Constraints on kimberlite ascent mechanisms revealed by phlogopite compositions in kimberlites and mantle xenoliths
Introduction
Kimberlites are of great economic and scientific value because they are the major hosts to diamonds, entrain abundant mantle and crustal xenoliths and represent the deepest magmas (> 150–200 km) produced in the mantle that we may observe at the Earth’s surface. At the surface, kimberlites occur as diatremes, sills and dykes. Mapping indicates that kimberlite diatremes typically comprise multiple units that originated from discrete magma batches, whereas sills and dykes represent one or more discrete magma pulses at surface (e.g., Clement, 1982, Dawson and Hawthorne, 1970, Field and Scott Smith, 1999, Mitchell, 1986, Mitchell, 1995, Nowicki et al., 2004, Sparks, 2013).
Despite the importance of kimberlites, and at least 40–50 years of dedicated studies, several unresolved issues remain in relation to different aspects of kimberlite petrology, geochemistry, mineralogy and volcanology. These include, but are not limited to:
- 1)
The composition of kimberlite melts upon formation in the deep Earth and at the time of emplacement in the upper crust (e.g., see contrasting views of Kopylova et al., 2013 and Kamenetsky et al., 2014) and the extent to which kimberlite melts are modified by interaction with, and assimilation of, wall-rock material (e.g., Donnelly et al., 2012, Kamenetsky and Yaxley, 2015, Kirkley et al., 1989, Malarkey et al., 2010, Mitchell, 2008, Patterson et al., 2009, Pilbeam et al., 2013).
- 2)
The depth of kimberlite melt formation, i.e. the asthenosphere (e.g., Griffin et al., 2014, Haggerty, 1994, Paton et al., 2009, Ringwood et al., 1992, Smith, 1983, Tappe et al., 2011, Tappe et al., 2013) vs the lithosphere (e.g., Becker and le Roex, 2006, Gaffney et al., 2007, le Roex et al., 2003, Tainton and McKenzie, 1992).
- 3)
The causes of the very fast ascent (several m/s; Canil and Fedortchouk, 1999, Kelley and Wartho, 2000, Peslier et al., 2008) of kimberlite magmas (cf. Brett et al., 2015, and Russell et al., 2012vsKamenetsky and Yaxley, 2015, Wilson and Head, 2007vsSparks et al., 2007, Gregoire et al., 2006).
- 4)
The extent to which kimberlite rocks are modified by syn- and post-emplacement processes including alteration by ground waters (e.g., see contrasting views of Mitchell, 2008, Mitchell, 2013 and Afanasyev et al., 2014, Giuliani et al., 2014b, Sparks, 2013, Stripp et al., 2006).
The purpose of this contribution is to provide new insights into two of the above controversies, namely, the ascent mechanism(s) of kimberlite magmas and the evolution of kimberlite melts during ascent through the lithospheric mantle. We leave the remaining questions to future studies.
The driving forces that promote the rapid ascent of kimberlite magmas are poorly constrained. For example, Russell et al. (2012) proposed that exsolution of a CO2-rich vapour phase at mantle depths provides the main propellant for kimberlite melt ascent. However, this hypothesis is at odds with the carbonate-rich nature of kimberlite matrices and, hence, the parental magma(s) (e.g., Brooker et al., 2011, Kamenetsky et al., 2004, Kamenetsky et al., 2014, Nielsen and Sand, 2008, Patterson et al., 2009), as well as the non-explosive emplacement of many kimberlite magmas (e.g., magmatic dykes several km in length) (e.g., Andrews and Emeleus, 1975, Mitchell, 1986, Nielsen and Sand, 2008). Conversely, if magma ascent is not driven by volatile exsolution as proposed by Gregoire et al. (2006), we might expect the highly reactive kimberlite melts to be entirely consumed through reaction with mantle wall rocks. In fact, experiments using bulk kimberlite rock (Canil and Fedortchouk, 1999, Chepurov et al., 2013) and Na-carbonate compositions (Kamenetsky and Yaxley, 2015) as proxies for kimberlite melts have shown that orthopyroxene, clinopyroxene, garnet and, to a lesser extent, olivine in mantle peridotites are unstable and dissolve in kimberlite magmas at mantle pressure and temperature conditions. Petrographic observations of resorbed orthopyroxene, clinopyroxene and garnet in kimberlite rocks (Hunter and Taylor, 1982, Kamenetsky et al., 2009, Mitchell, 2008, Shee et al., 1994) confirm the instability of these minerals in kimberlite magmas. The experimental and petrographic evidence for assimilation of entrained mantle material by kimberlite melts implies that (1) the composition of kimberlite melts must vary during their rise to surface; and (2) the ascent of kimberlite magmas to the Earth’s surface requires unique conditions, such as partial shielding from the wall rocks lining the magma conduit (Giuliani et al., 2014a).
