopite phenocrysts preserve up to four distinct growth zones (zones 1–4) from the core to the rim, which can be distinguished on the basis of major-element chemistry (see discussion below). Most phenocrysts only have cores of zone 2 composition overgrown by zone 4 rims (Fig. 3a). In this study, we have recognized a number of new varieties of mica from the MPK, and several mantle and crustal xenoliths, and explore the implications of this phenocryst zonation for the magmatic evolution of the kimberlite.
|
Samples of hypabyssal kimberlite and altered mantle and crustal xenoliths that were used for petrographic examination and microanalysis in this study came from diamond drill-holes (DDH) AN 35, 54, 55 in the central lobe; DDH AN 1, 2, 15, 29, 58, 68 in the north lobe; and DDH AN 4, 5, 65, 69, 70 in the north extension lobe (Fig. 2). These samples range in depth from 25 to 285 m. Within the three diatremes, units of hypabyssal kimberlite most often preserve fresh minerals, free from deuteric or hydrothermal alteration, or oxidation because of weathering, which are suitable for microanalysis. Units of volcaniclastic kimberlite are commonly strongly hydrothermally altered or oxidized, and their primary mineralogy is only rarely preserved. Thus, we concentrated on the mineralogy of the relatively fresh hypabyssal kimberlite intrusions in this study. No samples of hypabyssal kimberlite from the south lobe diatreme proved to be suitable for microanalysis because of the degree of weathering and alteration in this lobe.
The following sections describe the different hypabyssal kimberlite and xenolith lithologies encountered in the Aries diatremes.
Macrocrystic phlogopite kimberlite (MPK), macrocrystic kimberlite (MK), macrocrystic kimberlite breccia (MKB) and globular segregationary-textured macrocrystic kimberlite (SK)
Hypabyssal kimberlite dykes and plugs, with widths ranging from several centimetres to 70 m, intrude each of the lobes of the Aries kimberlite (Fig. 2). Macrocrystic phlogopite kimberlite (MPK) dykes generally have sharp intrusive contacts and cross-cutting relationships with the surrounding volcaniclastic breccias and country rocks. Smaller dykes and stringers of MPK, generally tens of centimetres to <1 m wide, may be highly irregular and discontinuous in form. Hypabyssal MPK has an inequigranular or macrocrystic texture, in which 20 mm olivine macrocrysts and 4 mm phenocrysts comprise 50% of the rock. The olivine content is generally low, consisting of <10% macrocrysts and <20% phenocrysts. MPK also contains a population of euhedral–subhedral phlogopite phenocrysts, ranging up to 10% in abundance and up to 1·5 mm long (Fig. 3a–d, f, j). These macrocrysts and phenocrysts are set in a variably fine- to coarse-grained phlogopite-rich groundmass composed of randomly orientated phlogopite laths up to 3 mm long in coarse-grained zones. The groundmass also contains variable amounts of spinel (magnesiochromite to magnetite; 30% in abundance), serpentine, talc, calcite, diopside ± fluorapatite, along with a diverse accessory mineral assemblage that is detailed in Table 1. The margins of the MPK dykes are commonly aphanitic, containing <5% olivine macrocrysts. Aphanitic zones at dyke margins are generally <8 cm wide, but can be up to 28 cm in width. Olivine-rich bands and aphanitic zones, up to 40 cm wide, also occur within dykes away from their margins.
|
The groundmass of MPK contains widespread segregationary textures, and is transitional to globular segregationary-textured kimberlite (SK) in some regions. In general, groundmass segregations in MPK are irregularly shaped areas, up to
10 mm wide, infilled with serpentine, calcite, phlogopite and/or fluorapatite, with or without minor diopside, magnetite, ilmenite–pyrophanite, pyrite, chalcopyrite and/or galena (Fig. 4a). These segregations have relatively sharp boundaries with the surrounding uniform-textured groundmass. Crystals of apatite and/or phlogopite, each 1 mm long, commonly intrude from the margins of segregations into their interiors. A sample of MPK from the north lobe (AN15-124·04m) contains segregated veinlets and patches in the groundmass (<8 mm long), that contain <1 mm long inclusion-rich crystals of diopside, 0·3 mm long subhedral, equant crystals of andradite, serpentine, apatite, magnetite (0·1 mm long), calcite and rare phlogopite (Fig. 4d).
