Origin & Host Rocks for Diamonds

There are literally thousands of known kimberlites and many hundreds of lamproites and lamprophyres but only a handful contain commercial amounts of diamond. One estimate made many years ago suggested that less than 1% of all kimberlites are commercially mineralized (Lampietti and Sutherland, 1978). Although, many hundreds of new discoveries have been made since that paper was published, this statistic remains essentially valid.

Diamondiferous kimberlites and lamproites are essentially restricted to cratons and cratonized terrains. These include stable Archean cratonic cores (known as Archons) as well as cratonized Proterozoic margins (referred to as Protons) (Figure 4). Some unconventional diamondiferous host rocks have also been identified in cratons as well as outside cratonic terrains within tectonically active regions along the margins of cratons. Because high ore grades have been detected in some of these, unconventional commercial host rocks are anticipated to be found in the future (Erlich and Hausel, 2002). Presently, diamond exploration programs are designed to search for conventional host rocks (i.e., kimberlite and lamproite) or for placers presumably derived from these.

The North American Craton showing regions of favorability for conventional
host rocks. The diamond shapes are locations for reported diamond finds
From Hausel (1998).

Most diamonds are considered xenocrysts that separated from disaggregated mantle peridotite and eclogite during transportation to the earth’s surface in kimberlitic, lamproitic and some lamprophyric magmas.  Kimberlites, lamproites, and lamprophyres tend to occur in clusters of a few to more than 100 occurrences.  Structural control is thought to be important in the emplacement of these, and several structural orientations are often recognized within each district.

Kimberlite

The majority of diamond mines are developed in kimberlite such as the Wesselton, DeBeers, Kimberley, Dutoitspan and Ekati, or in placers, particularly beach placers along the west coast of Africa.  Lampietti and Sutherland (1978) reported only about 10% of the known kimberlites were mineralized with diamond. This statistic may no longer be valid in that as many as 50% of kimberlites found in Canada and Wyoming in recent years, and possibly as many as 90% in Colorado have yielded diamond. Even so, only a very small portion are commercially mineralized.  When economic, kimberlites may contain hundreds of millions to billions of dollars worth of stones, thus kimberlite should be a priority target in any exploration program.

Kimberlites are essentially carbonated alkali peridotites that exsolve CO₂ during ascent to the surface from the earth’s upper mantle (according to the EPA, such a volcano would be poluting our atmosphere and require taxation under tax and trade) resulting in diatremes with considerable brecciation and dissolution-rounded xenoliths and cognate nodules.  The diatremes appear as sub-vertical to vertical pipes that taper down at depth forming steeply inclined cylindrical bodies. The average angle of inclination of the walls of various pipes in the Kimberley region of South Africa (Wesselton, DeBeers, Kimberley and Dutoitspan) is 82° to 85°.  Ideally, the pipes have rounded to ellipsoidal horizontal cross sections filled with kimberlitic tuff or tuff-breccia.  Many continue from the surface to depths of 1 to 1.5 miles (2-2.4 km), where they pinch down to narrow root zones emanating from a feeder dike.

The Kimberley pipe, which was mined out by 1915 (about 20 years after discovery), contracted sharply at depth.  At the lowest level of mining (3,465 feet), it was no longer pipe shaped but rather had the appearance of three intersecting dikes (Kennedy and Nordlie 1968). Combined with the estimated 5,100 feet (1,600 m) of erosion since the time of emplacement, the depth to the original point of expansion was probably 1.5 miles (2.4 km).

Kimberlitic magmas are interpreted to originate from depths as great as 120 miles (200 km) and travel to the Earth’s surface in a matter of hours (O’Hara and others, 1971). The magma is thought to rise rapidly, possibly 6 to 18 mi/hr (10 to 30 km/hr), in order to transport high-density ultramafic xenoliths.  Within the last few miles of the surface, emplacement rates are thought to increase dramatically to several hundred miles per hour. Such velocities could bring diamonds from the mantle to the surface in less than a day. McGretchin (1968) estimated that the speed of the fluidized material near the surface increased to as much as 870 mi/hr (400 m/s), or about the speed of sound (Mach 1 or 331 m/s).  Some estimates have even suggested kimberlite emplacement at the Earth’s surface may have achieved velocities exceeding Mach 3 (Hughes 1982)! 

Contact between kimberlite & granite
shows no evidence of baking. The
contact is knife sharp and the adjacent
granitic rocks show no alteration. This
could only happen if the kimberlitic
magma was cool upon emplacement.

The temperature of the magma at the point of eruption is relatively cool (Figure 5).  Watson (1967) indicated a magma temperature of less than 600°C on the basis of the coking effects on coal intruded by kimberlite.  A low temperature of emplacement is also supported by the absence of any visible thermal effects on country rock adjacent to most kimberlite contacts.  Davidson (1967) suggested the temperature of emplacement may have been as low as 200°C based on the retention of argon.  Hughes (1982) pointed out that the near-surface temperatures of the gas-charged kimberlite melt may be as low as 0°C owing to the adiabatic expansion of CO₂ gas as kimberlite erupts at the surface.

