Kimberlites

The first Kimberlites were described by Vanuxen in 1837 from Ludlowiville near Ithaca, New York state; however, the term Kimberlite was introduced by Lewis (1887) to describe the diamond-bearing, porphyritic mica peridotites of the Kimberley area of South Africa. Kimberlites are highly magnesian (MgO > 25% weight) magmatic rocks which are enriched in volatiles (water, carbon dioxide, and fluorine) and carry anomalously high contents of elements such as K, Na, Ba, Sr, rare earth elements, Ti, Zr, Nb, and P. In simple terms, Kimberlites, constitute a hybrid group of rocks that encompass a group of volatile rich (dominantly CO2) potassic, ultrabasic rocks and that displays a pronounced inequigranular texture, resulting from the presence of macro-crysts (and/or mega-crysts) that are set in a fine grained matrix.

Owing to the great diversity in terms of their textural, mineralogical, petrographic, and geochemical characteristics, diverse definitions and classifications for Kimberlites were proposed.

Classification based on the textural and genetic variations:

This model proposed by Clement and Skinner, (1979) relying on textural features identifies three genetic facies of kimberlite rocks.

1) Crater Facies Kimberlite
2) Diatreme Facies Kimberlite
3) Hypabyssal Facies Kimberlite

1) Crater Facies: The surface morphology of un-weathered kimberlite (Fig.1) is characterized by a crater, up to 2 km in diameter, whose floor may be 150 to 300 m below the surface. The crater is generally deepest in the middle and round the crater is a tuff ring which is relatively small, generally less than 30 meters, when compared to the diameter of the crater. Crater facies are represented by pyroclastic (formed as a result of eruptive forces) and epiclastic rocks (fluvial alteration of pyroclastic material) and are distinguished by sedimentary (layer) deposition.

kimberlites2020(4).png

Fig.1: Crater Facies Kimberlite. Modified after Mitchell 1986.



Two main categories of rocks are found in crater facies kimberlites; pyroclastic, those deposited by eruptive forces; and epiclastic, which are rocks reworked by water.

Pyroclastic Rocks: These rocks are found preserved in tuff rings around the crater and within the crater. Tuff rings have small height. Igwissi Hills in Tanzania and Kasami in Mali are the pipes with well-preserved tuff rings (Fig.2). Deposits are commonly bedded, vesicular and carbonatised. Tuff deposits preserved within the crater are also rare; however, the Igwissi Hill pipes in Tanzania have been examined and revealed three distinct units. From top to bottom, they are:

1. Well-stratified tuffs layers defined by lapilli and ash size particles.
2. Poorly stratified coarse pyroclastics.
3. Basal breccias.

Epiclastic Rocks: These sediments represent fluvial reworking of pyroclastic material from the tuff ring in the Crater Lake formed on top of the diatreme. They are complex and resemble a series of overlapping alluvial fans mixed in with lacustrine deposits.

kimberlites2020(10).jpg

Fig.2: Igwisi Hills kimberlite crater. From The earth story.



2) Diatreme Facies: The diatreme facies in kimberlite is characterized by a carrot shaped body with near circular or elliptical outline on the surface and steeply dipping (80°-85°) walls. These facies sometimes may exceed 2 km in depth. The diatreme facies are characterized by fragmental nature and the presence of angular to rounded country rock fragments (ranging from a few centimeters to sub-microscopic size) imparts a distinct identity. This facies is constituted by autoliths (rounded fragments of earlier generations of kimberlite), pelletal lapilli, (large rounded to elliptical lapilli sized clasts represented by a large anhedral olivine or phlogopite in the form of a nucleus, that is enclosed in a optically unresolvable micro-phenocrystal matrix), fragmented mantle xenoliths that are represented by discrete and fractued grains of olivine garnet, clinopyroxene and ilmenite set in a product of magnetic crystallization consisting of micro-phenocrysts and groundmass.