|
Macrocrystic kimberlite (MK) refers to hypabyssal kimberlite that has a macrocrystic texture, but in which the primary mineral assemblage of macrocrystic phlogopite kimberlite has been extensively altered and replaced by quartz, serpentine–talc, chlorite, clay and/or carbonate. Hypabyssal MPK or MK that contains
15% country-rock xenoliths is classified as macrocrystic kimberlite breccia (MKB).
Olivine–phlogopite–richterite kimberlite (OPRK)
Massive hypabyssal MPK exposed in the north extension lobe varies in composition to an olivine–phlogopite–richterite kimberlite (OPRK) adjacent to the walls of the lobe (Fig. 2). The OPRK contains potassian richterite, diopside and/or sanidine as fine-grained granular groundmass phases, and may include 1 mm long euhedral–subhedral poikilitic potassian richterite laths. Similar to MPK, OPRK contains up to 10% phlogopite phenocrysts by volume (<0·5 mm long). The groundmass of OPRK is commonly fine-grained and varies from phlogopite-rich to diopside–amphibole-rich on a millimetre scale. OPRK is also distinguished by the morphology of the contained olivine macrocrysts and phenocrysts; the 5 mm diameter macrocrysts are typically sparse (<5% in abundance), and commonly exhibit resorption features such as heavily embayed margins; the abundant (<30%) 2 mm long euhedral phenocrysts may form twins and complex saw-tooth crystal groups. As there is a lack of sharp, well defined contacts between OPRK and MPK, the relative timing of intrusion could not be defined, and the precise relationship between the two units remains uncertain, but may be gradational.
The texture of the groundmass in OPRK varies from uniform to segregationary. Groundmass segregations range from 1 to 12 mm in length, and contain varying proportions of 1 mm long amphibole, phlogopite, fluorapatite and/or diopside crystals, surrounded by serpentine, sanidine, calcite and/or siderite, along with minor titanite, chalcopyrite ± rare quartz (Fig. 4b). These segregations are irregularly shaped areas within the groundmass that have relatively sharp boundaries, grading into the surrounding groundmass over tens of millimetres. The euhedral to subhedral amphibole crystals vary in composition from potassic richterite to magnesio–arfvedsonite or actinolite, and often intrude from the margins of segregations into their interiors. The volume of amphibole in different segregations varies from <10 to 90%.
Macrocrystic serpentine–diopside ultramafic dykes (MSDU)
Several narrow macrocrystic serpentine–diopside ultramafic dykes (MSDU), each <1 m wide, intrude the MPK in the north extension lobe (DDH AN4). These ultramafic dykes contain 40% rounded olivine macrocrysts, up to 10 mm in diameter, separated by interstitial crystals of diopside (<2 mm wide), some of which are poikilitic and enclose olivine macrocrysts, along with serpentine, and <2 mm wide spinel grains (chromite to magnetite), minor fluorapatite, phlogopite, calcite, allanite–(Ce), monazite–(Ce) and aeschynite–(Ce). These dykes are only observed in DDH AN4 and thus appear very limited in extent. They have sharp intrusive contacts with the surrounding macrocrystic phlogopite kimberlite.
