Ca isotope fingerprints of early crust-mantle evolution

Abstract The utility of 40 Ca/ 44 Ca as a tracer of pre-existing crustal contributions in early Archaean cratons has been explored to identify traces of Hadean crust and to assess the style of continental growth. The relatively short half-life of 40 K (∼1.3 Gy) means that its decay to 40 Ca occurs dominantly during early Earth History. If Archaean crust had a significant component derived from a more ancient protolith, as anticipated by “steady state” crustal evolution models, this should be clearly reflected in radiogenic 40 Ca/ 44 Ca ratios (or positive initial e Ca ) in different Archaean cratons. A high precision thermal ionisation technique has been used to analyse the 40 Ca/ 44 Ca ratios of plagioclase separates and associated whole rocks in ∼3.6 Ga (early Archaean) samples from Zimbabwe and West Greenland. Three out of four tonalite, trondhjemite, granodiorite (TTG) suite samples from Zimbabwe display initial 40 Ca/ 44 Ca ratios indistinguishable from our measured modern MORB value (i.e., e Ca (3.6) ∼ 0). Greenland samples, however, are very diverse ranging from e Ca (3.7) = 0.1 in mafic pillow lavas and felsic sheets from the Isua supracrustal belt, up to very radiogenic signatures (e Ca (3.7) = 2.9) in both mafic rocks of the Akilia association and felsic TTG from the coastal Amitsoq gneisses. At face value, these results imply the Zimbabwe crust is juvenile whereas most Greenland samples include an earlier crustal component. Yet the west Greenland craton, as with many Archaean localities, has experienced a complex geological history and the interpretation of age-corrected initial isotope values requires great care. Both felsic and mafic samples from Greenland display e Ca (3.7) so radiogenic that they are not readily explained by crustal growth scenarios. The presence of such radiogenic 40 Ca/ 44 Ca found in low K/Ca plagioclases requires Ca isotope exchange between plagioclase and whole rock during later metamorphic event(s). In addition the unexpectedly radiogenic Ca isotope ratios in some mafic samples reflect anomalous K/Ca ratios as a result of intense K-metasomatism ∼3.6 Ga. Thus Ca isotope measurements are not a robust tracer of crustal growth in the presence of intense tectono-metamorphic processes. Coupled with other isotope data, however, the degree of overprint can be estimated and the 40 Ca/ 44 Ca ratio of a little disturbed sample hints at a small contribution of Hadean protocrust in the coastal part of the Godthabsfjord area (Southwest Greenland). In the majority of Zimbabwe TTG samples, unradiogenic initial Ca isotope ratios point to very little prior crustal history and minor subsequent disturbance. We thus infer that the modest initial e Nd ∼0.8 of the Zimbabwean samples is representative of the depleted mantle at ∼3.6 Ga. Furthermore, Ca isotope systematics provide little support for a “steady state” model of crustal growth.