3) Hypabyssal Facies: The hypabyssal facies kimberlites are rocks formed by the crystallization of volatile rich kimberlite magma. Macroscopically they are massive rocks in which the macro-crystal olivine and other macro-crysts (ilmenite, phlogopite, garnet) are commonly visible. They show the igneous textures and effects of magmatic differentiation. Some of the characteristic textural features of this facies include: 1. Absence of pyroclastic fragments and textures, 2. Presence of late stage poikilitic growth of phlogopite, 3. Segregation textures involving segregation of calcite and serpentine. 4. Flow banding marked by the preferred orientation of micro-phenocrysts.

kimberlites2020(3).png

Fig.3: Model of an idealized kimberlite system, illustrating the hypabyssal, the diatreme and the crater facies. From Mitchell (1986).



Based on difference in their isotopic composition, Smith (1983) classified the kimberlites in to two groups: Group-I and Group-II kimberlites.

Group I kimberlites: Group I includes the most classical kimberlites, originally termed basaltic kimberlites: that is, ultrabasic (SiO2 < 45 wt%), potassic (K/Na atomic ratio > 1), volatile-rich (dominantly CO2) rocks, characterized by the presence of macro- and mega-crysts of magnesium-rich minerals such as olivine, ilmenite, pyropic garnet, variably chromium-rich diopsidic pyroxene, phlogopite, enstatite, and Ti-poor chromite, set in a fine matrix of olivine, serpentine, carbonate, and other accessory Mg- and/or Ca-rich minerals. Both the macro- and the megacrysts are at least in part xenocrysts, or accidental crystalline components derived from disruption of country-rocks (essentially deep-seated mantle peridotites and eclogites) crosscut by the rising kimberlite magma.

Group II kimberlites (orangeites): originally termed micaceous or lamprophyric kimberlites, are ultrapotassic (K/Na > 3), peralkaline ([K + Na]/Al > 1), volatile-rich (dominantly H2O) rocks, characterized by the presence of phlogopite and olivine as macro-crysts, in a groundmass made of phlogopite, olivine, and diopside, commonly zoned to titanian aegirine, spinel ranging in composition from Mg-bearing chromite to Ti-bearing magnetite, perovskite, and other minerals. They have greater mineralogical affinity with lamproites than with group I kimberlites.

Distribution of Kimberlites in the world

Kimberlites find distributed in all the continents of the world (Fig.4). On the basis of the distribution patterns of the kimberlites across the world, Clifford (1966), observed that the economically viable kimberlites occur primarily on Pre-Cambrian Cratons, particularly those of Archaean age (older than ca. 2.5 Ga). This observation later on came to be known as Cliffords Rule. No primary diamond deposit is known in crustal terranes younger than 1.6 Ga. This peculiar association suggests a link between the presence of diamonds and the age of the subcontinental lithosphere, and Clifford’s rule has long been considered as a valuable selection criterion in diamond exploration programs. It is worth noting here that diamondiferous kimberlites are usually young compared to the age of the lithosphere in which they have intruded. Many (including most South African examples) are Cretaceous, many others are Paleozoic (as in the Sakha Republic, Siberia), but the whole array extends from the Proterozoic to the Neogene (such as some 22 Ma examples in Western Australia).

kimberlites2020(2).png

Fig.4: Worldwide distribution of kimberlites.



Kimberlite Emplacement Models

Various models of kimberlite pipe emplacement have been proposed over the years. These include: 1) The explosive-boring theory, 2) The Fluidization Theory, 3) the hydrovolcanic theory and 4) Embryonic Pipe Theory.

Explosive Volcanism Theory

The volcanic nature of kimberlite was soon recognized (Lewis 1887, Bonney 1899) and under the influence of ideas advanced by Geikie (1902) to explain the origins of similar diatremes in Scotland, it was proposed that kimberlites were emplaced by explosive-boring (Wagner 1914). Kimberlitic diatremes were thus regarded as volcanic vents erupting explosively from depths of up to 2 km. The eruption was considered to have originated from the violent explosive liberation of highly compressed vapors and gases of magmatic origin. The level at which this occurred is now marked by the transition from feeder dike to diatreme.