Investigations of the evolution of kimberlite magmas require phases that preserve geochemical signatures of magma interaction during transport through both the crust and the mantle. One such phase is phlogopite. Phlogopite is a common mineral in kimberlites where it spans a range of size from macrocryst (~> 0.5 mm) to groundmass phase (~< 0.1 mm), and preserves large compositional variations, thus providing a useful repository of information on the evolution of kimberlite and related melts (Beard et al., 2000, Downes et al., 2006, Fritschle et al., 2013, Hunter et al., 1984, Mitchell, 1986, Mitchell, 1995, O'Brien et al., 1988, Reguir et al., 2009). Phlogopite is also a common mineral in metasomatised mantle rocks (Boettcher and O'Neil, 1980, Carswell, 1975, Dawson and Smith, 1977, Erlank et al., 1987, Grégoire et al., 2002, Jacob et al., 2009, Jones et al., 1982, Konzett et al., 2000, Rehfeldt et al., 2008, Waters, 1987), including those preserving evidence of interaction with kimberlite-related melts immediately prior to kimberlite eruption (e.g., Boyd et al., 1983, Dawson et al., 2001, Farmer and Boettcher, 1981, Giuliani et al., 2012, Giuliani et al., 2013a, Giuliani et al., 2014b).
In this contribution, we compare the major and trace element compositions of phlogopite macrocrysts, microcrysts and groundmass grains in the Bultfontein kimberlite (Kimberley, South Africa (Fig. 1); i.e. the kimberlite type-locality) with phlogopite crystals in mantle xenoliths entrained by the Kimberley kimberlites. Phlogopite grains in the Bultfontein kimberlite exhibit multiple concentric zoning patterns that record a complex history of crystal growth episodes and interaction with mantle wall rocks. These zoning profiles reveal that individual kimberlite units in diatremes, sills and dykes are not necessarily the product of a single pulse of kimberlite melt, but can result from complex, multi-stage emplacement and interaction mechanisms that are often obscured by the very complexity of these rocks. Outcomes of this work also support our previous inference that kimberlite magmas probably only reach the surface if they ascend along conduits that have already been metasomatised by previous pulses of kimberlite melt (Giuliani et al., 2014a).
Limited petrographic and compositional data exists for phlogopite hosted in the Kimberley kimberlites. Clement (1982) reported that phlogopite occurs as megacrysts (> 10 mm), macrocrysts (0.3–10 mm), phenocrysts and groundmass grains in these kimberlites. Macrocrysts, phenocrysts and groundmass grains display complex, multiple zoning; in particular dark-coloured overgrowths often rim partially resorbed cores of macrocrysts. Shee (1985) provided additional details on the compositions of phlogopite in the Wesselton kimberlite, which is part of the Kimberley cluster (see Section 2). He reported that macrocryst cores have compositions analogous to phlogopite in metasomatised mantle rocks from the Kimberley area. In addition, phlogopite macrocrysts display deformation features (e.g., kink-banding, undulose extinction) attributed to transport from mantle depths. Shee (1985) and Clement (1982) suggested that most phlogopite macrocrysts probably originated from disaggregation of metasomatised mantle rocks as previously proposed for some phlogopite megacrysts from southern African kimberlites (Dawson and Smith, 1975). Interestingly, Shee (1985) reported high Ti-Cr compositions for some macrocryst rims, which are distinct from the compositions of groundmass mica in the Wesselton kimberlite; these rims host inclusions of spinel of undefined composition.