Phlogopite–biotite-bearing serpentinized ultramafic xenoliths
Three serpentinized ultramafic xenoliths (AN29-234·15m, AR51, AR61) that contain grains of high-Ti phlogopite–biotite were found in DDH AN29 in the north lobe (Fig. 2). Red-brown to dark-brown high-Ti phlogopite–biotite occurs either as minor (0·8 mm long) euhedral–subhedral laths, or as 1 mm long anhedral grains, both occurring at former grain boundaries between silicate minerals. The other silicate minerals in these xenoliths have been completely replaced by serpentine–talc, but the original grain boundaries are still evident and each xenolith has a coarse-grained, equigranular texture. The complete replacement of all silicate minerals, apart from high-Ti phlogopite–biotite, by serpentine–talc suggests that the original rocks were ultramafic in composition. Minor anhedral grains of rutile that are 0·7 mm long, also occur at former grain boundaries, and are associated with the high-Ti phlogopite–biotite grains in all three xenoliths. One xenolith (AR61) contains minor apatite (0·4 mm) and very rare grains of zircon (0·2 mm wide).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHYRESULTSWHOLE-ROCK GEOCHEMISTRY40AR/39AR GEOCHRONOLOGYDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Mineral chemistry
The composition of phlogopite–biotite is focused on in this study because of the insights into magma evolution provided by the chemical zonation of the phlogopite–biotite phenocrysts and the composition of high-Ti phlogopite–biotite macrocrysts, as described above. In an effort to define more precisely the provenance of the various mica macrocrysts and xenocrysts that occur in the MPK and the OPRK, LA–ICP-MS trace-element data were obtained for phlogopite–biotite macrocrysts and phenocrysts (zones 1 and 2 composition) from the MPK and the OPRK lithofacies, and phlogopite–biotite grains from a granite (AN69-59·65m) and a serpentinized ultramafic xenolith (AR51). Amphiboles in the OPRK also show significant chemical variation, and this is important in rock classification and in understanding the differentiation of the kimberlite magmas. The compositions of minor silicate phases, including diopside, sanidine and andradite, are summarized in Tables 1 and 2. The oxide mineralogy of the groundmass of the kimberlite was described in detail by Edwards et al. (1992)
; new data for spinels occurring in the groundmass and in mantle xenoliths are the subject of a separate paper (Downes et al., in preparation).
|
Texture and composition of phlogopite–biotite
High-Mg phlogopite xenocrysts
The MPK contains rare
1 mm long euhedral–subhedral xenocrysts of F-rich (1·31–3 wt %), highly magnesian phlogopite (Mg-number 95·3–98·5, Al2O3 15·3–17·13 wt %; Table 3; Fig. 5a, b) that is colourless and non-pleochroic in thin section. In samples from the central and north lobes, these phlogopite xenocrysts are overgrown by thin, continuous, orange-brown rims of phlogopite (Fig. 3i).
|
|
Trace-element data for one of these xenocrysts in a MPK sample from the central lobe (AN55-131·57m) indicate that it contains low levels of most minor elements (Nb, Zr, Rb, Sr, Co, Ni, Zn, Li, Ba, Y) compared with the other analysed phlogopite–biotites from the Aries kimberlite (Table 4; Fig 6a–f).
|
|
High-Ti phlogopite–biotite macrocrysts
Both the MPK and OPRK contain minor
1·1 mm long macrocrysts of red-brown (bt 1) and dark-brown (bt 2) high-Ti biotite, that vary in morphology from subhedral to strongly resorbed, and generally are overgrown by later generations of phlogopite (zones 2, 3, 4; Fig. 3c, e–h). Rare high-Ti biotite xenocrysts (bt 2), some of which are deformed, are more Fe-rich than the red-brown high-Ti biotite (bt 1) macrocrysts and are not overgrown by later generations of phlogopite (Figs 3h, 5a).