Kimberlite magma is considered to rise from the deep mantle along cracks and fissures. The magma is believed to contain insufficient volatiles to allow direct explosive eruption, and its ascent is therefore halted when some impermeable level is reached. Pooling produces magma chambers, at relatively shallow depths, termed intermediate chambers. Crystallization in these chambers results in volatile build up. Sufficient pressures are eventually generated to cause up-warping and fracturing of the roof. Explosive eruption of kimberlite with concomitant brecciation of the conduit then follows until the excess pressure is reduced. Repetition of the process can account for multiple intrusion at a single vent, or the occurrence of closely spaced diatremes if the roof fracture occurs at slightly different points above the magma chamber.

Through extensive mining it is clear, that this theory, is untenable. The principal arguments against the hypothesis, either in its original form (Wagner 1914) or in modified are the following:

1) No evidence for forceful intrusion exists, there is an absence of up-doming concentric fracturing; 2) there are no explosion centers at depth, either at the base of the diatremes or the root zones; 3) deep mining has not revealed intermediate chambers either; 4) explosive-boring is not consistent with the restriction of breccias to the interior of conduits, some of which have never reached the surface. 5) The zonal arrangement of xenoliths; the sinking of xenoliths; and the preservation of the pre-existing country rock stratigraphy in the mega-xenolith assemblage, are not consistent with explosive vent clearing; 6) Projections of country rock into the diatreme are found which could not survive explosive activity.

Fluidization Theory
Dawson (1962, 1967a, 1971, 1980) has been the principal advocate of the fluidized emplacement of kimberlite diatremes. He believes that the distribution, rounding, and striation of inclusions, the juxtaposition of xenoliths derived from various depths, the surrounding and partial detachment of blocks of country rock, the absence of up-doming, and lack of thermal metamorphic effects can only be explained by this process. Dawson thus envisions a gas-charged kimberlite magma rising from the upper mantle through a fracture system. At suitable points of crustal weakness, breakthrough to the surface occurs from depths of 2-3 km. Adiabatic expansion of magmatic gases (dominantly CO2) occurs and the explosion vent is enlarged and infilled by fluidized fragmental kimberlite, drilling upward with a sandblasting effect and following major joint systems. In some diatremes later gas surges emplace distinctive tuff columns, while cavities in the vent may be infilled with magma which consolidates as massive kimberlite or incorporates clasts to form kimberlite breccia.

Diatreme formation by fluidization has not been accepted by all volcanologists and has been rejected in particular by those who believe that diatremes form by hydrovolcanic processes. The principal arguments advanced against fluidization are the following:

It is very unlikely that high vapor pressures and large volumes of gases will be exsolved from slowly cooling magmas deep in the crust. Rapid vesiculation can occur only at shallow depths, moreover these intrusions are of such small volume that it is doubtful whether they could produce sufficient quantities of volatiles to support a 2-km-long fluidized bed. It is not specified why the gas phase should exsolve all at once and disrupt the magma into pyroclasts, or why other batches of kimberlite do not exsolve gases in the same manner. Autholithic clasts in diatreme facies kimberlites are not vesicular or shard-like; commonly they are angular and fractured and show no signs of the abrasion features expected as a consequence of involvement in gas-tuff streaming. Most of the xenolith clasts are angular and have not therefore been subjected to extended periods of aggregative bubbling fluidization. The concentration of xenoliths at specific horizons and the preservation of a crude stratigraphy in the mega-xenolith suite is not consistent with long periods of bubbling fluidization. The presence of the xenoliths is not compatible with the required earlier period of erosional widening of the pipe by high velocity gas-tuff streaming.

In summary, while the fluidization hypothesis has been widely accepted as a mechanism of diatreme emplacement, it does not play a significant role in the formation of kimberlitic diatremes.