Kimberlites typically transport xenoliths and xenocrysts to the surface.  Many of these are derived from mantle depths and some form a distinct suite of minerals that are referred to as kimberlitic indicator minerals. The traditional indicator minerals used to explore for kimberlite include pyrope garnet, chromian diopside, chromian enstatite, picroilmenite, chromite, and diamond.

Lamproite

Serious interest in lamproite intensified following the discovery of a world-class diamond deposit in olivine lamproite in 1979 in the Kimberley region at Argyle, Western Australia. Prior to this discovery, geologists around the world were focused only on kimberlite and argued that kimberlite was the only source rock for diamonds. It was essentially impossible to get exploration funding for projects other than those associated with kimberlite. This closed minded philosophy would have continued to the present if it were not for the fact that the drainage adjacent to Argyle had been filled with diamonds. The discovery led to the recognition of other diamondiferous lamproites in Australia, Brazil, China, Gabon, Zambia, Ivory Coast, India, Russia, and the US and also led to recognition that previous diamond producers in Arkansas and India were actually in olivine lamproite rather than kimberlite..

Scott-Smith (1996) subdivides lamproites into two general groups: phlogopite-leucite lamproites (~60% SiO₂) and olivine lamproites (>20% MgO, 35–45% SiO₂ and 7% K₂O) with abundant serpentine pseudomorphs after olivine. Instead of pipes with steep walls that slowly diminish in diameter with increasing depth, lamproites are characterized by “champagne-glass” vents filled by tuffaceous rocks, often with massive volcanic rocks in the core.

Lamproites appear to have formed under variable thermal gradients originating from depth and in some cases extending into the diamond stability field (Nixon 1995). A qualitative correlation between diamond and olivine in lamproite is confirmed in both the Ellendale and Kapamba provinces, in which diamond grades are consistently higher in olivine lamproites than leucite lamproites (Figure 6). When found, diamonds occur primarily in pyroclastic rocks; the magmatic phases are notoriously diamond poor owing to the high temperatures sustained in the flows during eruption (Scott-Smith, 1986).

Ellendale 7 diamondiferous lamproite
exposed in dozer trench. This commercial
pipe was hidden under a few feet of soil.

Where vents flare out, a potential for substantial tonnages exist in larger craters.  At Argyle, Western Australia, past reserve estimates of 94 million tons of ore at an average grade of 750 carats/100 tonnes led to its classification as a world-class deposit. Some of the richer portions of this deposit yielded grades as high as 2,000 carats/100 tonnes. However, large numbers of the Argyle diamonds are graphitized and partially resorbed; more than 60% are irregular in shape and include macles, polycrystalline forms, and rounded dodecahedrons. The largest Argyle diamond weighed 42.6 carats with the overall size of diamonds being quite small (average < 0.1 ct).  Nearly 80% are brown with the remaining stones dominantly yellow or colorless. Very significant are rare but economically important pink to red diamonds that bring Argyle fame.

At one point, Argyle mined nearly 40% of the world’s annual diamonds: by the end of 2000, the mine had produced an extraordinary 558,400,000 carats (Shigley and others, 2001).

Argyle diamond mine in 1986 shortly after discovery. This became the
largest producing diamond mine in the world, for many years.

Many lamproitic diamonds are relatively small and include common “fancy” yellow to brown stones. For example, macro diamonds (>1 mm) from the Ellendale field in Western Australia are dominantly yellow dodecahedra with many micro diamonds being colorless or pale-brown, frosted, step-layered octahedral (Shigley and others, 2001).

Placers

Because of a relatively high specific gravity (3.5), diamonds are often found in secondary stream or marine placers with other minerals of relatively high specific gravity such as magnetite, spinel, ilmenite, rutile, garnet, gold, etc.  Historically, there have been many reports of gold prospectors finding diamonds while searching for placer gold.  Examples include California, Colorado, Georgia, North Carolina and Wyoming in the United States and New South Wales in Australia (Hausel, 1998). 

Placer diamond deposits formed throughout geological history as is evident by diamonds in ancient Proterozoic paleoplacers in the Witwatersrand metaconglomerates of South Africa and the Snowy Range Group in the Wyoming Province, United States, as well as modern placers along the Atlantic Coast of Africa and Smoke Creek near the Argyle mine, Australia. 

Because of extreme hardness, diamonds can transport great distances in fluvial systems with little to no evidence of wear. Some of the more productive deposits include stream and marine placers where a large percentage of diamonds are gem-quality owing to fracturing and disaggregation of imperfect industrial diamonds during stream transport. Considerable numbers of diamonds have been mined from stream sediments along the Orange River basin in southern Africa and continuing in beach sands down current from the mouth of the Orange River along the Atlantic Coast.