Analysis of the platinum-group elements in a particular type of ancient volcanic rock provides clues about Earth's early history as well as a fresh approach to understanding mantle dynamics.
In a paper on page 620 of this issue1, Maier et al. provide a provocative hypothesis for those engaged in the study of Earth's evolution. Their evidence comes from measurements of the concentrations of platinum-group elements (PGE) in ancient volcanic rocks known as komatiites, which originated from deep within the mantle.
Maier et al. find that the contents of these elements in 3.5-billion- to 3.2-billion-year-old komatiites from Barberton in South Africa and Pilbara in Western Australia are lower than those in younger komatiites. To explain the difference, they propose that the deep-mantle source of the older komatiites was deficient in PGE. According to a widely accepted hypothesis, the mantle acquired these elements as a surface layer of meteoritic material, the 'late veneer', whose deposition terminated with the 'late heavy bombardment' 3.8 billion years ago. Maier et al. propose that alloys of iron and the PGE trickled slowly down through the mantle and — so their thinking goes — older komatiites did not receive their full dose whereas younger ones did.
For Earth scientists, the value of the PGE is that they provide unique information about the composition and evolution of the mantle. Three of these elements, osmium (Os), iridium (Ir) and ruthenium (Ru), normally behave compatibly — that is, they tend to be retained in the solid phase during partial melting of the mantle or crystallization of rock. The other three, platinum (Pt), palladium (Pd) and rhodium (Rh), as well as the geochemically similar element rhenium (Re), are incompatible — they partition preferentially into the melt. This contrasting behaviour leads to changes in the proportions of the two types of element, which can be used to monitor how the mantle melts. The compatible isotope 187Os is produced by radioactive decay of incompatible 187Re, and the resultant changes in the ratio of radiogenic 187Os to stable 188Os can be used to establish the timing of mantle melting.
Basalt, the most common mantle-derived magma, is the product of 'low-degree melting'. Under these conditions, small amounts of sulphides and alloys remain in the residue of melting, as do most of the PGE, which are all highly compatible with these minor phases. Komatiitic magmas form by high levels of mantle melting, when even the refractory sulphides and alloys enter the melt, and the komatiites thereby acquire high concentrations of all the PGE. Analyses of PGE have served to date komatiites and trace the composition of their mantle source2,3,4, but until the work by Maier et al.1 temporal changes in their PGE contents had received little attention. Although there was some evidence that the oldest komatiites contained a lower content of PGE than younger komatiites5,6, a systematic analysis had not been done.
Maier et al. chose Pt and Ru to represent the overall behaviour of all the PGE. They show that the concentrations of these elements are lowest in 3.5-billion-year-old komatiites; are slightly higher in samples aged about 3.2 billion years; and are higher still and relatively constant in komatiites between 2.7 billion and 90 million years old. Having considered and rejected alternative explanations for the increase in PGE levels, the authors opt for an explanation that invokes the late veneer and a protracted timescale for mixing of PGE into the mantle.
It is generally thought7 that the PGE from the late veneer of meteoritic material mixed rapidly through the mantle, but Maier et al.1 suggest that this material took several hundred million years to percolate down to the lower mantle. Working on the principle that komatiites formed through melting of mantle plumes that rose from the base of the mantle, they propose that, 3.5 billion to 3.2 billion years ago, this part of the mantle had not received its full quota of PGE. By 2.7 billion years ago, trickle-down was complete, and from then on komatiites were derived from a homogenized PGE-rich source.
But how firm are their data? To measure PGE content, Maier et al.1 used a technique called the Ni-sulphide collection method, which may underestimate Pt contents by 15% or more. But they present convincing evidence that their analyses are sufficiently accurate for their purposes. Data obtained using a more precise method — the Carius-tube approach — yield comparably low PGE contents in some Barberton komatiites5. Moreover, any systematic bias should equally affect komatiites that are both older and younger than 3 billion years.
The bigger picture of komatiite geochemistry has to take into account studies of neodymium (Nd) isotopes8. These analyses tell us that the mantle source of komatiites had become depleted in incompatible elements very early in Earth's history, perhaps during extraction of the melt that formed the first continental crust9. The extraction of this material should also have depleted Re (which is incompatible) more than Os (which is compatible), leading to an Os isotope ratio that is distinctly lower than that of the original, undifferentiated mantle7. Yet the Os isotopic compositions of most komatiites are almost identical to that of the meteoritic material that we accept as representative of the mantle. Maier and colleagues' results provide an explanation for the contrasting nature of the two isotope systems — the Nd isotopes in komatiites retain a record of differentiation during early melt extraction, but their Os isotopic composition was swamped by later input of meteoritic material from the late veneer.
Maier and colleagues' work sets the scene for analysis of the trace-element and Nd and Os isotopic compositions of komatiites, allied with modelling of mixing of meteoritic material into the lower mantle. Such a combination has the potential to set important constraints on the evolution and geodynamics of the early Earth.
Maier, W. D. et al. Nature 460, 620–623 (2009).
Luck, J. M. & Arndt, N. T. Terra Cognita 5, 323 (1986).
Keays, R. L., Ross, J. R. & Woolrich, P. W. Econ. Geol. 76, 1645–1674 (1981).
Puchtel, I. S., Brügmann, G. E., Hoffmann, A. W., Kulikov, V. S. & Kulikova, V. V. Contr. Mineral. Petrol. 140, 588–599 (2001).
Puchtel, I. S., Walker, R. J., Anhaeusser, C. R. & Gruau, G. Chem. Geol. 262, 355–369 (2009).
Maier, W. D., Roelofse, F. & Barnes, S.-J. J. Petrol. 44, 1787–1804 (2003).
Walker, R. J. Chemie Erde 69, 101–125 (2009).
Arndt, N. T., Barnes, S. J. & Lesher, C. M. Komatiite (Cambridge Univ. Press, 2008).
Harrison, T. M. et al. Science 310, 1947–1950 (2005).
About this article
Contributions to Mineralogy and Petrology (2016)
Journal of Petrology (2011)