Apart from the detailed work of Clement (1982) and Shee (1985), Dawson and Smith (1975) and Boettcher and O'Neil (1980) reported the major element compositions of megacrysts from the Kimberley kimberlites. Allsopp and Barrett (1975) reported the Rb-Sr isotope systematics of phlogopite from some of the Kimberley kimberlites and Fitch and Miller (1983) analysed the K-Ar isotope composition of mica from the De Beers pipe, also part of the Kimberley cluster.
Several studies have documented the petrography and geochemistry of mica in mantle xenoliths from the Kimberley kimberlites (e.g., Boettcher and O'Neil, 1980, Boyd et al., 1983, Carswell, 1975, Dawson and Smith, 1977, Delaney et al., 1980, Erlank et al., 1987, Farmer and Boettcher, 1981, Giuliani et al., 2013a, Giuliani et al., 2013c, Giuliani et al., 2014a, Giuliani et al., 2014b, Giuliani et al., 2014c, Grégoire et al., 2002, Jacob et al., 2009, Jones et al., 1982, Konzett et al., 2000, Matson et al., 1986, Rehfeldt et al., 2008, Rosenbaum, 1993, Shee, 1985, Simon et al., 2007, Waters, 1987). Phlogopite in mantle xenoliths commonly occurs as laths of variable size (rarely exceeding 1 cm in size) associated with other metasomatic minerals such as clinopyroxene, K-richterite, ilmenite, rutile, titanates, sulphides, etc. The composition of these phlogopite laths is usually homogeneous, but, in some xenoliths, darker rims enriched in Ti are observed (Delaney et al., 1980, Farmer and Boettcher, 1981, Giuliani et al., 2014a, Jones et al., 1982, Rosenbaum, 1993). Irrespective of the presence of large phlogopite laths, several mantle xenoliths host a second generation of phlogopite, which occurs as interstitial, fine-grained, dark-coloured grains occasionally associated with pyroxene and/or Fe-Ti oxide minerals (e.g., Boyd et al., 1983, Dawson and Smith, 1977, Delaney et al., 1980, Erlank et al., 1987, Rehfeldt et al., 2008). The formation of large phlogopite laths is probably linked to mantle metasomatism prior to xenolith entrainment in kimberlite magmas (e.g., Giuliani et al., 2014c, Konzett et al., 1998). In contrast, the fine-grained generation is attributed to metasomatism associated with the kimberlite magmatic event (e.g., Boyd et al., 1983, Dawson and Smith, 1977).