Rare dark-brown high-Ti phlogopite macrocrysts, 0·9 mm long, have similar subhedral to strongly resorbed habits to the high-Ti biotite macrocrysts (bt 1). These grains are similarly overgrown by later generations of phlogopite (Fig. 3b, d). There is some overlap in composition with phlogopite from two of the serpentinized ultramafic xenoliths recovered from Aries, in terms of Al–Fe–Ti content (Fig. 5a, b; Table 3). One high-Ti phlogopite macrocryst was analysed to determine its trace-element content and it is distinguishable by its relatively high Nb (84–109 ppm; Fig. 6a, c), Sr (76–99 ppm; Fig. 6d), Co (94–104 ppm), Li (33·7–45·2 ppm) and Ba (6157–7494 ppm) contents (Table 4). Very rare high-Ti phlogopite macrocrysts have sieve textures (Fig. 3d) and contain numerous small inclusions such as high-Al micas. A dissolution texture seen in high-Ti phlogopite macrocrysts in the OPRK is characterized by the crystallization along cleavage planes, of groundmass serpentine, chalcopyrite and sanidine grains, enclosing <100 µm long rectangular fragments of the larger phlogopite macrocryst.
Dark-brown phlogopite phenocryst cores (zone 1)
The cores of some phlogopite phenocrysts contain zones and patches of dark-brown phlogopite, up to 1 mm wide, with slightly higher Fe and Ti contents than the surrounding zone 2 composition phlogopite (Table 3; Figs 3j, 5a, b). The boundaries between these zones and the subsequent overgrowths of phlogopite can be highly irregular and may not be sharply defined by resorption surfaces. Phlogopite phenocrysts with zone 1 cores are rare, and have only been identified in samples of the MPK from the central and north lobes. The trace-element contents of zone 1 phlogopites generally plot quite close to those of zone 2 phlogopite phenocrysts (Fig. 6a–f). Zone 1 phlogopite from the central lobe (AN55-131·57m) is high in Sr (91–121 ppm), whereas zone 1 phlogopite from the north lobe (AN15-195·27m) is higher in Nb, but lower in Sr (35·5–37·3 ppm; Table 4; Fig. 6a, d). Both Co (71–82 ppm) and V (153–242 ppm) are slightly elevated in zone 1 phlogopite cores in comparison with the zone 2 phlogopite phenocrysts (Table 4; Fig. 6e).
High-Cr phlogopite phenocrysts (zone 2)
A widespread population of phlogopite phenocrysts (10% in volume; 1·5 mm long) is common to both the MPK and the OPRK phases. These phenocrysts are commonly euhedral, but some show evidence of resorption and rounding at the ends of laths and generally lack inclusions of other minerals (Fig. 3a–i). In regions of hypabyssal kimberlite with globular segregationary textures, these phlogopite phenocrysts lack phlogopite overgrowth rims. In terms of Al, Ti, Fe content, these phenocrysts are quite tightly constrained (Fig. 5a, b; 1·2–1·9 wt % TiO2), and partially overlap with the compositions of phlogopite phenocrysts in both lamproites and Group 2 kimberlites (orangeites) as defined by Mitchell (1995). However, overall, the Aries phenocrysts are generally more aluminous. As a population, the phenocrysts exhibit a distinctive trend of increasing Cr2O3 content, up to a maximum of 1·7 wt %, with increasing Mg-number from 83 to 90 (Fig. 5c).
Zone 2 phlogopite phenocrysts are tightly constrained in trace-element contents (Fig. 6a–f), including 50–70 ppm Co, and low V (<150 ppm). Their content of Ni varies over quite a large range (477–1061 ppm; Fig. 6f) and shows an apparent correlation with Cr2O3 (Table 4).
High-Ti phlogopite–biotite in rims and groundmass (zone 3)
One of the generations of phlogopite overgrowth rims on earlier phenocrysts and macrocrysts is a dark-brown, high-Fe–Ti phlogopite–biotite. These rims are very thin (5 µm) (Fig. 3b, d, f, g) and phlogopite–biotite of this composition also occurs in the groundmass of OPRK as a very fine-grained phase (Fig. 3d, g). This generation of phlogopite–biotite in rims and groundmass shows a limited compositional trend to lower Al2O3 with an increase in FeO in comparison with phlogopite phenocryst cores (Fig. 5a). This trend in groundmass and rim compositions correlates with the biotite trend observed by Mitchell (1995) in Group 2 kimberlites (orangeites), although the Aries phlogopites are more aluminous.