Hydrovolcanic theory
Hydrovolcanism refers to volcanic phenomena produced by the interaction of magma or magmatic heat with an external source of water, such as a surface body or aquifer. The main proponent of this theory is Lorenz (1999). Lorenz proposes that diatremes and maars form at hydraulically active zones of structural weakness such as faults or lineaments. Magma rising as a dike enters the fracture and contacts circulating groundwater; the resulting hydrovolcanic explosion fragments and chills the magma and brecciates the country rock. Hydroclastic debris may be ejected as a tuff ring surrounding a maar. Continued activity results in the enlargement of the fissure by further wall rock brecciation and spalling of the rock into the fracture as a consequence of pressure differences between the wall rocks and the explosion chamber formed where the water and magma interact.

Lorenz's hypothesis of diatreme formation is attractive in that the following features of kimberlitic diatremes may be explained: Diatremes (and maars), in general, are clearly related to linear features. Many kimberlite diatremes at their lower levels are seen to be located at the intersection of dikes and fractures. Feeder dikes appear to have risen into pre-existing fracture systems. All of these zones of weakness may be hydraulically active. Diatremes are commonly developed in thick sequences of sedimentary and volcanic rocks of high porosity and permeability. Diatremes are less commonly found in rocks of low permeability such as granite gneiss terrains. Kimberlite diatremes occur in groups. Modem maars and Tertiary diatremes also occur in clusters and their close geographical association is apparently related to the local hydrological regime. Mega-xenoliths (floating reefs) are interpreted as down-faulted and/or subsidence features. The occurrence of epiclastic kimberlite& indicates that the crater above the diatreme was at times filled with water. The presence of blocks of these kimberlites at depth in the diatreme indicates that the crater lake can be disturbed by later eruptions. Drainage will obviously promote hydrovolcanic eruptions in the underlying diatreme.

Embryonic Pipe Theory
Recognizing the complexity of kimberlite pipes, Clement (1979, 1982) believes that no single process can account for their diverse geological and petrographic characteristics. In his model, root zones are interpreted as embryonic pipes which are modified by post-surface breakthrough fluidization into diatremes.

According to this theory, Kimberlite magma dikes rising from depth are believed to develop a precursor volatile phase due to exsolution of CO2 liberated as a consequence of pressure decrease. This volatile phase, being under high pressure, penetrates fractures and joints in the wall rocks above and at the margins of the intrusion. The advancing front of contact brecciation is followed by magma which penetrates the breccias and any joints or fractures present. Intrusion breccias are formed and wall rocks are wedged into the conduit. The path of the advancing magma is controlled by pre-existing structures. The change from fissure filling to root zone development may be due to increasing volatile exsolution as pressure falls upon ascent, intersection of the dike with a fracture that can be exploited, or which contains groundwater.

This process is envisioned to continue until the magma reaches a level where explosive breakthrough to the surface is possible. Clement (1979, 1982) believes that this occurs at 300-400 m and may be promoted by groundwater-magma interactions. As a consequence of breakthrough and pressure release, the magma in the root zone is believed to de-gas rapidly and to form a vapor-liquid-solid fluidized system.

The vapor exsolution surface is considered to migrate rapidly downward as a consequence of expansion and further pressure release (Fig.5). During this period of fluidization, pre-existing root zone hypabyssal kimberlites, high level contact breccias, and degassing magma are thoroughly mixed. Lack of rounding of country rock clasts indicates that the fluidized system existed only briefly. Repetition of the whole process will produce diatremes containing several distinct varieties of diatreme facies kimberlites and very complex root zones.

kimberlites2020(6).png

Fig.5: Embryonic pipe development. Front of contact brecciation in red. Modified from Mitchell, R. H. (1991).



kimberlites2020(1).png

Fig.6: Stages in the development of a diatreme as envisioned by Clement (1982). The period of embryonic pipe development is followed by either fluidization (A) or hydrovolcanism (B). Modified from Mitchell, R. H. (1991).



The complex structure of kimberlite pipes indicates that no single process is responsible for their formation. Pipe development is initiated by sub-surface brecciation processes which lead to the formation of a complex root zone above a feeder dike. Surface breakthrough is not the result of explosive-boring but the gradual ascent of the root zone complex to levels where crater formation by hydrovolcanic explosion can occur. Diatremes appear to be secondary structures formed by the subsequent modification of the underlying root zone or embryonic pipe, by fluidization or downward migrating hydrovolcanism.

