Which meteorites built the Earth? – Constraints from the Cu isotope signal of plate tectonics

Cover image: artists impression of the Hadean Earth (copyright Ron Miller)

I have spent my summer thinking about some of the biggest questions in Earth & Planetary Science:

Which materials built the Earth?

How has the process of plate tectonics served to mix as well as segregate layers of the Earth in terms of their composition?

Does does the answer to the second question inform the first?

It’s been great fun, and whilst big advances in science are hard to come by I certainly value this unique opportunity to spend my time grappling with such topics – it’s certainly given me a taste of what a career in academia would bring. Aiding my quest was my supervisor, the St Andrews Isotope Geochemistry Lab (STAiG) and my own wits.

My supervisor, Dr Paul Savage, is a specialist in non-traditional stable isotope geochemistry. This means that he is a world-leading expert in the science of using mass variants of a given element (in this case, copper or ‘Cu’) to trace magmatic, metamorphic and metasomatic (high temperature fluid) processes. His work has already yielded insights into the formation of Earth’s core, a process that separated metal and highly siderophile (literally ‘metal-loving’) from lithophile (oxygen-loving) elements on a vast scale around 4.5 billion years ago.

Many questions remain in this area and Cu isotopes are uniquely poised as a geoscientific tool to help us solve them. In particular, the cycling of sulphur (S) during the process of plate tectonics – the single most important aspect of terrestrial geology – must be constrained in order to understand how Cu isotope compositions have evolved over deep time. Without knowing this for certain, any models that try to match primitive meteorites (chondrites) to the Earth on the basis of their Cu will be forever doubtful (Fig. 1).

(Figure 1: the Cu isotope range of primitive meteorites and the offset between an model chondritic bulk composition for the Earth and estimates of the actual composition – processes involved in core formation may explain this discrepancy; adapted from Savage et al, 2015)

I studied an aspect of this problem by analysing Cu isotopes from rocks that sample the depths of a ‘fossil’ subduction zone. Subduction zones are places where dense oceanic crust sinks beneath a buoyant continent; this is the key distinguishing feature of terrestrial geology and acts to continually resurface the Earth, bring continents together and split them apart (the cycle of plate tectonics). When subduction beneath a given continent stops, usually due to another encroaching continental mass, a portion of the oceanic crust literally rebounds back up the subduction channel – a process known as obduction (Fig. 2).

(Figure 2: simplified illustrations of key tectonic processes – adapted from msnucleus.org)

Eventual collision of the two continents results in folding of the oceanic crust into layers far up in mountain ranges (‘ophiolites’) – such as the Zermatt-Saas ophiolite, which I visited last summer on a university-lead 4th year field excursion (Fig. 3). It’s a spectacular place and instantly fascinated me, so I jumped at the opportunity to work on samples from this area.

(Figure 3: The Matterhorn – this piece of once subducted oceanic crust now stands tall as an icon of the Western Alps)

The idea here is that, as rocks subduct down into the mantle, pressure and temperature conditions steadily rise. This often results in dehydration of the down-going slab, at which point the chemistry of those released fluids becomes important. If they are rich in S then, depending on whether we have sulphide or sulphate, Cu isotopes should be preferentially stripped from the sinking oceanic rock (as has been previously demonstrated for zinc – see Inglis et al, 2017) Overall, this would lead to an evolution of Cu isotope compositions in the deep and upper mantle (as well as the crustal rocks derived from it) over time. Our models of the initial Cu isotope composition of the Earth would therefore need to take this into account, with probable implications for which meteorites they predict to be the most likely building blocks of our world.

Using the world class clean lab and mass spectrometry set up of the STAiG lab, I set about dissolving my rocks in acid and stripping out all the unwanted elements from solution until all I had left was Cu (hopefully). This is a somewhat agonizing process that takes weeks and requires long hours in the lab, but as a geochemist at heart this was no great trouble. In the last few weeks of my internship, my supervisor and I ran the isolated Cu solutions on the Neptune mass spectrometer.

The early results are that subduction does influence Cu isotopes, although the story is complex and I am wary of making definitive statements just yet. I intend to continue working on this outside of my funded research time, compile more data and read ever deeper – hopefully by the time of the poster session later this month I will have a more complete story to share about the dynamic processes that govern our world and its origin, more than four and a half billion years ago…


Inglis, E. C. et al, 2017, Geochemistry, Geophysics, Geosystems, DOI 10.1002/2016GC006735.

Savage, P. S. et al, 2015, GPL, 1, pp. 53-64, doi: 10.7185/geochemlet.1506.

msnucleus.org – accessed 10/9/17


Leave a Reply