Groundmass phlogopite (zone 4)
A later stage of phlogopite growth occurs as overgrowths on phlogopite phenocrysts and macrocrysts, and as fine-grained masses to coarse-grained interlocking laths, <3 mm long, in the groundmass of MPK. These phlogopite rims commonly contain inclusions of groundmass chromite and magnetite. This stage of phlogopite growth also shows a limited trend to depletion in Al with increasing Fe content with respect to the composition of zone 2 phlogopite phenocrysts (Fig. 5a). However, none of the groundmass micas in the hypabyssal phases at Aries shows the strong Al depletion that is seen in groundmass micas from lamproites and Group 2 kimberlites (orangeites), which have a large tetra–ferriphlogopite component (Mitchell, 1995).
Barian aluminous phlogopite
Rare barian phlogopite occurs in an unusual groundmass segregation in one sample from the central lobe (AN55-131·57m; Fig. 4c). These grains contain 7·52–12·68 wt % BaO and 18·22–18·98 wt % Al2O3 (Table 3). The subhedral phlogopite grains are 0·5 mm long, and colourless in thin section. They occur in a concentrically zoned 9 mm wide groundmass segregation that contains abundant calcite, with fewer numbers of disseminated pyrite grains, apatite (forming sprays of acicular crystals in calcite), magnetite and serpentine.
Phlogopite–biotite in ultramafic and granite xenoliths
The compositional ranges of micas in the three serpentinized ultramafic xenoliths are presented in Table 3. The phlogopite–biotites in the xenoliths are TiO2-rich and show variations in Mg-number, but are relatively constant in Al2O3 content (Fig. 5a–c). The composition of high-Ti biotite from a granite xenolith (AN69-59·65m) from the north extension lobe is distinct from that of the high-Ti phlogopite–biotite contained in the other xenoliths, and any of the high-Ti phlogopite–biotite macrocrysts in the Aries kimberlite, being higher in Fe (Table 3; Fig. 5a–c).
High-Ti phlogopite grains from two of the serpentinized ultramafic xenoliths (AR51 and AN29-234·15m) have low Nb, Zr, Rb, and very low Sr contents, and high Co (86–121 ppm) and V (642–1289 ppm) concentrations (Table 4; Fig. 6a–e). Ni contents vary from 100–154 ppm in AN29-234·15m, to 509–574 ppm in AR51, and correlate with the Cr2O3 contents (Fig. 6e, f; Table 4). The high-Ti biotite from the granite xenolith also contains relatively low concentrations of Zr (0·251–0·608 ppm) (Fig. 6a) and very low Sr (1·40–2·04 ppm), but is high in Rb (621–786 ppm) (Fig. 6d). The very low Ni contents (37·2–46·3 ppm) also appear to correlate with the Cr2O3 contents (52–261 ppm Cr; Fig. 6e, f; Table 4). The most distinctive feature of the trace-element chemistry of the phlogopite–biotite from the xenoliths is the very low Sr contents that result in high Rb/Sr ratios of 308–496 for the granite xenolith, and 67–211 for the ultramafic xenoliths (Fig. 6b).