Petrogenesis

In spite of vast research, the origin of kimberlites remains controversial, in particular to the nature and depth of their source region. Kimberlites are characteristically associated with a suite of mafic and ultramafic xenoliths whose mineralogy indicates an origin within the upper mantle. Such xenoliths are fragments of conduit wall-rock detached by kimberlite magma during its rapid ascent through the lithosphere, and they place useful constraints on where and under what conditions the kimberlite melt formed. Kimberlitic magmas are thought to form through partial melting deep in the mantle.

Kimberlites, like carbonatites, are rare, but have been found on almost every continent, and are also the principal transporter of a variety of xenoliths from crustal and mantle depths. Importantly, these mantle xenoliths brought up by kimberlites are the primary source of information on the nature of phsico-chemical processes in the mantle, and even more so, the continental mantle (Pearson et al., 2004). Kimberlites form part of a spectrum of silica-undersaturated rocks that vary widely in composition and include such rock types as melilitites, lamprophyres, and nephelinites (Fig. 7). The petrogenesis of kimberlites is, however, controversial, with disagreements over the nature and depth of the source region, whether they are primary in origin, and the cause of melting (e.g., plume vs. volatile fluxing) (Keshavet al., 2005).

Three general types of hypotheses have long been considered for the genesis of kimberlites:

1. Kimberlites are a mechanical mixture of a H2O-rich ankeritic magma and a granitic lower crust (Dawson, 1967).
2. Kimberlites are resulted directly from the partial melting, at high pressures, of a mafic to ultramafic mantle (Wagner, 1929; Holmes, 1936).
3. Kimberlites are formed by high pressure differentiation of a mafic magma (proto-kimberlite) by a process of continued fractional crystallization (Williams, 1932; O'Hara, 1968).

The geological association of kimberlites with specific suites of xenoliths, and the comparison with experimental data, give support to the last hypothesis (n.3) previously proposed by a number of other authors (MacGregor, 1970). The initial melt or proto-kimberlite (Kamenetsky et al. 2008) is assumed to be a chloride-carbonate-rich fluid with a very low SiO2 content. During its passage towards the surface, its composition becomes more like that of kimberlitic magma as it interacts with its mantle wall-rocks: the assimilation of olivine and other mantle minerals increases the silica content of the fluid, driving it towards the low-SiO2, high-MgO composition characteristic of kimberlite. However, despite the significant advances in the petrology and geochemistry of kimberlite magmatism, determination of kimberlite melt compositions both in the hypabyssal facies and in the mantle remains aproblem under debate (Kamenetsky et al., 2009;Russell et al.2012;Sparks et al. 2009; Pesikov et al.,2015).

kimberlites2020(11).jpg

Fig.7: Schematic cross section of an Archean craton, with an extinct mobile belt (once associated with subduction) and a young rift. The low cratonal geotherm causes the graphite-diamond transition to rise in the central portion. Lithospheric diamond therefore occurs only in the peridotites and eclogites of the deep cratonal root, where they are them incorporated by rising magmas (mostly kimberlitic K). Lithospheric orangeites (O) and some lamproites (L) may also scavenge diamonds. Melilitites (M) are generated by more extensive partial melting of the asthenosphere; depending on the depth of segregation they may contain diamonds. Nephelinites (N) and associated carbonatites develop from extensive partial melting at shallow depths in rift areas, and they do not contain diamonds. From Mitchel 2005.



Diamond and kimberlites

Kimberlites are the most important source of primary diamonds. Many kimberlite pipes also produce rich alluvial or eluvial diamond placer deposits. About 6.400 kimberlite pipes have been discovered in the world, of those about 900 have been classified as diamondiferous, and of those just over 30 have been economic enough to diamond mine.