Section snippets
Geological setting
The kimberlite and mantle xenolith samples examined in this study derive from the Bultfontein Dumps, which hosts waste material from mining of the Bultfontein kimberlite. The Bultfontein kimberlite is part of the Kimberley cluster of kimberlites, which also includes the De Beers, Dutoitspan, Wesselton, Wesselton Floors and Kimberley kimberlites, along with a number of smaller pipes (Field et al., 2008), and is located in the SW part of the Kaapvaal craton (Fig. 1). On the basis of their
Sample descriptions
This study examines phlogopite grains from a sample of hypabyssal kimberlite and 9 variably metasomatised mantle xenolith samples from the Bultfontein kimberlite (Kimberley; Fig. 1). Sample BK is an off-cut (8 × 6 × 1.5 cm) of a larger sample of massive macrocrystic kimberlite. In this work, macrocrysts are distinguished from microcrysts and groundmass grains based on the exposed surface area in thin section: macrocrysts have a surface area > 50 × 103 μm2 (i.e. ~> 0.3–0.5 mm for the longest axis of
Analytical methods
Initial studies of the texture and chemical zoning of phlogopite grains in the Bultfontein kimberlite were performed on carbon-coated thin sections using a Philips (FEI) XL30 environmental scanning electron microscope (ESEM) TMP, equipped with an OXFORD INCA energy-dispersive X-ray spectrometer (EDS) at the University of Melbourne. X-ray mapping of phlogopite macrocrysts was undertaken using a Hitachi SU-70 field-emission scanning electron microscope (FE-SEM) equipped with an OXFORD
Results
Phlogopite macrocrysts and microcrysts commonly show zoning patterns that are evident from changes in pleochroic colours (Fig. 2b and c and Table 2) and back-scattered electron response (Fig. 3). Phlogopite cores are often resorbed with rounded or embayed outlines and are overgrown by one or more (up to four) zones with different compositions (Figs. 2b, c, 3). The partially resorbed cores host inclusions of clinopyroxene and chromite, whereas carbonate (dolomite, calcite, strontianite) and
Origin of low Ti-Cr macrocryst cores
Zoning is a typical feature of phlogopite macrocrysts in kimberlites (e.g., Beard et al., 2000, Clement, 1982, Dawson, 1980, Mitchell, 1986, Mitchell, 1995, Reguir et al., 2009, Shee, 1985). The resorption (i.e. embayments, rounded shape) and deformation textures (Clement, 1982, Shee, 1985) of the macrocryst cores (Figs. 2b, c and 3) and the occurrence of mineral inclusions (e.g., clinopyroxene; Fig. 3e) that are absent from magmatic assemblages in kimberlites (e.g., Kamenetsky et al., 2009,
Conclusions
Despite decades of dedicated study, several aspects of kimberlite magma formation and evolution remain intensely debated and often enigmatic. This detailed study of the composition of phlogopite grains in the Bultfontein kimberlite has provided additional new constraints on the ascent mechanisms of kimberlite magmas and their geochemical evolution.
In sample BK, phlogopite macrocrysts (and some microcrysts) exhibit zones with similar Ti-Cr-rich compositions to rims of macrocrysts in the
Acknowledgments
AG acknowledges funding from the Australian Research Council through a Discovery Early Career Research Award (DECRA, grant n. DE150100009) and from the Australian Academy of Sciences through a J.G. Russell Award. We are grateful to Graham Hutchinson and Alan Greig for support with the electron microprobe and LA-ICPMS analyses. De Beers Consolidated Mines and Bruce Wyatt are thanked for providing access to the studied samples. This and previous versions of the manuscript were improved by
References (111)
- et al.
Rb-Sr age determinations on South African kimberlite pipes
Physics and Chemistry of the Earth
(1975) - et al.
Structural aspects of kimberlite dyke and sheet intrusion in south-west Greenland
Physics and Chemistry of the Earth
(1975) - et al.
LAM-ICPMS U-Pb dating of kimberlitic perovskite: eocene-oligocene kimberlites from the Kundelungu Plateau, D.R. Congo
Earth and Planetary Science Letters
(2008) - et al.
Geochemistry and mineralogy of kimberlites from the Arkhangelsk Region, NW Russia: evidence for transitional kimberlite magma types
Lithos
(2000) - et al.
Origin of olivine in kimberlite: phenocryst or impostor?
Lithos
(2009) - et al.
The ascent of kimberlite: insights from olivine
Earth and Planetary Science Letters
(2015) - et al.
Garnet dissolution and the emplacement of kimberlites
Earth and Planetary Science Letters
(1999) Primary and secondary phlogopites and clinopyroxenes in garnet lherzolite xenoliths
Physics and Chemistry of the Earth
(1975)- et al.
The stability of ortho- and clinopyroxenes, olivine, and garnet in kimberlitic magma
Russian Geology and Geophysics
(2013) - et al.
Chemistry of micas from kimberlites and xenoliths-II. Primary- and secondary-textured micas from peridotite xenoliths
Geochimica et Cosmochimica Acta
(1980)