Amphiboles
The OPRK contains laths and fine-grained granular groundmass grains of potassian richterite (Na/K 4·2–5·7) containing 0·73–1·49 wt % TiO2 and 1·32–2·32 wt % Al2O3 (including one grain of magnesiokatophorite with a Na/K ratio of 3·48, 1·35 wt % TiO2, and 3·2 wt % Al2O3; Table 2; Fig. 7). The groundmass segregation amphiboles within the OPRK vary in composition from potassic richterite (Na/K ratio of 1·93–3·05, 0·33·–0.86 wt % TiO2, and 0·11–1·1 wt % Al2O3), to actinolite (at the rims of some crystals), and magnesio-arfvedsonite (Na/K ratio of 2·96–3·35, and 0·57–0·66 wt % TiO2) (Table 2). The trends in amphibole composition follow a decrease in Al, Ti, Mg and Ca and an increase in K and Fe from groundmass potassian richterite to potassic richterite in the segregations. Within the segregations, the magnesio–arfvedsonite has higher contents of Na than the potassic richterite (Table 2), and some grains are zoned from potassic richterite cores to actinolite rims showing a decrease in Na and K, and a large increase in Ca from the core to rim, respectively.
|
|
WHOLE-ROCK GEOCHEMISTRY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHYRESULTSWHOLE-ROCK GEOCHEMISTRY40AR/39AR GEOCHRONOLOGYDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
The geochemistry of the main hypabyssal phase of the Aries kimberlite has been described in detail by Edwards et al. (1992)
and Taylor et al. (1994) based on earlier drill-core material. In our study, we analysed eight new drill-core samples to determine the extent of geochemical variations that occur with the observed petrographic variations from MPK to the minor, differentiated OPRK and MSDU phases. Three of these analyses of MPK were rejected on the basis of (1) the contamination index (CI) of Clement (1982), and (2) examination of the CaO/Al2O3 ratios as an indicator of the degree of talc alteration. The compositions of two unaltered samples of MPK from the north lobe are similar to analyses reported by Edwards et al. (1992) and Taylor et al. (1994) (Table 5; Fig. 8a, b), the only differences being lower concentrations of Na2O, Cr and Ni in the samples analysed in our study.
|
|
The MSDU dykes show a depletion in SiO2, Al2O3, TiO2 and K2O, and an enrichment in Fe2O3 and MgO in comparison with MPK (Table 5). The depletion in K2O is pronounced and reflects the very low phlogopite content of the MSDU dykes. Sr contents are relatively similar between MSDU and MPK (Fig. 8a), and Ni (1200 ppm) contents are higher in MSDU than in the MPK samples reported in this study (Table 5). Notably, the MSDU is higher in Pb, U and Th than MPK or OPRK (Fig. 8a; Table 5). The incompatible elements, Y, Zr, Nb and Ta, are relatively constant in the MPK and MSDU (Fig. 8a; Table 5). The contents of rare earth elements (REE) and the chondrite-normalized REE patterns of MPK and MSDU are also similar (Fig. 8a, b).
Two samples of OPRK from the north extension lobe show marked increases in SiO2, Na2O and, to a smaller extent, K2O in comparison with the MPK. This is accompanied by a substantial decrease in MgO and Ni contents (Table 5). The OPRK is substantially enriched in Sr (1020–1180 ppm) and depleted in Ta and Th compared with the MPK and MSDU (Fig. 8a). These samples of OPRK are also depleted in light REE (LREE) relative to MPK and MSDU (Fig. 8b). Zr is moderately enriched, and Nb depleted, in comparison with the MPK and MSDU (Fig. 8a). Although the contamination indices for these OPRK samples were high (2·44–2·57; Table 5), we have reported them to provide an indication of the general geochemical variation from the MPK to the differentiated OPRK phase.