Although diamond crystals are found in Kimberlite and related rocks, the origin of diamond (Fig. 7) is more closely related to the fragments of peridotite and eclogite which are derived from the upper mantle, below cratonic (shield) areas. In order for diamonds to form, they require extremely high pressures and temperatures which are only found in these deep levels of the earth. It is here that the rock, eclogite, forms consisting of red pyrope garnet and green clinopyroxene; diamond crystals develop alongside the garnet and pyroxene crystals. Peridotite fragments (xenoliths) composed of garnet, olivine, and orthopyroxene also contain diamonds and are similarly derived from the upper mantle. However, these fragments commonly disaggregate during the emplacement process resulting in a matrix containing the disaggregated minerals of olivine, pyroxene, and diamond (xenocrysts).

Although diamond crystals form in the upper mantle below cratonic areas, they can only remain stable at these high pressures and temperatures. The mantle xenoliths and diamond crystals that are brought quickly to surface in a Kimberlite magmatic fluid are able to survive near surface in a quenched or meta-stable state. If the intrusion of kimberlite is delayed during its rise to surface or is trapped in the lower crust, diamond crystals will not be stable in the P-T environment and will revert to graphite.

It is under shield areas or cratons that the diamond crystals can remain stable at shallower depths due to the low geothermal gradient related to the sub-cratonic keel beneath continental crust (Fig. 7) . This P-T environment has been referred to as the diamond storage area (Kirkley, M. B. et. al., 1991). The keel area is an optimal source for diamonds since fractures below the craton are more likely to tap this area and remain accessible to the surface.

kimberlites2020(5).png

Carbonate rich Peuyuk kimberlite from Somerset Island, Canada. From Andrea Giuliani.



kimberlites2020(5).jpg

Kimberlite from Bellsbank, north of Kimberley, South Africa. From James St. John.



kimberlites2020(6).jpg

Kimberlite from Premier Kimberlite Pipe, Cullinan, northeastern South Africa. From James St. John.



kimberlites2020(13).jpg

Hypabyssal kimberlite. From Reddit.



kimberlites2020(8).jpg

Hypabyssal facies kimberlite, Masontown, Pennsylvania. This kimberlite dike is enclosed by black shale. From Wyoming Diamond and Gemstone Province.



kimberlites2020(7).jpg

Diatreme facies kimberlite breccia from Lake Ellen, UP, Michigan. From Wyoming Diamond and Gemstone Province.



kimberlites2020(3).jpg

Tuffaceous, crater facies kimberlite from the Iron Mountain district. From Wyoming Diamond and Gemstone Province.



kimberlites2020(2).jpg

Large fractured chromian diopside (chrome diopside gemstone) megacryst in Sloan kimberlite from Colorado. From Wyoming Diamond and Gemstone Province.



kimberlites2020(14).jpg

Diamond in kimberlite. Bultfontein Mine, Kimberley, Baard District. From e-rocks.



kimberlites2020(15).jpg

Diamond in kimberlite. Bultfontein Mine, Kimberley, Baard District. From e-rocks.



kimberlites2020(1).jpg

Diamond (6.51 mm) in kimberlite. Bultfontein Mine, Kimberley, Baard District. From Geology for investors.



kimberlites2020(4).jpg

Diamond (6.51 mm) in kimberlite. Bultfontein Mine, Kimberley, Baard District. From Geology for investors.






Bibliography



• Brown, R. J., Manya, S., Buisman, I., Fontana, G., Field, M., Mac Niocaill, C., & Stuart, F. M. (2012). Eruption of kimberlite magmas: physical volcanology, geomorphology and age of the youngest kimberlitic volcanoes known on earth (the Upper Pleistocene/Holocene Igwisi Hills volcanoes, Tanzania). Bulletin of volcanology, 74(7), 1621-1643.
• Mitchell, R. H. (1991). Kimberlites and lamproites: primary sources of diamond. Geoscience Canada, 18(1).
• Mitchell, R. H. (2013). Kimberlites: mineralogy, geochemistry, and petrology. Springer Science & Business Media.
Nimis, P. (2009). Diamonds, Kimberlites and Lamproites. Geology Vol. IV, 154.