Among the more curious materials that occur naturally on the Earth are volcanic rocks called kimberlites. Unusually rich in magnesium, many of them also host notable quantities of elements as exotic as lanthanum. Kimberlite magmas are generated at great depths, and it seems that they just cannot wait to reach the Earth's surface: they literally drill their way up to form pipe-like bodies. And, if these peculiarities were not enough, they are the primary source of those dazzling objects that have been the source of much fascination and conflict — diamonds.

Credit: VADIM KAMENETSKY

The Udachnaya-East kimberlite pipe in Siberia, less than a kilometre in diameter, contains an unusually high content of sodium and chlorine, far greater than other rocks of its kind. Kimberlites are rather susceptible to alterations that modify their original composition. Even though this pipe appears fresh there has been some debate about whether the saltiness of the magma that fed this pipe was inherited at the mantle source or imparted by the incorporation of sediments near the surface. Isotopic data obtained by Vadim Kamenetsky and colleagues seem to seal the case in favour of a salty mantle source (Earth Planet. Sci. Lett. 285, 96–104; 2009).

A mineral called perovskite received particular attention from the researchers. Crystals of this mineral intermingle with those of salt in the Udachnaya-East kimberlite. Perovskite forms directly from the parent magma of kimberlites, and its isotopic composition is hence expected to reflect that of the kimberlite magma. If the magma composition were to be modified by incorporation of, say, salt-rich sediments, the isotopic composition of perovskite would change too — it would then fall somewhere in between that of the parent magma and the contaminant.

It turns out that the strontium isotopic composition of perovskites from this particular kimberlite is very close to what would be expected for uncontaminated, primary magmas. This is borne out by the mineral relationships in the rock and various diagnostic elemental ratios, which reveal that perovskite and salt minerals grew from the same magma. When looked at in light of previous geochemical work by Kamenetsky's group (for example, J. Petrol. 49, 823–839; 2008), it seems unlikely that the salt in this kimberlite had anything to do with the addition of crustal material.

The violent and rapid rise of kimberlite magmas to the surface is traditionally thought to be driven by large quantities of dissolved water. Intriguingly, the Udachnaya-East kimberlite magma shows evidence of having been poor in water. Perhaps, suggest the researchers, the answer is in the salt: it is the salty composition that imparts the high fluidity and propensity to rise up, both of which are required for kimberlite emplacement.