|
40AR/39AR GEOCHRONOLOGY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHYRESULTSWHOLE-ROCK GEOCHEMISTRY40AR/39AR GEOCHRONOLOGYDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Thirteen grains of mica, ranging in size from 188 to 1185 µm long, were selected for UV laser 40Ar/39Ar analysis from two kimberlite samples (AN65-36·82m OPRK; AN15-99·55m MPK), one granite xenolith (AN69-59·65m) and two serpentinized ultramafic xenoliths (AR51, AR61 from DDH AN29) from the north and north extension lobes. Cores and rims of grains were analysed by UV laser spot analysis and 15 mm wide line traverses were undertaken on the larger 476–974 mm diameter grains. One high-Ti phlogopite–biotite macrocryst from an OPRK sample (AN65-36·82m) retained an apparent 40Ar* diffusive loss profile with older
951–952 Ma core ages and younger rim ages (Fig. 9; Electronic Appendix 1, available at http://www.petrology.oxfordjournals.org/). Excluding four young 745–793 Ma 40Ar/39Ar ages from a total of 19 ages obtained from small mica grains or the rims of larger grains yielded a weighted mean 40Ar/39Ar age of 815·4 ± 4·3 Ma (95% confidence, n = 15), which is interpreted to be the kimberlite eruption age (Electronic Appendix 1).
|
The duration of kimberlite ascent for the north extension lobe was estimated via modelling of the core to rim 40Ar* diffusive loss profiles obtained from one 590 µm diameter high-Ti phlogopite–biotite grain (see Analytical Methods section). This approach is identical to that of Pearson et al. (1997)
, Kelley & Wartho (2000), Kempton et al. (2001) and Wartho & Kelley (2003), although the grains used in our study were generally smaller than those used in previous studies.
The high-Ti phlogopite–biotite macrocryst from the OPRK sample AN65-36·82m exhibited a 10 µm thick overgrowth rim of phlogopite. This thin overgrowth was not thought to unduly modify the 40Ar* diffusive loss profile in this 590 µm diameter grain. Because of the composition of the high-Ti phlogopite–biotite macrocryst in the OPRK sample AN65-36·82m (17 wt % FeO (Table 3)), it was difficult to assign a definite set of Ar diffusion parameters to this sample; therefore, the data were modelled using the Ar diffusion parameters for both Fe–biotite (Grove & Harrison, 1996) and phlogopite (Giletti, 1974). The 40Ar/39Ar data and the result of the modelling are shown in Fig. 9 and Electronic Appendix 1.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL METHODSPETROGRAPHYRESULTSWHOLE-ROCK GEOCHEMISTRY40AR/39AR GEOCHRONOLOGYDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAREFERENCES |
|---|
Classification: is Aries a kimberlite or a lamproite?
The classification of alkaline rocks is a difficult and often controversial process. Edwards et al. (1992)
classified Aries as a micaceous kimberlite on the basis of its mineralogical, geochemical and isotopic characteristics, in agreement with the criteria recommended by Woolley et al. (1996). In light of the occurrence of a minor richterite- and sanidine-bearing hypabyssal phase (OPRK) in the north extension lobe of the Aries kimberlite—mineralogical characteristics that might suggest lamproitic affinities—it is worth re-examining the basis for its classification. In Table 6, the mineralogical characteristics of the Aries kimberlite are compared with those of olivine lamproites and Group 2 kimberlites (orangeites). At Aries, hypabyssal kimberlite (MPK and OPRK) contains phlogopite phenocrysts that are similar in chemistry to, but, as a group, are more aluminous than, ‘primitive’ phlogopite from lamproites and Group 2 kimberlites (orangeites; Fig. 5a, b, using fields defined by Mitchell, 1995). Phlogopite rims on macrocrysts and phenocrysts, and groundmass phlogopite show similar compositional trends to groundmass phlogopite in lamproites and Group 2 kimberlites (orangeite), but lack the strong Al depletion which is characteristic of a large tetra–ferriphlogopite component. Both MPK and OPRK contain high-Ti phlogopite–biotite macrocrysts with close textural and compositional similarities to those described from South African Group 2 kimberlites (orangeites) and other micaceous kimberlites worldwide (‘type 1’ mica of Smith et al., 1978; Mitchell, 1995; Figs 5a, b, 10). Serpentine and calcite are common in the groundmass and groundmass segregations of MPK (Table 6).
|
