Kimberlites (pipe-like bodies) are the host rocks for the majority of diamonds, and kimberlites are dominantly located in regions of ancient continental crust.
However, only a small proportion of kimberlites are diamond-bearing. It has been long realized that diamonds are not inherently part of kimberlite melts but are “stowaways.” They reside in the mantle roots of ancient crustal blocks and are grabbed by ascending kimberlite magmas (see Figure). The potential for diamonds in kimberlites depends on their abundance and distribution in the deep lithospheric mantle roots beneath continental crust.
We recently published a paper that provides an explanation for why kimberlites in certain regions of the Earth would lack diamonds. (Ernst, R.E., Davies, D.R., Jowitt, S.M., Campbell, I.H. (2018). Can mantle plumes destroy diamonds? Earth and Planetary Science Letters, v. 502, p. 244-252). In this paper, we modeled the effect of a hot mantle plume arising underneath a region of ancient crust.
Mantle plumes are thermally buoyant upwellings from the deep mantle, and upon reaching areas of thin lithosphere (<100 km thick), they can partially melt and produce huge volumes of liquid rock of mainly basaltic composition termed Large Igneous Provinces (LIPs, for short) and they can also cause formation of kimberlite magmas under adjacent areas of thicker lithosphere (see Figure). [By the way, LIP events are of such a scale that they would cover the entirety of Canada (or the US) with liquid rock to a depth of at least 10 m up to nearly 8 km. LIPs are also associated with the breakup of supercontinents such as Pangea, and also older examples, with some types of major ore deposits (notably Ni-Cu-PGE), and with catastrophic climate change including mass extinction events where up to 90% of life on Earth was wiped out. However, don’t worry, LIP events are rare; they only happen every 20-30 million years. LIPs are also present on other terrestrial bodies, notably Venus and Mars]. Where portions of the rising mantle plume are stopped beneath thick lithosphere (>100 km thick) the plume transmits heat into the deep lithospheric mantle and the resulting increase in temperature can cause deep lithospheric diamonds to be turned into graphite.
A slightly younger pulse of kimberlite magmas rising through the lithosphere would therefore only see graphite and not diamonds and would reach the paleosurface as barren kimberlites (lacking diamonds). However, once the thermal pulse from the plume had faded (over about 100 million years), then the lower lithospheric mantle “root” would again be cool enough for diamonds to reform from the graphite. Subsequent rising kimberlite magmas would, therefore, encounter diamonds and bring them to the paleosurface, where they could be mined (and make romantic couples happy).
We have identified examples of both situations (loss of diamonds and regrowth of diamonds). For instance, in Siberia, 250-225 million-year-old kimberlites lack diamonds because of a prior (370 million-year-old) mantle plume event (that destroyed the diamond potential). In northern Canada, 1100 million old kimberlites are barren because of an 1140 million-year-old plume/LIP event. However, in the same area much younger 180-150 million-year-old kimberlites are again diamond-rich, because of the long period of cooling (between 1100 and 180 million years ago) allowed diamonds to reform.