Rare volcanic rocks known as kimberlites are produced from magmas that originate in Earth’s mantle and then erupt onto the planet’s surface. These rocks have a violent eruption style, and a chemical and mineralogical composition that is unlike any other magmatic rock on Earth. In particular, kimberlites can contain centimetre-sized crystals of rare minerals such as garnets, zircons and, most notably, diamonds. Moreover, they have exceptionally high amounts of incompatible trace elements — those that preferentially enter a magma formed by melting of the mantle. These peculiar characteristics raise questions about the nature of the kimberlite source and its location in the mantle. Writing in Nature, Woodhead et al.1 suggest that all kimberlites originate from a single deep reservoir that has survived for most of Earth’s history.
There is a general consensus on several aspects of kimberlite formation. First, kimberlites must be extremely enriched in water and carbon dioxide to explain their violent eruption style and the presence of associated diatremes — conical or pipe-like structures that extend from Earth’s surface to depths of more than one kilometre. Second, some kimberlites must form exceptionally deep in the mantle, as evidenced by inclusions in kimberlitic diamonds of minerals that are unstable at the planet’s surface. These minerals include ringwoodite2, which is stable only in the transition zone between the upper and lower mantle (at depths of 410–660 km), and bridgmanite3, which is the dominant mineral in the lower mantle.
Third, in addition to containing minerals that crystallized from the ascending magmas, kimberlites contain a large assemblage of minerals and xenoliths (rock fragments) that were collected from surrounding material during the rapid ascent from the kimberlite source (Fig. 1). Some minerals, such as the diamonds that have ringwoodite inclusions, come from the deep mantle, some derive from shallower mantle and some originate from the planet’s crust.
By contrast, there is little consensus on the exact location of the kimberlite source in the mantle and, even more crucially, on the nature of this source. It could be a rather primitive material — one that has survived deep in the mantle from soon after Earth’s formation. Alternatively, it might be a material that was at some stage present at or near the planet’s surface and has since been recycled into the deep mantle. Both interpretations exist in the literature4 and a clear argument for the existence of the two types of source is the presence of two groups of kimberlites that have contrasting mineralogy and geochemistry.
Minerals in the first group, often referred to as archetypal kimberlites, have compositions of strontium and neodymium isotopes that resemble those of the primitive mantle. Those in the second group, commonly called orangeites, have much more enriched strontium and neodymium isotopic compositions that resemble those of continental materials4. The enriched nature of orangeites is usually attributed to interaction of the magmas with continental crust or the uppermost solid part of the mantle during ascent and probably does not represent the composition of the kimberlite source.
Woodhead and colleagues present a compilation of newly acquired and previously published neodymium and hafnium isotopic data, measured on archetypal kimberlites. These kimberlites cover a large age range, from less than 200 million years old up to 2 billion years old. The authors demonstrate that, over this long time period, kimberlites seem to always tap a source whose isotopic composition resembles that of the primitive mantle. This observation puts constraints on the nature of the kimberlite source, and favours a pristine reservoir — one that has survived untouched deep in the mantle for most of Earth’s history.
The idea that part of the deep mantle has remained isolated from its surroundings is supported by the discovery of traces of primitive material in volcanic rocks called ocean island basalts, which might originate from regions known as seismically anomalous zones that are found at the core–mantle boundary5,6. A primitive source has also been attributed to many other types of rock, such as granitoids7. The case for a primitive kimberlite source is bolstered by the evidence that this source is deep.
For the other rock types, a near-primitive isotopic composition might be explained by the presence of recycled crust in the rock source. Woodhead et al. dismiss this interpretation for kimberlites by arguing that the contribution of recycled oceanic crust would have had to have been constant over the two billion years of recorded history. Moreover, they suggest that the presence of high helium ratios (ratios of helium-3 to helium-4) in diamonds of some kimberlites indicates a deep source, close to the core–mantle boundary.
The authors’ interpretation might be correct, but a few independent observations need to be reconciled before the model can be applied to all kimberlites. For example, the presence of anomalous amounts of sulfur-33 in kimberlitic diamonds suggests that the source contains material that was present at Earth’s surface more than 2.5 billion years ago, when the planet’s atmosphere was not yet oxidized8. How this recycled material can coexist with the rest of the source is unclear.
Another potential concern is the unknown relationship between high helium ratios and isotopes produced by radioactive decay that are measured in diamonds. Some diamonds have low helium ratios, and strontium and lead isotopic compositions that are similar to those of Earth’s crust. But no strontium and lead isotopic data are available for previously analysed diamonds that have high helium ratios9. As a result, such high ratios might or might not trace a pristine deep source.
Finally, kimberlitic diamonds are plucked from the mantle during ascent, and the information that they provide might be irrelevant in terms of the kimberlite source. To confirm a pristine and deep origin of kimberlites, we need to demonstrate that the kimberlite magmas themselves have pristine characteristics, such as high helium ratios, tungsten isotopic anomalies that could trace interaction of the magmas with the planet’s core, and so on. A lot of work is still ahead of us.
Nature 573, 498-499 (2019)